The hilarious thing to me is that Universal format for PowerPC never went away. So you can actually compile one binary for PowerPC, Intel and ARM and they’ll work on like every Mac to ever run macOS lol
@@GaryExplains I think Apple is being disingenuous regarding the performance of there SOC, when they compare its performance to a mac laptop, it may look amazing. However when one consider the significant thermal and power handicap these computers have, and you start looking at it from the wider PC industry, its not as impressive. Earlier this year I ran Cinebench on friends brand new i9 equipped MacBook pro, it was beaten bey my I7 t490 and utterly destroyed by my Zephyrus g14. The comparison with the mac mini, is with a three year old i3 and that is a specific processor ( i3-8100B) , when you start looking at it from that point of view, and knowing that manufacturers will give the numbers (and graphs) to show their products in the best possible light. You realize that all the what apple say don't really matter until you have a device and having a real world experience. There is a part of their presentation i think they straight up lied, Rosetta running x86 apps on ARM faster than native, nope, wont happen. Thats BS. Still I cant wait to get one, my son loves mac OS
@@nicholaswjamrock I don't know if PC users are in some deep denial mode or what it is but these things have been tested independently several places now. Benchmarks not produced by Apple has shown Rosetta beating all the x86 Macs running x86 binaries. Not hard to see why given that M1 is that much faster than their existing offerings. Anyway you are missing the mark if you try to compare the M1 to the the biggest most powerful computers the PC world can offer. M1 is for their low end stuff and it is already beating top end Macs. The point is that each individual core on the M1 is more powerful than anything else out there, certainly per watt. That means it is relatively easy for Apple to destroy the top of the line PCs in the future. They have a huge thermal budget they can use if they want. They can add in a ton of more cores. They have faster cores AND can add more of them. The M1 simply shows what is possible with Apple Silicon. When the M2, M3 or whatever they will call it is realized, the top end of the PC market will be conquered. Dell, HP, Lenovo etc will have their iPhone moment.
@@erikengheim1106 I am not a PC user, I am a computer user, I have used all major architectures and operating systems released over the past 30 years, my first mac was a Wall Street. I own several macbooks and i own a working powerbook, that i bought in 2006. I am not anti- Apple, i a just a pragmatist. With the exception of my zephyus, i only buy laptops types of laptops thinkpads and macbooks (powerbooks). 1. where are the benchmarks you speak of?? 2.you said Rosetta beats X86 macs, ---- Now I am in a unique situation, i have thinkpads and macbook pros with the same processor, i used these thinkpads as hackintosh and they were faster, way faster. Now may be asking why?, let me explain, apple for the last 5 years have being optimizing their laptop based on the limits of power delivery and cooling system that can be accommodated by the physical size of the laptop. this is a fact and has been proven my reviewers and independent publication. 3. if you take a look into how ARM instruction are structured and executed, vs x86, there is no way an emulation can out perform bear metal, sorry that's just not so. when i was younger , i used to emulate old game console on my pc, i was running on hardware which on paper was up to 5 X faster than the original consoles and it was not always pretty. 4. i did not compare the M1 to the the biggest most powerful computers the PC world can offer, Apple did. when they said its faster than 95% of all windows laptops, to which they may be correct, but what i was looking at was the pro level laptop and laptops within the price range of apple offerings. when compared to those its not that impressive. 5 Now regarding the M1 Beating top end macs, "www.apple.com/mac/m1/" follow the link and read the fine print at the bottom. 6. One thing i agree with you whole heartedly , the performance per watt is going to be amazing 7. some thing you should understand about, electronics, performance and power consumption dont scale linearly, eg a cpu that consumes 5 watt at 2 gHz may just consume only 1 watt at 1 gHz, so the long battery life that was talked about will only be available under certain conditions, and this applies to all electronics. 8. I am not making speculations on future apple SOCs, i am talking about what they presented and how it was presented when compare to what it on the market right now. Its is possible two years from now apple may have the best processor on the market, I remember the days when the fastest laptop you could buy was a power book, no gimmicks no marketing BS, it was just flat out the best. I dont buy marketing, i buy performance. you may think i am against apple, i am not i am an apple product user for over 25 years and i am proud of that, but i am will never be blinded by marketing or hype. I want this to succeed, not because of apple, but because when ever apple starts a trend the entire consumer electronic industry follows, (blue tooth, wifi, high resolution displays, long battery life, the ultra portable laptop, etc....) . Apple being successful will show other manufactures that is possible to make a compelling ARM based device, and being that are devices are usually cheaper to own and uses less power. it would allow more people to have access to modern computing. Even though there are companies already Offering computers that runs linux and windows based on the ARM architecture they are not good value for the average user, Apple is in a position to show them how its done when its done properly.
@@nicholaswjamrock Problem is, that it makes no sense for you to be a Mac user given your views. Why would you buy products which according to you are clearly inferior? Schizophrenia? 1. Here is an example. They run Geekbench under Rosetta 2 and still get great performance. www.macrumors.com/2020/11/15/m1-chip-emulating-x86-benchmark/. 2. I think Apple mostly makes the right choice in focusing on battery life and portability. I honestly think Macs have had high enough performance for 90% of my needs for at least 10 years now. Yes for some very compilation heavy tasks I got a PC some years ago. But I don't do that often. 3. Rosetta 2 isn't emulation per say though. It performs a translation of x86 code upon installation. So you are running ARM code. Of course there is a performance hit of about 20-30% or so, running ARM code that was no properly optimized. Yet given that M1 is so much faster than existing Macs that still puts them ahead of the pack. Emulation has come pretty far since your gaming days I think. Rosetta 1 worked pretty decently as far as I remember and it was not as good as Rosetta 2. 4. Do you have benchmarks showing that most pro level intel laptops beat the M1 Mac Pro performance wise? 5. I don't need to read the fine print on Apple's page, as this is what independent non-Apple benchmarks show. 7. But that is kid of the big deal about the M1. They can increase its performance a lot with much lower watt increases than what is required for intel chips. Basically they have way more potential to jack up performance. Hence it is premature to write off Apple. 8. "Best processor" by what metric? You move the goalposts around so much I am not sure what you are really talking about. You talk about Apple performance down due to thermal constraints in the laptops. Sure... but if ask how powerful is the M1, then that is a separate question from how powerful is Mac Book Air with an M1 inside and no cooling. You seem to conflate the two a lot. Apple clearly has a high performance chip, whether they are choosing to utilize the full performance of that or not is a different question. --- Not sure what kind of Apple user you are. You seem to only care about the performance. If performance is the only thing somebody cares about, then I fail to grasp why they would be a Mac user at all. Macs are about the total experience. I use a Mac because of a combination of ease of use, elegant design, access to a rich Unix eco-system in a system that just works. I switched from Linux because I got tired of spending so much of my time configuring stuff, and lacking access to quality GUI apps. Using a Mac has saved me time due to superior experience not due to superior peformance. I remember years back when doing home video editing in iMovie. My brother did the same on a high performance PC, but it took him ages to get stuff done. Horrible interface. Stuff didn't work. Softwae would crash. It was frustrating experience for him, as well as for most PC users at the time it seemed. I am sure PCs are better at this today. But certainly when I made the switch to Mac during the OS X switch a lot of stuff like video editing just worked a lot better on Mac. And the system was no prone to crashes like Windows PCs at the time. --- I am glad you are positive about the future, but I don't get why the M1 doesn't impress you then. Is it not way beyond what Micorsoft has offered in the same space? If it beats most existing intel laptop Macs, then I don't get why you don't consider this a major achievement. I wrote a prediction 4 years ago that Apple would switch to ARM. In that prediction I took it for granted that their new ARM laptops would have lower performance than their intel laptops. My rational for ARM still being worth it, was that I assume consumers would prefer slightly less performance with longer battery time, and cheaper laptops. Apple completely blew past my predictions. Not only do they offer these computers cheaper, than they did before and with longer battery life, but also with significantly higher performance. Something I had deemed impossible to achieve with ARM. I just don't comprehend what kind of fantasy world you live in where you expect even more than this.
Why? Nothing is better for the development and customers then competition. I am user of Ubuntu so M1 does not much my needs. Please be more "polite" for the innovative work the Intel delivers (and others). All companies which contribute in innovation, development and improvement of human live admire our applause. Have a nice day.
Since this is only gonna benefit Apple, the comparison to Intel is gonna be pretty moot overall. Intel isn't going anywhere. What this shows is the fact Apple switched 3 times to many the fans keep spending money, just shows how power the x86 platform is. Not has been able to replace it. Sure AMD have made some strides towards faster performance per clock cycle, but it isn't significant enough for desktop oems to make a complete move. Since Apple isn't gonna sell its chip to 3rd parties, then PC oems will have no choice but to still choose AMD or Intel. No matter how good this M1 is, the Mac still be be more limited than a Windows PC. Go faster to nowhere is not better at being slower going somewhere. Like Gary said, for users of the Mac who need real power, they should wait for later versions for the iMac and Mac Pro, because the M1 isn't gonna be powerful enough handle heavy sustained tasks. His claim about Rosetta not taking that much of a performance hit is to me not very true. A virtualize environment or an emulator can't run exactly as nature hardware. Maybe its OK for level apps like Office, but for PS/ FCP? No way!
@@TechieXP "FCP will be slow and Rosetta will be garbage!" Source: "Dude trust me" Quite different to the testimonies from people who had their hands on the DTK and said "the DTK was running x86 code faster than x86" ARM won't replace x86, at least for right now, but it will be the de facto alternative for the majority of people who want lighter and longer-lasting laptops. Only Apple's core, though, do I see really taking on the desktop. They were not lying when they said it was the "fastest core in the world" seeing the preliminary benchmarks, it outpaces the best Zen3 core AMD has right now. Nvidia will be the "intel" of the ARM PC land, so since you hate being locked down, I hope you don't buy one of these Nvidia ARM chips when they come out and are comparable or maybe even better to x86 chips in years time.
Something to think about: The end of the Intel era also means a "more closed, tightly-controlled" environment. Good for Apple, but the consumer will have less and forced choices.
Thanks Gary. You consistently use enough detail to explain without getting bogged down in the weeds! Some video I would like you to cover the differences in Intel architecture to the Apple architecture, and how the M1 is conceptual similar to IBM's Cell processor and its adoption in Sony Playstation 3.
This is what I like and admire Gary, he explained the most complicated things when it comes to computing or computer as a whole even I don't really understand what he is talking as I'm not into computer so much but I like listening to him. How I wish I'm as knowledgeable as he is. He deserve a million subscription. Thank you Gary for sharing your knowledge, I enjoyed watching your video.😂😂😂👌 😀 🇵🇭
I remember few years back you mention in one video that we need more arm computers, that could be the push we need! I could see what aws/annapurna graviton could do 2 years ago in a very small module and I can't wait till we can get on an a78 architecture cpu from nvidia next year hopefully :)
You did it on a whole different level than the rest of UA-cam. I wonder when windows and other ARM based CPU manufacturers will catch up? Does AMD have a chance or x86 is just cannot be made so power efficient.
Apple has gone RISC before and went back to Intel from the Power PC. Apple does not give a F about speed. It's all about control. ARM is great for small devices but when you need power it doesn't have the capacity.
Ever heard of Galaxy book S? Windows on ARM is already there. The issue is third party app support for ARM, which is just not there yet. I think x86 still has a few years left before ARM takes over, despite its handicap of power consumption.
@@jordonberkove7438 Arm doesn't have the capacity? That is nonsense for two very important reasons. 1. The world's fastest supercomputer is based on the Arm instruction set. 2. There is nothing intrinsic to the Arm instruction set that makes it slower or faster than say x86. Please tell me why you think it is?
@@jordonberkove7438 The leaked benchmarks tell a different story. It seems very powerful. We'll know more very soon. Control is surely important for Apple, but just one factor.
@@zonyyu8412 The A14 is faster but with higher power consumption. Stock ARM cores are focused more on efficiency instead of outright performance. A performance-focused core design like the X1 should narrow the gap to Apple's Firestorm cores. That said, Apple made some intriguing design decisions that enabled the A14 to challenge Intel and AMD.
@@fallinginthed33p the firestorm cores are indeed not the most efficient cores out there. They are less efficient than the A77 found in S865, and about the same efficiency as the A77 found in S865+. Efficiency takes a hit when driving an 8-wide behemoth of a processor to 3GHz, where wider designs tend to prefer lower clock speeds. That said, the A14 CPU is more efficient as a whole compared to S865 CPU, and that is because the little cores are fast enough to be significantly utilized, and because those are MUCH more efficient than any of the big cores, overall efficiency is greater. The Cortex A55 are so slow that most programs simply switch to the big cores for execution, and that is why in recent years, we have seen “medium cores” to handle tasks that little cores cannot. All in all, efficiency of the CPU complex depends on the workload - how often the large cores are used, how often the small cores are used, etc. FYI: the little cores on A14 are as fast as the big cores in A9. Quite shocking how far CPU’s have come.
@@zonyyu8412 All this is probably why we aren't seeing 2x or 3x battery life for M1 Macbooks. The Firestorm cores are wide, deep, fast and hungry. Running them at full speed will consume x86-like levels of power but with more performance. The new ARM X1 cores are meant to be paired with A77 and A57 cores; I'm wondering if Apple will follow a similar high-medium-low speed design for later chips.
I did thank. I‘m curious how Apple might incorporate discrete graphics when they seemingly have made a great job of the integrated graphics. Is it just me or does it seem that the innards are becoming more modular as in, now just replacing one or two chips?
Before watching it I already know this video is the best explaining the technical aspect of the M1 chip while at the same time making it digestable for most people.
With the throughput being demanded from a portable computer only because it is portable instead of a desktop form factor, give me an inch thick, 5 lb macbook pro that doesn't scorch my lap or the power cord AND with the extra 3/4" of thickness, give me so much more performance it'll bring me tears of joy. You can make the Mac Pro as big as it was in 2013 and pack it with paralleled processors on paralleled 4 layer pcbs, memory, and cooling fans on a computer that will finish rendering all my 8k video before I have the chance to click the Render button.
Amazing explanation as always, I just wonder what Qualcomm, AMD, Intel, Samsung, even ARM itself is going to do next, I just hope the integrated ram doesn't catch up or maybe it can be hybrid system because the idea of having to select the amount of ram at the moment of the purchase without having the option of upgrading it's just awful (unless they give you way more than enough on the baseline... which is not the case) which also makes you wonder, what are going to do about the Mac Pro and discrete GPU'S
The Firestorm M1 cores all run at 3.2ghz. The only difference between the low end MacBook Air and the other M1 computers is one less graphics core (probably binning to recover some SoCs with a bad GPU core). The Air will thermal around the 9 or 10 minute mark if heavily pushed.
intel is not going to throw the towel; dont forget that in 2000 apple was at the edge of bankrupt with their slow underperforming overpriced powerpc cpus (by ibm + motorola); they moved to intel (after a decade of market bashing the pentium in their bullshitty comercials) and they gained super performance and reliability, even making it possible the birth of the hackintosh; let's see if apple's ARM is on pair with the all-mighty Xeons of todays in the power desktop and server fields
@@alerey4363 apple will not and will never be competing on the server side business unless they have really good reason to do that, Apple is more towards consumer computer business and software.
@@reobating8826 www.apple.com/macos/server/ you were saying? ah, and let me remind you that the XServe was a failed line of blade servers apple sold in the 2000s
@@alerey4363 this will be true pre iPhone era, of course Apple has server side business but what I meant to say is that they never will compete seriously in that market just because they do not want to focus on it.
@@reobating8826 they can't compete with industry standard HP, Dell, IBM servers but they'd want to, given they still SELL their OS SERVER software, to be run on their *INTEL* macpros (you know the dreaded trashcan, heater, whatever u call it); so if they don't make macoserver runnable on their new ARM platform then intel cpus are still KING in that domain (and prosumer desktop too); so long at launch apple only showed glorified versions of the iphone/ipad pro processors embedded into macbook chassis and macmini case, nothing more, and oh yes..."one more thing": those ridiculous graphic charts with curves showing "performance" without even a real data compared to real world computers..well, it's apple, it always did the same, zero specs, pure adjectives (this is the best notebook apple has ever produced, fantastic speed, blazingly fast performance, super long battery life and the bla bla alike)
Geekbench results show all three machines running at 3.20 GHz. They seem like they are the same chip running with different cooling solutions (and differences in GPU binning).
Thank you for the video. An off-topic question from a foreign language teacher: Do you prefer "Apple has" (which is what you say at the beginning) or "Apple have" (which you say at 1:52 and 16:21)? Is there a difference or is it just the notion of collective nouns vs plural nouns, i.e. just as "the crowd is cheering" (looking at a crowd as a whole) or looking at it as "every single member of ..." -> "the crowd are cheering"?
The correct way is "Apple has", "Apple is" etc, as Apple is a company, a legal entity. However I often think about the engineers so really I am saying "the Apple engineers have" or "the Apple engineers are" and that gets shortened to "Apple have" etc.
Hey gary could have it been possible to translate x86 instruction set to risc at hw level or is it limited to intel licencing we have wine that translate nt system calls to unixs we have Rosetta that translates x86 to arm how about using a hw level translator
Translating it at a HW level basically means you are building a x86 chips and that requires a license from Intel. Transmeta tried it and the rumor is that Stexar/NVIDIA tried it with Project Denver. According to Charlie Demerjian, the Denver CPU was originally intended to support both ARM and x86 code using code morphing technology from Transmeta, but was changed to the ARMv8-A 64-bit instruction set because Nvidia could not obtain a license to Intel's patents.
Why does it support only two thunderbolt ports? Does the M1 need to have two thunderbolt controllers to support four thunderbolt ports? Also it doesn't support eGPUs or more than one display, does this mean the M1 doesn't have enough bandwidth for those? How will the performance of the SoC be affected if the ram & gpu are separated? Which I suspect they'll do for mbp 16", imac, imac pro & mac pro
i’m trying to decide which macbook pro to get (either intel 2020 or M1 2020) and i’m stuck on which one to get. is M1 reliable at this time considering it is a brand new chip?? and i need to run adobe apps as well, will M1 support my adobe photoshop and premier pro??
It is not really brand new though. It is a continuation of the A14 chip in the iPhone. People don't ask if the next iPhone chip is reliable. I think it is more of an issue if there will be x86 software that will all run as expected. Apple was showing Adobe Photoshop getting ported. Not sure about Premier Pro. I am personally not in a hurry. I am still using my 2013 Mac Pro model and still don't have performance issues. I suppose one could just wait a few more weeks and see what kind of experience people have with these new Macs.
Jonathan Morris and Linus Tech shots fired! Jonathan shows below a new iPhone 12 Mini exporting a 4K video in seconds while his Intel based i9 Mac is still rendering minutes later.
The iPhone is using hardware based encode while the iMac is using software based encode. You should always use software based encode unless you’re live-streaming and your system cannot keep up.
Thanks Gary, great video! The bit that puzzles me about the M1 is the Unified Memory Architecture (UMA). These new machines support either 8 or 16 GB of memory, so if I buy a Mac mini with 16GB is the M1 I get physically different or is Apple handling a memory upgrade with some kind of clever software switch?
The defacto-method to improve IPC is to slap a large L2 cache in the die.Nvidia did it AMD did it and also Apple did the same. I wonder if the desktop version of these chips will completely ditch the power efficiency cores and just use a homogeneous architecture all chips running at the same clock speed.Also something like a 16 or 32 core chip would probably be quite powerful.
What Apple has done is extraordinary... No other ARM licensee, or manufacturer using chips from some other ARM chip maker, has yet matched the sophistication and pace of Apple's silicon engineering. And it hasn't been for want of trying. Samsung dumped massive investments into an effort to build its own M series chips before giving up. Furthermore, Huawei and other Chinese makers have manufactured ARM reference designs under their own brand names, without achieving Apple's results. Nvidia desperately tried to beat Apple with its Tegra chips before throwing in the towel in smartphones. Qualcomm has been embarrassed by its fall from mobile chip leader to merely a runner up by Apple ever since it was caught flat-footed by the 64-bit A7 back in 2013. Intel once spent billions subsidizing Android licensees to produce tablets with its mobile x86 Atom chips before giving up on mobile. And now, Apple is taking its mobile expertise developed to power iOS devices and using it to drive its notebook and desktop Macs, starting with the MacBook Air, 13 inch MacBook Pro, and Mac mini. The company expects to complete a transition over the next two years. That would be hard to believe from anyone else, certainly after Microsoft's eight-year struggle to make little progress with Windows for ARM. Apple's move to its own SoCs isn't just a chip migration. It's a radical rethinking of desktop PCs that leverages massive amounts of engineering work already completed to deliver iPhones and the light and thin iPad platform, to make them ultra-responsive, radically mobile, and blur the line between hardware and software.
I’m excited for the Mac Mini! I keep breaking iPads by running games and keeping them in a kiddy case with handle so they only last me 2 years 😕😩🌱 they thoughtfully put a fan in the Mac Mini 😀
Is this wide execution architecture only possible because it's a RISC based processor? Or would it be feasible to create something similar in a CISC based processor? Because as far as I understand, a CISC processor splits its instructions into smaller ones, so maybe that makes it harder to do this kind of parallelization...?
I have a question about RAM and the M1 processors. I have seen several videos wherein iPads Pro have no trouble editing and rendering 4k and even 6k video often faster than powerful desktop computers. I have also seen videos of people playing demanding video games (IIRC, Fortnite before the recent kerfuffle) with no trouble. The iPads Pro featured less powerful processors operating in a more thermally constrained physical environment than a laptop or, especially, a Mac Mini. They also had less RAM, 6gb vs 8 or 16gb. Is it possible that the transition to Apple Silicon might require us to rethink the issue of how much RAM we need, especially with apps that are optimized for these processors and the operating systems? Thank you for your wisdom.
Video playback and rendering can be accelerated through specific hardware blocks so you don't need to use the CPU. It's a lot more efficient and faster that way. I don't know how this would affect RAM usage but Apple are leaders in image and video processing in hardware.
@@fallinginthed33p However, even the casual video editor should avoid using hardware accelerated video encoding. Software encode is superior. I hope apps updated to be native to the M-series SOCs allows the user to not use the specialized hardware and solely use the CPUs to encode video.
I think it will be very interesting to see what they come up with for M1X and M1Z... Anyway, I believe that they also have different TDPs on MacBook Air and MBP13... maybe 12W in MBA and 25W in MBP? Or maybe 9W in MBA and 15W in MBP... Anyway, I really dig these ARM chips, and I hope that Qualcomm/Samsung/Nvidia are able to put out a true competitor to this M1, although IMHO it'll be harder to take advantage of the power optimizations on Windows machines, even if they use only ARM-compiled apps... That said, I don't quite like what degree of control has now Apple on his entire hardware... Look, the not-upgradable RAM on MacBook is not surprising to me: I don't like it, and I personally buy only user-uogradable devices, but it's nothing new, and there are plenty of Windows devices on which the RAM is soldered on the motherboard itself... if I'm gonna have soldered-on RAM, I'd better have it directly linked or soldered on the CPU, as it guaranteed lower access latency to it (which is extremely useful, especially for on-board graphics). What concerns me is the fact that RAM is now not-user-upgradable even on the Mac Mini... I'm worried that they will make RAM soldered out of the factory even on iMacs! The 2 things that make me uncomfortable the most, though, are the lack of support to eGPU, the lack of support to Windows and the possible lack of support of any other operative system that's not macOS Big Sur or newer: we are used to iPads being tablets, and therefore being stuck on their OS, but MacBooks and Mac Minis are still PC nonetheless, and I expect the user to have control on what runs on its machine. But Apple doesn't want to give users this freedom, because every single Apple device is, to them, just a way to get users in their software ecosystem, and get money just by gathering data, selling apps on their store and sell people services on subscription (SSaS)... THIS I don't like!
@@tri-a7655 I partially disagree... 1) I'm aware of other people working on ARM transition, but let's be honest. Firstly, if we want to take ChromeOS into consideration, we'd rather also consider other Linux-based systems, which have been running on ARM for years already. That said, ChromeOS is more similar to iPad OS than it is to a true computer operative system, like Windows and macOS: it doesn't have the flexibility and the power of Windows, macOS, Debian, Ubuntu, Arch Linux and other stuff like that... And I'm aware that there are already devices with specific Qualcomm and MediaTek ARM processors around (Surface Pro X, some Chromebooks...), but to me the Qualcomm and MediaTek options seem a lot weaker than Apple M1 in terms of power and efficiency. Moreover, the Windows, ChromeOS and Linux ecosystems don't have the same degree of vertical integration. This means that Apple is able to better optimize the experience, from hardware down to software, while it's harder to have a truly optimized Windows machine. This means that you have to account for losses in efficiency from the lack of optimization. So if Qualcomm and MediaTek truly want to compete in this space, they need an even stronger offering than Apple, not a weaker one. For now, they are a step behind. 2) Please, don't say soldered RAM is OK... lack of user upgradeability, serviceability and customization is ALWAYS bad. It's a compromise we are now accustomed to because it gives us thin and light designs like the XPS 13 or the MacBook Pro 13, but it's not idea. To me, it's unacceptable on bigger, more powerful and pricier machines like MacBook Pro 16, iMacs and other non-Apple computers... You can't say that it's the same as an high end phone... firstly, in phones there's not the actual space to put user-upgradable RAM, so the compromise HAS to be done. Moreover, phones DO NOT behave like computers, even if they'll happen to have the same CPU designs... on high end phones, the need for RAM usually comes from the habitude of keeping many many apps in the background, many of which need to constantly update informations for notifications. But the single apps themselves do not use all that RAM, and there are very very few apps (mostly games and foto/video editin apps) that actually need much RAM at all. On PCs on the other hand, you still have the multitasking factor, maybe even stronger. But on PCs, if you multitask, you probably are more active on those apps (unless they are messaging apps like Whatsapp Desktop, Telegram, Discord...). And, most importantly, there are many more apps and use cases that actually NEED as much ram as possible (batch editing in Lightroom, Photoshop editing, video editing, music production, simulation, calculus...). You can optimize how the system manages apps and programs in the background, but you can't optimize all that much all those program that need juice... and before saying that those chips are not supposed to be used for that, remember that Apple is transitioning its entire lineup to ARM and that they've already presented the MacBook Pro 13, which is aimed at professionals... Moreover, in the lifetime of a phone, the kind of usage from the user doesn't change all that much, while the needs of an individual could dratically evolve in time. Think about a student who buys a laptop for school (and therefore needs 8GB RAM at most on macOS). This student may get into music production, photography or stuff like that down the line. When this appens, the processor of the laptop may still be capable of doing that, but the RAM may now hold him back. So the student may upgrade to a new laptop, even if his previous one would have still done OK if he could have upgraded the RAM... I know because I've been there: I went in 2/3 years from truly needing 4GBs on my laptop to now have 32GBs! "if best performing iPhones were sufficient with just 6GB of RAMs. Then this 16GB max should translate to much more capable than that." 8GB at the moment is the bare minumum for a decent Windows experience, and 16GB is what most users should have (unless all they do is basically only writing documents and reading stuff on the internet) if they use heavier programs. 32GB is actually a decent RAM amount for serious professionals, while truly demanding ones actually use 64GBs or more. I'm not saying that 64GBs are needed NOW on MacBook Pro, but I can see a professional buying a MacBook Pro 13 with 16GB of RAM encountering some limitations 2/3 years down the line. If all you need is more RAM, you should be guaranteed the possibility to upgrade it and not throw away your perfectly functioning laptop... And let's not forget that Apple is picturing themselves are ambientalists (using recycled alluminum, removing accessories to have a smaller iPhone box that is more efficient to transport...), while doing NOTHING for right to repair and right to upgrade. They are basically doing NOTHING to limitate e-waste...
@@tri-a7655 What I'm saying is that macOS is already more optimized than Windows (although I also use Linux and I find it to be blazing fast). Maybe macOS can be improved, but there's not much you can do to make heavy tasks use less memory. You can't do magic. Many apps could be absolutely improved, starting from the entire Adobe software suite (DaVinci Resolve runs better than Premiere Pro, for example...), but if you have 60 audio tracks on a project, there's not much you can do
The geekbench of the AIR and the Pro shows the M1 is running at the same exact boost frequency in both cases, 3.2ghz. I take it the better cooling in the Pro just lets the CPU sustains the boost longer.
Sir please explain in details How apple CPU and GPU better than other or different from other? In fully detail but in understanding language for class 11th student. 😫🙏🙏💓Pleased!!!
Now what is wanting is a full performance review on the new apple products that use this new cpu without that nobody can make a full decision about the potential hardware
Not an expert by any means but from what I’ve read and saw, the speed of cache generally gets slower, so L1 is fastest (I think operating at the same or maybe half the clock cycles of the processor), then L2 might operate at half that again and then L3 maybe half again. As they get slower they also generally get larger though as they’re cheaper to manufacture (being that they’re not as high performance memory). The new trend now is to then also make your previously very slow mass storage to be a kind of itself a cache (see AMD RDNA 2 graphics/PS5/Xbox Series X), leveraging the now very fast SSD speeds, there’s not so much of a penalty by paging memory to your mass storage. Of course this also usually has the drawback that your mass storage isn’t as mass as what it could be, so really SSDs are adding a memory/storage category rather than entirely replacing one.
Greetings from near Albuquerque, New Mexico, USA. Great explanation of the Apple M1 processor. A lot of interesting info in a short time. Your description of the "Ultra-wide execution architecture" sounds exactly like vector processing, which Cray and others (most notably the Cell processor, a collaborative result between IBM, Sony, and Toshiba) were doing decades ago. I wonder why Apple or Intel took so long to implement the concept for something other than graphics. It's a logical path to much better performance than scalar processors.
Vector processing and wide instructions pipes are two different things. The modern equivalent of vector processing are SIMD instructions. Having said that, some of the techniques used in the Cray-1 did allow for pipeline parallelism, but they were limited to the vector operations. In a modern CPU the wide pipelines are used for all instructions.
I understand it has 2 ports ,either one for charging and if you are using one for charging you will be left with only one ...is there an adapter you can recommend if you want more port .... thanks
So how will the App Store feature iOS / iPadOS? Or differentiate between launching an app on Rosetta or Universal? I could see when users download Photoshop in iOS to try it out and then having the standard app. Imagine using spotlight (doing command+spacebar, type "photoshop", press enter) and you might not know what version of the app will launch. I feel like it can confuse some users.
If the RAM is on the chip then won't the access times be even lesser resulting in performance boost???? Also A78 has only 4 stage pipeline? Or 4 parallel pipelines like in superscalar structure??
Yes, RAM will also help in performance boost but not much, because the speed of the RAM is more dependent on the type of the RAM and not only the difference of distances. And it's still far far a way from the CPU, So not a huge performance jump can be seen due to RAM-on-chip. It'll be a small boostof performance. It's definitely 4 parallel pipelines, because you can not run any processor with just 4 stage pipeline 🤣🤣🤣. You'll need 5th stage too to execute IO instructions. And here for a-78 he's talking about the independent 4 different instructions in parallel, so that it'll execute all 4 instructions as one instruction. And took very less time.
@@meetmarakana3275 yeah and cache miss highly unlikely if L2 cache is 8 way set associative....but may be loading time of a new application reduces as RAM is on board....
@@meetmarakana3275 I was wandering if the ram is in the same die or if its just in the same package but in separate die. and if its in the same die then are the cpu and ram sharing the same clock? because if they are then there will be a huge performance boost because all the handshaking will be gone, many clock cycles will be saved
@@AnirbanKar4294 I can not see the main RAM having the same clock. In a PC you just need to much RAM, it needs to be replaceable upgradeable and would just be to costly. And remember with big/little your CPU clock is going to keep changing. 1.x GHz on the little and 2.x or even 3.x on the Big core. Remember even the rasbury PI dose not have the Main RAM in the SOC, even at 1.5ghz it would get cooked!
Intel has been notorious in. Their use of clock speed as that has been the confusion engendered by intel when they were trying to compete with and and Chris processors, they used this to rid themselves of a second competitor and leave and as the only competition but their confusion didn’t work eventually because the and CPUs of AMD got ore work don for the clocking claimed. Now they can’t attack AMD. Because they have newer technology.
I have heard that the M1 machines in partnership with the OS will send a report via the internet to Apple whenever you start an application. Data includes the application ID, your IP address, time and location data, etc. The OS also gives Apple the ability to decide whether or not the application is allowed to run on the machine. The user will not have the ability to disable the reporting and control functions. Anyone able to confirm this?
Do you think M1 will be coming to the Ipad Pro and universal pro apps such as full photoshop in the next ipads? Ram should increase to 8gb first like on the base model M1 computers?
What’s the advantage to having the ‘auto-on, auto-wake’ feature when I lift the lid? I’d prefer to turn on my MacBook Pro myself, manually. How can I disable this ‘auto-wake’ feature?
@@GaryExplains yes. Often. I often move it here and there, then open the lid. It awakes every time I open it. I do not like this feature. It always comes on. But sometimes when I travel with it and set it up, I don’t want it on immediately. I’d prefer it open - but still remain off until I expressly turn it on.
OK, interesting, I have never seen anyone do that. After a quick Google search it seems that the old method of setting a value in the "nvram" doesn't work with the M1/M2 models (in fact it can brick your device 😬).
It would be very easy for Apple to clearly specify the base for comparison, both on battery life as well as on performance. Apple chose not to do it for some reason.
While some of its graphs were obviously unclear and intentionally unspecific. All of the comparisons to the previous generation Macs are very clear and specific.
Can you tell us what your projections are, considering this move by Apple. Where does this leave Intel/AMD and the x86 ISA. Will AMD/intel have to come out with an arm chip to compete in the mobile space. I just find this very interesting, the competition is so intense, pros and cons going each ways. Will there be definitive winners and losers. Yea thanks!
Best Battery Life (Laptopmag.com) 1. Dell Latitude 9510 (18:17) 2. Dell Latitude 9410 2-in-1 (16:54) 3. Asus ExpertBook B9450 (16:42) 4. Apple MacBook Pro, 13-inch, M1, 2020 (16:32) 5. Samsung Galaxy Book Flex 15 (15:44) 6. MacBook Air with M1 (14:41) 7. LG Gram 14 2-in-1 (2020) (14:00) 8. Dell Latitude 7400 (13:23) 9. HP Spectre x360, 13-inch, Late 2019 (13:20) 10. Dell Latitude 7400 2-in-1, (13:08) Good to see Apple finally catch up...lol
The hilarious thing to me is that Universal format for PowerPC never went away. So you can actually compile one binary for PowerPC, Intel and ARM and they’ll work on like every Mac to ever run macOS lol
That's insane
Can always count on Gary to explain these things in an easy to digest way. He's not the professor for nothing
*FRANKLINGOODWIN!*
@@1MarkKeller MARK!
@@franklingoodwin *FRANKLINGOODWIN!*
The only Mac I can afford is big Mac
I can afford a Double Big Mac. I know, I'm doing well for myself 😂
Yeah Big Mac from McDonald burger 😂😂😂
I am hungry because of you, thank you very much!!
You’ll be getting the Mac Pro then 😀👍
Nonsense. You can afford a 20 year old Mac with floppies.
Hands down the best video out there about the M1 chip! 💪🏻
Wow, thanks!
@@GaryExplains I think Apple is being disingenuous regarding the performance of there SOC, when they compare its performance to a mac laptop, it may look amazing. However when one consider the significant thermal and power handicap these computers have, and you start looking at it from the wider PC industry, its not as impressive. Earlier this year I ran Cinebench on friends brand new i9 equipped MacBook pro, it was beaten bey my I7 t490 and utterly destroyed by my Zephyrus g14. The comparison with the mac mini, is with a three year old i3 and that is a specific processor ( i3-8100B) , when you start looking at it from that point of view, and knowing that manufacturers will give the numbers (and graphs) to show their products in the best possible light. You realize that all the what apple say don't really matter until you have a device and having a real world experience.
There is a part of their presentation i think they straight up lied, Rosetta running x86 apps on ARM faster than native, nope, wont happen. Thats BS.
Still I cant wait to get one, my son loves mac OS
@@nicholaswjamrock I don't know if PC users are in some deep denial mode or what it is but these things have been tested independently several places now. Benchmarks not produced by Apple has shown Rosetta beating all the x86 Macs running x86 binaries. Not hard to see why given that M1 is that much faster than their existing offerings.
Anyway you are missing the mark if you try to compare the M1 to the the biggest most powerful computers the PC world can offer. M1 is for their low end stuff and it is already beating top end Macs. The point is that each individual core on the M1 is more powerful than anything else out there, certainly per watt. That means it is relatively easy for Apple to destroy the top of the line PCs in the future. They have a huge thermal budget they can use if they want. They can add in a ton of more cores. They have faster cores AND can add more of them.
The M1 simply shows what is possible with Apple Silicon. When the M2, M3 or whatever they will call it is realized, the top end of the PC market will be conquered. Dell, HP, Lenovo etc will have their iPhone moment.
@@erikengheim1106 I am not a PC user, I am a computer user, I have used all major architectures and operating systems released over the past 30 years, my first mac was a Wall Street. I own several macbooks and i own a working powerbook, that i bought in 2006. I am not anti- Apple, i a just a pragmatist. With the exception of my zephyus, i only buy laptops types of laptops thinkpads and macbooks (powerbooks).
1. where are the benchmarks you speak of??
2.you said Rosetta beats X86 macs, ---- Now I am in a unique situation, i have thinkpads and macbook pros with the same processor, i used these thinkpads as hackintosh and
they were faster, way faster. Now may be asking why?, let me explain, apple for the last 5 years have being optimizing their laptop based on the limits of power delivery and cooling system that can be accommodated by the physical size of the laptop. this is a fact and has been proven my reviewers and independent publication.
3. if you take a look into how ARM instruction are structured and executed, vs x86, there is no way an emulation can out perform bear metal, sorry that's just not so. when i was younger , i used to emulate old game console on my pc, i was running on hardware which on paper was up to 5 X faster than the original consoles and it was not always pretty.
4. i did not compare the M1 to the the biggest most powerful computers the PC world can offer, Apple did. when they said its faster than 95% of all windows laptops, to which they may be correct, but what i was looking at was the pro level laptop and laptops within the price range of apple offerings. when compared to those its not that impressive.
5 Now regarding the M1 Beating top end macs, "www.apple.com/mac/m1/" follow the link and read the fine print at the bottom.
6. One thing i agree with you whole heartedly , the performance per watt is going to be amazing
7. some thing you should understand about, electronics, performance and power consumption dont scale linearly, eg a cpu that consumes 5 watt at 2 gHz may just consume only 1 watt at 1 gHz, so the long battery life that was talked about will only be available under certain conditions, and this applies to all electronics.
8. I am not making speculations on future apple SOCs, i am talking about what they presented and how it was presented when compare to what it on the market right now. Its is possible two years from now apple may have the best processor on the market, I remember the days when the fastest laptop you could buy was a power book, no gimmicks no marketing BS, it was just flat out the best. I dont buy marketing, i buy performance.
you may think i am against apple, i am not i am an apple product user for over 25 years and i am proud of that, but i am will never be blinded by marketing or hype. I want this to succeed, not because of apple, but because when ever apple starts a trend the entire consumer electronic industry follows, (blue tooth, wifi, high resolution displays, long battery life, the ultra portable laptop, etc....) . Apple being successful will show other manufactures that is possible to make a compelling ARM based device, and being that are devices are usually cheaper to own and uses less power. it would allow more people to have access to modern computing.
Even though there are companies already Offering computers that runs linux and windows based on the ARM architecture they are not good value for the average user, Apple is in a position to show them how its done when its done properly.
@@nicholaswjamrock
Problem is, that it makes no sense for you to be a Mac user given your views. Why would you buy products which according to you are clearly inferior? Schizophrenia?
1. Here is an example. They run Geekbench under Rosetta 2 and still get great performance.
www.macrumors.com/2020/11/15/m1-chip-emulating-x86-benchmark/.
2. I think Apple mostly makes the right choice in focusing on battery life and portability. I honestly think Macs have had high enough performance for 90% of my needs for at least 10 years now. Yes for some very compilation heavy tasks I got a PC some years ago. But I don't do that often.
3. Rosetta 2 isn't emulation per say though. It performs a translation of x86 code upon installation. So you are running ARM code. Of course there is a performance hit of about 20-30% or so, running ARM code that was no properly optimized. Yet given that M1 is so much faster than existing Macs that still puts them ahead of the pack.
Emulation has come pretty far since your gaming days I think. Rosetta 1 worked pretty decently as far as I remember and it was not as good as Rosetta 2.
4. Do you have benchmarks showing that most pro level intel laptops beat the M1 Mac Pro performance wise?
5. I don't need to read the fine print on Apple's page, as this is what independent non-Apple benchmarks show.
7. But that is kid of the big deal about the M1. They can increase its performance a lot with much lower watt increases than what is required for intel chips. Basically they have way more potential to jack up performance. Hence it is premature to write off Apple.
8. "Best processor" by what metric? You move the goalposts around so much I am not sure what you are really talking about. You talk about Apple performance down due to thermal constraints in the laptops. Sure... but if ask how powerful is the M1, then that is a separate question from how powerful is Mac Book Air with an M1 inside and no cooling. You seem to conflate the two a lot. Apple clearly has a high performance chip, whether they are choosing to utilize the full performance of that or not is a different question.
---
Not sure what kind of Apple user you are. You seem to only care about the performance. If performance is the only thing somebody cares about, then I fail to grasp why they would be a Mac user at all. Macs are about the total experience. I use a Mac because of a combination of ease of use, elegant design, access to a rich Unix eco-system in a system that just works. I switched from Linux because I got tired of spending so much of my time configuring stuff, and lacking access to quality GUI apps.
Using a Mac has saved me time due to superior experience not due to superior peformance. I remember years back when doing home video editing in iMovie. My brother did the same on a high performance PC, but it took him ages to get stuff done. Horrible interface. Stuff didn't work. Softwae would crash. It was frustrating experience for him, as well as for most PC users at the time it seemed. I am sure PCs are better at this today. But certainly when I made the switch to Mac during the OS X switch a lot of stuff like video editing just worked a lot better on Mac. And the system was no prone to crashes like Windows PCs at the time.
---
I am glad you are positive about the future, but I don't get why the M1 doesn't impress you then. Is it not way beyond what Micorsoft has offered in the same space?
If it beats most existing intel laptop Macs, then I don't get why you don't consider this a major achievement.
I wrote a prediction 4 years ago that Apple would switch to ARM. In that prediction I took it for granted that their new ARM laptops would have lower performance than their intel laptops. My rational for ARM still being worth it, was that I assume consumers would prefer slightly less performance with longer battery time, and cheaper laptops.
Apple completely blew past my predictions. Not only do they offer these computers cheaper, than they did before and with longer battery life, but also with significantly higher performance. Something I had deemed impossible to achieve with ARM.
I just don't comprehend what kind of fantasy world you live in where you expect even more than this.
I've been waiting for your video since the announcement.
Hope you enjoyed it!
@@GaryExplains oh we enjoyed it for sure.
I wish you were my microprocessor teacher during engineering.
I love that they brought back the pc guy... truly the end of the intel era for apple
Why? Nothing is better for the development and customers then competition. I am user of Ubuntu so M1 does not much my needs. Please be more "polite" for the innovative work the Intel delivers (and others). All companies which contribute in innovation, development and improvement of human live admire our applause. Have a nice day.
Since this is only gonna benefit Apple, the comparison to Intel is gonna be pretty moot overall.
Intel isn't going anywhere.
What this shows is the fact Apple switched 3 times to many the fans keep spending money, just shows how power the x86 platform is.
Not has been able to replace it.
Sure AMD have made some strides towards faster performance per clock cycle, but it isn't significant enough for desktop oems to make a complete move.
Since Apple isn't gonna sell its chip to 3rd parties, then PC oems will have no choice but to still choose AMD or Intel.
No matter how good this M1 is, the Mac still be be more limited than a Windows PC.
Go faster to nowhere is not better at being slower going somewhere.
Like Gary said, for users of the Mac who need real power, they should wait for later versions for the iMac and Mac Pro, because the M1 isn't gonna be powerful enough handle heavy sustained tasks.
His claim about Rosetta not taking that much of a performance hit is to me not very true.
A virtualize environment or an emulator can't run exactly as nature hardware.
Maybe its OK for level apps like Office, but for PS/ FCP? No way!
@@TechieXP "FCP will be slow and Rosetta will be garbage!"
Source: "Dude trust me"
Quite different to the testimonies from people who had their hands on the DTK and said "the DTK was running x86 code faster than x86"
ARM won't replace x86, at least for right now, but it will be the de facto alternative for the majority of people who want lighter and longer-lasting laptops. Only Apple's core, though, do I see really taking on the desktop. They were not lying when they said it was the "fastest core in the world" seeing the preliminary benchmarks, it outpaces the best Zen3 core AMD has right now.
Nvidia will be the "intel" of the ARM PC land, so since you hate being locked down, I hope you don't buy one of these Nvidia ARM chips when they come out and are comparable or maybe even better to x86 chips in years time.
Something to think about: The end of the Intel era also means a "more closed, tightly-controlled" environment. Good for Apple, but the consumer will have less and forced choices.
@@markusbuchholz3518 But Intel has been overcharging and underdelivering so I'm hoping it will push them to actually make good CPUs
Thanks Gary. You consistently use enough detail to explain without getting bogged down in the weeds! Some video I would like you to cover the differences in Intel architecture to the Apple architecture, and how the M1 is conceptual similar to IBM's Cell processor and its adoption in Sony Playstation 3.
Was waiting for this. Gary does this best! Brings all the critical thinking and logical analysis that's missing from other videos.
This is what I like and admire Gary, he explained the most complicated things when it comes to computing or computer as a whole even I don't really understand what he is talking as I'm not into computer so much but I like listening to him. How I wish I'm as knowledgeable as he is. He deserve a million subscription. Thank you Gary for sharing your knowledge, I enjoyed watching your video.😂😂😂👌 😀 🇵🇭
Same clock speed. Likely 7 GPU cores because of yield. Apple’s speed comparisons are between the machine they’re replacing to the new ones.
Thanks a lot for this detailed description of the M1 chips! I don't think any other UA-camr has done a dive this deep!
Thanks Gary for educating us!
I remember few years back you mention in one video that we need more arm computers, that could be the push we need! I could see what aws/annapurna graviton could do 2 years ago in a very small module and I can't wait till we can get on an a78 architecture cpu from nvidia next year hopefully :)
best man to explain the context of these fan debates. thanks a lot!
You did it on a whole different level than the rest of UA-cam. I wonder when windows and other ARM based CPU manufacturers will catch up? Does AMD have a chance or x86 is just cannot be made so power efficient.
Apple has gone RISC before and went back to Intel from the Power PC. Apple does not give a F about speed. It's all about control. ARM is great for small devices but when you need power it doesn't have the capacity.
Ever heard of Galaxy book S? Windows on ARM is already there. The issue is third party app support for ARM, which is just not there yet. I think x86 still has a few years left before ARM takes over, despite its handicap of power consumption.
@@jordonberkove7438 Arm doesn't have the capacity? That is nonsense for two very important reasons. 1. The world's fastest supercomputer is based on the Arm instruction set. 2. There is nothing intrinsic to the Arm instruction set that makes it slower or faster than say x86. Please tell me why you think it is?
@@jordonberkove7438 The leaked benchmarks tell a different story. It seems very powerful. We'll know more very soon. Control is surely important for Apple, but just one factor.
@@nileshkumarvaishnav9550 Thanks. I'll have a look.
Anandtech has an interesting deep dive into the details of the A14 Firestorm core
I just read it. It challenged willow cove and zen3, and was an absolute bloodbath for stock ARM cpus. Disappointed they didn’t write about GPU...
@@zonyyu8412 The A14 is faster but with higher power consumption. Stock ARM cores are focused more on efficiency instead of outright performance. A performance-focused core design like the X1 should narrow the gap to Apple's Firestorm cores. That said, Apple made some intriguing design decisions that enabled the A14 to challenge Intel and AMD.
@@fallinginthed33p indeed. For them to have gotten where they are, amazing. That being said, Intel and AMD shouldn't be counted out just yet
@@fallinginthed33p the firestorm cores are indeed not the most efficient cores out there. They are less efficient than the A77 found in S865, and about the same efficiency as the A77 found in S865+. Efficiency takes a hit when driving an 8-wide behemoth of a processor to 3GHz, where wider designs tend to prefer lower clock speeds.
That said, the A14 CPU is more efficient as a whole compared to S865 CPU, and that is because the little cores are fast enough to be significantly utilized, and because those are MUCH more efficient than any of the big cores, overall efficiency is greater. The Cortex A55 are so slow that most programs simply switch to the big cores for execution, and that is why in recent years, we have seen “medium cores” to handle tasks that little cores cannot.
All in all, efficiency of the CPU complex depends on the workload - how often the large cores are used, how often the small cores are used, etc.
FYI: the little cores on A14 are as fast as the big cores in A9. Quite shocking how far CPU’s have come.
@@zonyyu8412 All this is probably why we aren't seeing 2x or 3x battery life for M1 Macbooks. The Firestorm cores are wide, deep, fast and hungry. Running them at full speed will consume x86-like levels of power but with more performance. The new ARM X1 cores are meant to be paired with A77 and A57 cores; I'm wondering if Apple will follow a similar high-medium-low speed design for later chips.
Great video and really useful to have the explanations. 👍
Glad you enjoyed it!
I did thank. I‘m curious how Apple might incorporate discrete graphics when they seemingly have made a great job of the integrated graphics. Is it just me or does it seem that the innards are becoming more modular as in, now just replacing one or two chips?
I am interrested in the technology aspect of this. I think that ARM and Risc-V is the future of computing for the general public.
You explained how they did it so clearly and simply that now i am thinking why others can't do it
Before watching it I already know this video is the best explaining the technical aspect of the M1 chip while at the same time making it digestable for most people.
To be honest Gary, I'm quite looking forward to the old Acorn community getting Risc OS running on it :-)
" you you you come , we will execute at the same time " damn that sounded like straight out of middle ages
Simultaneous execution was not really a middle ages thing. I can think of a much more current example though.
With the throughput being demanded from a portable computer only because it is portable instead of a desktop form factor, give me an inch thick, 5 lb macbook pro that doesn't scorch my lap or the power cord AND with the extra 3/4" of thickness, give me so much more performance it'll bring me tears of joy. You can make the Mac Pro as big as it was in 2013 and pack it with paralleled processors on paralleled 4 layer pcbs, memory, and cooling fans on a computer that will finish rendering all my 8k video before I have the chance to click the Render button.
Excellent video. Thank for breaking this down without the usual Mac vs. PC melodrama (cough.. Linus). Subscribed!
Oh man... I love your channel!!! I always learn something new. Please 2021 be kind and keep him safe and well! 👍👍
Amazing explanation as always, I just wonder what Qualcomm, AMD, Intel, Samsung, even ARM itself is going to do next, I just hope the integrated ram doesn't catch up or maybe it can be hybrid system because the idea of having to select the amount of ram at the moment of the purchase without having the option of upgrading it's just awful (unless they give you way more than enough on the baseline... which is not the case) which also makes you wonder, what are going to do about the Mac Pro and discrete GPU'S
The Firestorm M1 cores all run at 3.2ghz.
The only difference between the low end MacBook Air and the other M1 computers is one less graphics core (probably binning to recover some SoCs with a bad GPU core). The Air will thermal around the 9 or 10 minute mark if heavily pushed.
Good! With AMD already on the ryze, and now Apple, Intel can toss out their stockpile of overpriced processors to the bargain bin
intel is not going to throw the towel; dont forget that in 2000 apple was at the edge of bankrupt with their slow underperforming overpriced powerpc cpus (by ibm + motorola); they moved to intel (after a decade of market bashing the pentium in their bullshitty comercials) and they gained super performance and reliability, even making it possible the birth of the hackintosh; let's see if apple's ARM is on pair with the all-mighty Xeons of todays in the power desktop and server fields
@@alerey4363 apple will not and will never be competing on the server side business unless they have really good reason to do that, Apple is more towards consumer computer business and software.
@@reobating8826 www.apple.com/macos/server/ you were saying? ah, and let me remind you that the XServe was a failed line of blade servers apple sold in the 2000s
@@alerey4363 this will be true pre iPhone era, of course Apple has server side business but what I meant to say is that they never will compete seriously in that market just because they do not want to focus on it.
@@reobating8826 they can't compete with industry standard HP, Dell, IBM servers but they'd want to, given they still SELL their OS SERVER software, to be run on their *INTEL* macpros (you know the dreaded trashcan, heater, whatever u call it); so if they don't make macoserver runnable on their new ARM platform then intel cpus are still KING in that domain (and prosumer desktop too); so long at launch apple only showed glorified versions of the iphone/ipad pro processors embedded into macbook chassis and macmini case, nothing more, and oh yes..."one more thing": those ridiculous graphic charts with curves showing "performance" without even a real data compared to real world computers..well, it's apple, it always did the same, zero specs, pure adjectives (this is the best notebook apple has ever produced, fantastic speed, blazingly fast performance, super long battery life and the bla bla alike)
I was specifically waiting for your video about the M1 chip :)
Hope you enjoyed it!
Incredibly lucid explanation. Thanks! So Gary, if you could pick one of the three, which would it be?
It would be one of the laptops. And I would get the MBP if I had the cash.
the graphs apple showed was impressive i can’t wait to see the tests on youtube
not really. even an i3 can do that. or amds apus.
@@ashishpatel350 you mean an i3 would outperform an i7 10th gen?
@@ashishpatel350
Name an i3 model based on 5nm process and has a dedicated neural engine let alone process multiple streams of 4k
@@amanagarwal1939 not to mention an iGPU that goes toe to toe with a GTX 1650 or rx 570
@@amanagarwal1939 Tiger Lake.
Geekbench results show all three machines running at 3.20 GHz. They seem like they are the same chip running with different cooling solutions (and differences in GPU binning).
Kjopp
I
Thank you for the video. An off-topic question from a foreign language teacher: Do you prefer "Apple has" (which is what you say at the beginning) or "Apple have" (which you say at 1:52 and 16:21)? Is there a difference or is it just the notion of collective nouns vs plural nouns, i.e. just as "the crowd is cheering" (looking at a crowd as a whole) or looking at it as "every single member of ..." -> "the crowd are cheering"?
The correct way is "Apple has", "Apple is" etc, as Apple is a company, a legal entity. However I often think about the engineers so really I am saying "the Apple engineers have" or "the Apple engineers are" and that gets shortened to "Apple have" etc.
Nice job Gary!!
This is actually the most informative video I 've seen regarding the M1
Glad it was helpful!
Surely you given enough details about which i even did not studied in my graduate level computer organization subject
In just 20 minutes!
Hey gary could have it been possible to translate x86 instruction set to risc at hw level or is it limited to intel licencing we have wine that translate nt system calls to unixs we have Rosetta that translates x86 to arm how about using a hw level translator
Translating it at a HW level basically means you are building a x86 chips and that requires a license from Intel. Transmeta tried it and the rumor is that Stexar/NVIDIA tried it with Project Denver. According to Charlie Demerjian, the Denver CPU was originally intended to support both ARM and x86 code using code morphing technology from Transmeta, but was changed to the ARMv8-A 64-bit instruction set because Nvidia could not obtain a license to Intel's patents.
What a concise summary. Tq Prof
*GARY!!!*
*Good morning Professor!*
*Good morning fellow classmates!*
Stay Safe Out There Everyone!
MARK!!!
Mornin' bud
Morning Professor
Good evening Mark!!!
Why does it support only two thunderbolt ports? Does the M1 need to have two thunderbolt controllers to support four thunderbolt ports? Also it doesn't support eGPUs or more than one display, does this mean the M1 doesn't have enough bandwidth for those? How will the performance of the SoC be affected if the ram & gpu are separated? Which I suspect they'll do for mbp 16", imac, imac pro & mac pro
i’m trying to decide which macbook pro to get (either intel 2020 or M1 2020) and i’m stuck on which one to get. is M1 reliable at this time considering it is a brand new chip??
and i need to run adobe apps as well, will M1 support my adobe photoshop and premier pro??
It is not really brand new though. It is a continuation of the A14 chip in the iPhone. People don't ask if the next iPhone chip is reliable. I think it is more of an issue if there will be x86 software that will all run as expected. Apple was showing Adobe Photoshop getting ported. Not sure about Premier Pro.
I am personally not in a hurry. I am still using my 2013 Mac Pro model and still don't have performance issues. I suppose one could just wait a few more weeks and see what kind of experience people have with these new Macs.
The definitive technical rundown on M1 comes from, as expected, Gary the Great
Jonathan Morris and Linus Tech shots fired! Jonathan shows below a new iPhone 12 Mini exporting a 4K video in seconds while his Intel based i9 Mac is still rendering minutes later.
The iPhone is using hardware based encode while the iMac is using software based encode.
You should always use software based encode unless you’re live-streaming and your system cannot keep up.
@@soravulpis96 And why is that?
Thanks Gary, great video!
The bit that puzzles me about the M1 is the Unified Memory Architecture (UMA). These new machines support either 8 or 16 GB of memory, so if I buy a Mac mini with 16GB is the M1 I get physically different or is Apple handling a memory upgrade with some kind of clever software switch?
The M1 is physically different.
The defacto-method to improve IPC is to slap a large L2 cache in the die.Nvidia did it AMD did it and also Apple did the same.
I wonder if the desktop version of these chips will completely ditch the power efficiency cores and just use a homogeneous architecture all chips running at the same clock speed.Also something like a 16 or 32 core chip would probably be quite powerful.
I would certainly think that the M2 would only have the performance cores, 8 as a minimum.
@@GaryExplains Does it have something like SMT though?
Properly not, but we will have to wait and see.
@@GaryExplains If by any chance they do that would really make it a competitor to the high end x86 chips.Lets see what happens!
I LOL'ed @ 8:38 "You you you come here!! We'll execute you at the same time!!" 🤯🔫😱🤣
Gary,why don't you explain the Intel and AMD processors in details as you do with smart phone processors?
You are always Informative! Thank you Gary! Keep up
What Apple has done is extraordinary... No other ARM licensee, or manufacturer using chips from some other ARM chip maker, has yet matched the sophistication and pace of Apple's silicon engineering. And it hasn't been for want of trying. Samsung dumped massive investments into an effort to build its own M series chips before giving up.
Furthermore, Huawei and other Chinese makers have manufactured ARM reference designs under their own brand names, without achieving Apple's results. Nvidia desperately tried to beat Apple with its Tegra chips before throwing in the towel in smartphones. Qualcomm has been embarrassed by its fall from mobile chip leader to merely a runner up by Apple ever since it was caught flat-footed by the 64-bit A7 back in 2013.
Intel once spent billions subsidizing Android licensees to produce tablets with its mobile x86 Atom chips before giving up on mobile. And now, Apple is taking its mobile expertise developed to power iOS devices and using it to drive its notebook and desktop Macs, starting with the MacBook Air, 13 inch MacBook Pro, and Mac mini. The company expects to complete a transition over the next two years. That would be hard to believe from anyone else, certainly after Microsoft's eight-year struggle to make little progress with Windows for ARM.
Apple's move to its own SoCs isn't just a chip migration. It's a radical rethinking of desktop PCs that leverages massive amounts of engineering work already completed to deliver iPhones and the light and thin iPad platform, to make them ultra-responsive, radically mobile, and blur the line between hardware and software.
I'm still amazed by the speed of their development
brings back memories of the transition from PowerPC to Intel.
Seems there approach to marketing and comparison is going back to the way it was during the PowerPC days too.
Or M68K to PowerPC, or when an entire industry moved away from 6502.
@@brostenen I don't remember if Apple provided their "fat" binaries during those transitions.
@@strobelightaudio Apple had an M68K emulator, build into PPC version of MacOS.
I’m excited for the Mac Mini! I keep breaking iPads by running games and keeping them in a kiddy case with handle so they only last me 2 years 😕😩🌱 they thoughtfully put a fan in the Mac Mini 😀
The iPad shouldn't break because of that
Great giving pictures of how the processor operations.
Gary can you say how it wil benifit for developer with different programming language
Is this wide execution architecture only possible because it's a RISC based processor? Or would it be feasible to create something similar in a CISC based processor?
Because as far as I understand, a CISC processor splits its instructions into smaller ones, so maybe that makes it harder to do this kind of parallelization...?
I have a question about RAM and the M1 processors. I have seen several videos wherein iPads Pro have no trouble editing and rendering 4k and even 6k video often faster than powerful desktop computers. I have also seen videos of people playing demanding video games (IIRC, Fortnite before the recent kerfuffle) with no trouble. The iPads Pro featured less powerful processors operating in a more thermally constrained physical environment than a laptop or, especially, a Mac Mini. They also had less RAM, 6gb vs 8 or 16gb.
Is it possible that the transition to Apple Silicon might require us to rethink the issue of how much RAM we need, especially with apps that are optimized for these processors and the operating systems?
Thank you for your wisdom.
Video playback and rendering can be accelerated through specific hardware blocks so you don't need to use the CPU. It's a lot more efficient and faster that way. I don't know how this would affect RAM usage but Apple are leaders in image and video processing in hardware.
This is more about OS X and tight integration with their arm soc! It’s not just a cpu! 🤯🤯🤯
@@hanniffydinn6019 Thank you. This is what I was wondering about
@@fallinginthed33p However, even the casual video editor should avoid using hardware accelerated video encoding. Software encode is superior.
I hope apps updated to be native to the M-series SOCs allows the user to not use the specialized hardware and solely use the CPUs to encode video.
Where can i find specific technical details about m1. Gary hits my mind
I think it will be very interesting to see what they come up with for M1X and M1Z...
Anyway, I believe that they also have different TDPs on MacBook Air and MBP13... maybe 12W in MBA and 25W in MBP? Or maybe 9W in MBA and 15W in MBP...
Anyway, I really dig these ARM chips, and I hope that Qualcomm/Samsung/Nvidia are able to put out a true competitor to this M1, although IMHO it'll be harder to take advantage of the power optimizations on Windows machines, even if they use only ARM-compiled apps...
That said, I don't quite like what degree of control has now Apple on his entire hardware...
Look, the not-upgradable RAM on MacBook is not surprising to me: I don't like it, and I personally buy only user-uogradable devices, but it's nothing new, and there are plenty of Windows devices on which the RAM is soldered on the motherboard itself... if I'm gonna have soldered-on RAM, I'd better have it directly linked or soldered on the CPU, as it guaranteed lower access latency to it (which is extremely useful, especially for on-board graphics).
What concerns me is the fact that RAM is now not-user-upgradable even on the Mac Mini... I'm worried that they will make RAM soldered out of the factory even on iMacs!
The 2 things that make me uncomfortable the most, though, are the lack of support to eGPU, the lack of support to Windows and the possible lack of support of any other operative system that's not macOS Big Sur or newer: we are used to iPads being tablets, and therefore being stuck on their OS, but MacBooks and Mac Minis are still PC nonetheless, and I expect the user to have control on what runs on its machine. But Apple doesn't want to give users this freedom, because every single Apple device is, to them, just a way to get users in their software ecosystem, and get money just by gathering data, selling apps on their store and sell people services on subscription (SSaS)... THIS I don't like!
@@tri-a7655 I partially disagree...
1) I'm aware of other people working on ARM transition, but let's be honest. Firstly, if we want to take ChromeOS into consideration, we'd rather also consider other Linux-based systems, which have been running on ARM for years already. That said, ChromeOS is more similar to iPad OS than it is to a true computer operative system, like Windows and macOS: it doesn't have the flexibility and the power of Windows, macOS, Debian, Ubuntu, Arch Linux and other stuff like that...
And I'm aware that there are already devices with specific Qualcomm and MediaTek ARM processors around (Surface Pro X, some Chromebooks...), but to me the Qualcomm and MediaTek options seem a lot weaker than Apple M1 in terms of power and efficiency. Moreover, the Windows, ChromeOS and Linux ecosystems don't have the same degree of vertical integration.
This means that Apple is able to better optimize the experience, from hardware down to software, while it's harder to have a truly optimized Windows machine. This means that you have to account for losses in efficiency from the lack of optimization. So if Qualcomm and MediaTek truly want to compete in this space, they need an even stronger offering than Apple, not a weaker one. For now, they are a step behind.
2) Please, don't say soldered RAM is OK... lack of user upgradeability, serviceability and customization is ALWAYS bad. It's a compromise we are now accustomed to because it gives us thin and light designs like the XPS 13 or the MacBook Pro 13, but it's not idea. To me, it's unacceptable on bigger, more powerful and pricier machines like MacBook Pro 16, iMacs and other non-Apple computers...
You can't say that it's the same as an high end phone... firstly, in phones there's not the actual space to put user-upgradable RAM, so the compromise HAS to be done.
Moreover, phones DO NOT behave like computers, even if they'll happen to have the same CPU designs... on high end phones, the need for RAM usually comes from the habitude of keeping many many apps in the background, many of which need to constantly update informations for notifications. But the single apps themselves do not use all that RAM, and there are very very few apps (mostly games and foto/video editin apps) that actually need much RAM at all. On PCs on the other hand, you still have the multitasking factor, maybe even stronger. But on PCs, if you multitask, you probably are more active on those apps (unless they are messaging apps like Whatsapp Desktop, Telegram, Discord...). And, most importantly, there are many more apps and use cases that actually NEED as much ram as possible (batch editing in Lightroom, Photoshop editing, video editing, music production, simulation, calculus...). You can optimize how the system manages apps and programs in the background, but you can't optimize all that much all those program that need juice... and before saying that those chips are not supposed to be used for that, remember that Apple is transitioning its entire lineup to ARM and that they've already presented the MacBook Pro 13, which is aimed at professionals...
Moreover, in the lifetime of a phone, the kind of usage from the user doesn't change all that much, while the needs of an individual could dratically evolve in time. Think about a student who buys a laptop for school (and therefore needs 8GB RAM at most on macOS). This student may get into music production, photography or stuff like that down the line. When this appens, the processor of the laptop may still be capable of doing that, but the RAM may now hold him back. So the student may upgrade to a new laptop, even if his previous one would have still done OK if he could have upgraded the RAM... I know because I've been there: I went in 2/3 years from truly needing 4GBs on my laptop to now have 32GBs!
"if best performing iPhones were sufficient with just 6GB of RAMs.
Then this 16GB max should translate to much more capable than that."
8GB at the moment is the bare minumum for a decent Windows experience, and 16GB is what most users should have (unless all they do is basically only writing documents and reading stuff on the internet) if they use heavier programs. 32GB is actually a decent RAM amount for serious professionals, while truly demanding ones actually use 64GBs or more.
I'm not saying that 64GBs are needed NOW on MacBook Pro, but I can see a professional buying a MacBook Pro 13 with 16GB of RAM encountering some limitations 2/3 years down the line. If all you need is more RAM, you should be guaranteed the possibility to upgrade it and not throw away your perfectly functioning laptop...
And let's not forget that Apple is picturing themselves are ambientalists (using recycled alluminum, removing accessories to have a smaller iPhone box that is more efficient to transport...), while doing NOTHING for right to repair and right to upgrade. They are basically doing NOTHING to limitate e-waste...
@@tri-a7655 What I'm saying is that macOS is already more optimized than Windows (although I also use Linux and I find it to be blazing fast). Maybe macOS can be improved, but there's not much you can do to make heavy tasks use less memory. You can't do magic. Many apps could be absolutely improved, starting from the entire Adobe software suite (DaVinci Resolve runs better than Premiere Pro, for example...), but if you have 60 audio tracks on a project, there's not much you can do
Waiting for you to make this soc run it with speed test G, hopefully.
Great video.. please Do AMD CPU naming scheme..
Well explained 💯❤
Glad it was helpful!
Gary, given the size of the MacBook air cpu's, could there be an M1 in non computer devices? (cars, homes, tablets, devices, etc)
The geekbench of the AIR and the Pro shows the M1 is running at the same exact boost frequency in both cases, 3.2ghz. I take it the better cooling in the Pro just lets the CPU sustains the boost longer.
Thanks Gary for this excellent explanation of M1 chip.
Excellent video 👍
Great video, very helpful
WOW , so informative, nice job.
iv been waiting for this vid the last 2 days!!! thank u Gary
Hope you enjoyed it!
Performance is really good if this cpu is single threaded only, I’m guessing double threaded feature would be for their iMac lineup
Sir please explain in details
How apple CPU and GPU better than other or different from other?
In fully detail but in understanding language for class 11th student.
😫🙏🙏💓Pleased!!!
I have several videos about that. This one is several years old but worth watching IMHO: ua-cam.com/video/TiVF_ZzzLM0/v-deo.html
@@GaryExplains thanks sir
You are only response on my comments from almost 20 youtubers
Really thankful
@@GaryExplains another question
Can we make killer cpu or gpu
By combining ARM and apple designes of cpu and gpu??
@@GaryExplains and make video on can under display speakers use on mobile
Like in sony tv??
Since Apple doesn't sell or license its designs to anyone else, there is no way to combine Apple's tech with anything. Not going to happen.
Now what is wanting is a full performance review on the new apple products that use this new cpu without that nobody can make a full decision about the potential hardware
Is the speed of all L1, L2, L3 cache the same, or is it just the name for the closest increasing numerically with distance?
Not an expert by any means but from what I’ve read and saw, the speed of cache generally gets slower, so L1 is fastest (I think operating at the same or maybe half the clock cycles of the processor), then L2 might operate at half that again and then L3 maybe half again. As they get slower they also generally get larger though as they’re cheaper to manufacture (being that they’re not as high performance memory).
The new trend now is to then also make your previously very slow mass storage to be a kind of itself a cache (see AMD RDNA 2 graphics/PS5/Xbox Series X),
leveraging the now very fast SSD speeds, there’s not so much of a penalty by paging memory to your mass storage. Of course this also usually has the drawback that your mass storage isn’t as mass as what it could be, so really SSDs are adding a memory/storage category rather than entirely replacing one.
I think in the future they'll put the battery in the thicker end of the Air, and the board into the thinner end.
Do they use Spin Locks to arbitrate cpu/gpu simultaneous access of memory?
Greetings from near Albuquerque, New Mexico, USA. Great explanation of the Apple M1 processor. A lot of interesting info in a short time. Your description of the "Ultra-wide execution architecture" sounds exactly like vector processing, which Cray and others (most notably the Cell processor, a collaborative result between IBM, Sony, and Toshiba) were doing decades ago. I wonder why Apple or Intel took so long to implement the concept for something other than graphics. It's a logical path to much better performance than scalar processors.
Vector processing and wide instructions pipes are two different things. The modern equivalent of vector processing are SIMD instructions. Having said that, some of the techniques used in the Cray-1 did allow for pipeline parallelism, but they were limited to the vector operations. In a modern CPU the wide pipelines are used for all instructions.
8:39
"that that that one, come one the same time..
we gonna execute you you, you now, COME HERE !!!!"
What do you think, A chip with 4X Cortex X1 cores and 4X Cortex A55 at 5nm can give a similar Performance and Battery life.
Just realised, as long as ARM is shiting that A55 core there is no way that can happen.
Excellent explanation! I’m learned! Great vid! 👍
Glad it was helpful!
Great video, thanks
I understand it has 2 ports ,either one for charging and if you are using one for charging you will be left with only one ...is there an adapter you can recommend if you want more port .... thanks
So how will the App Store feature iOS / iPadOS? Or differentiate between launching an app on Rosetta or Universal? I could see when users download Photoshop in iOS to try it out and then having the standard app. Imagine using spotlight (doing command+spacebar, type "photoshop", press enter) and you might not know what version of the app will launch. I feel like it can confuse some users.
What is not widely known is that the ssd is soldered on board making it VERY expensive to replace...
Just like your smartphone, tablet, and the iPad. 🤷♂️
@@GaryExplains Yes, but a PC/Linux user will replace it for $50.... while we Mac users will have to pay $700+ to Apple for the same replacement...
@@geoman1420 True, and if that isn't acceptable then don't buy a Mac.
i was waiting for this video
If the RAM is on the chip then won't the access times be even lesser resulting in performance boost????
Also A78 has only 4 stage pipeline? Or 4 parallel pipelines like in superscalar structure??
Yes, RAM will also help in performance boost but not much, because the speed of the RAM is more dependent on the type of the RAM and not only the difference of distances. And it's still far far a way from the CPU, So not a huge performance jump can be seen due to RAM-on-chip. It'll be a small boostof performance.
It's definitely 4 parallel pipelines, because you can not run any processor with just 4 stage pipeline 🤣🤣🤣. You'll need 5th stage too to execute IO instructions. And here for a-78 he's talking about the independent 4 different instructions in parallel, so that it'll execute all 4 instructions as one instruction. And took very less time.
@@meetmarakana3275 yes, even the ARM 6 from 1991/2 had a 5 stage pipeline.
@@meetmarakana3275 yeah and cache miss highly unlikely if L2 cache is 8 way set associative....but may be loading time of a new application reduces as RAM is on board....
@@meetmarakana3275 I was wandering if the ram is in the same die or if its just in the same package but in separate die. and if its in the same die then are the cpu and ram sharing the same clock? because if they are then there will be a huge performance boost because all the handshaking will be gone, many clock cycles will be saved
@@AnirbanKar4294 I can not see the main RAM having the same clock. In a PC you just need to much RAM, it needs to be replaceable upgradeable and would just be to costly. And remember with big/little your CPU clock is going to keep changing. 1.x GHz on the little and 2.x or even 3.x on the Big core. Remember even the rasbury PI dose not have the Main RAM in the SOC, even at 1.5ghz it would get cooked!
Great video, but what about the bottom right hand corner? Is that some kind of NPU?
This is the one we needed to explain apples marketing terms and see how good it really is.
valid points. liked and subscribed.
Intel has been notorious in. Their use of clock speed as that has been the confusion engendered by intel when they were trying to compete with and and Chris processors, they used this to rid themselves of a second competitor and leave and as the only competition but their confusion didn’t work eventually because the and CPUs of AMD got ore work don for the clocking claimed. Now they can’t attack AMD. Because they have newer technology.
I have heard that the M1 machines in partnership with the OS will send a report via the internet to Apple whenever you start an application. Data includes the application ID, your IP address, time and location data, etc. The OS also gives Apple the ability to decide whether or not the application is allowed to run on the machine. The user will not have the ability to disable the reporting and control functions. Anyone able to confirm this?
RISC is a clear win vs CISC
Do you think M1 will be coming to the Ipad Pro and universal pro apps such as full photoshop in the next ipads? Ram should increase to 8gb first like on the base model M1 computers?
Can't wait to see Speedtest G benchmarks for these Macs
What’s the advantage to having the ‘auto-on, auto-wake’ feature when I lift the lid? I’d prefer to turn on my MacBook Pro myself, manually. How can I disable this ‘auto-wake’ feature?
Do you often lift the lid and then don't turn it on?
@@GaryExplains yes. Often. I often move it here and there, then open the lid. It awakes every time I open it. I do not like this feature. It always comes on. But sometimes when I travel with it and set it up, I don’t want it on immediately. I’d prefer it open - but still remain off until I expressly turn it on.
OK, interesting, I have never seen anyone do that. After a quick Google search it seems that the old method of setting a value in the "nvram" doesn't work with the M1/M2 models (in fact it can brick your device 😬).
Wow. Then I’ll leave well enough alone! Thx for the tips, Sir. And for all the good on your channel!
So... M1 is kind of Itanium #2? Itanium VLIW == M1 Ultra wide execution or not?
Is these have thunderbolt 3 or just thunderbolt.? What is the defference between these 2?
The buzzing in the audio at the end had me freaking out lol!
Buzzing?
It would be very easy for Apple to clearly specify the base for comparison, both on battery life as well as on performance. Apple chose not to do it for some reason.
While some of its graphs were obviously unclear and intentionally unspecific. All of the comparisons to the previous generation Macs are very clear and specific.
@@GaryExplains Looks like I need to take another look. Thanks, Gary!
For full stack development (web, backend, app) which mb will you recommend?
Can you tell us what your projections are, considering this move by Apple. Where does this leave Intel/AMD and the x86 ISA.
Will AMD/intel have to come out with an arm chip to compete in the mobile space. I just find this very interesting, the competition is so intense, pros and cons going each ways. Will there be definitive winners and losers. Yea thanks!
I’ve actually always thought of you as the John Hodgman of tech.
Best Battery Life (Laptopmag.com)
1. Dell Latitude 9510 (18:17)
2. Dell Latitude 9410 2-in-1 (16:54)
3. Asus ExpertBook B9450 (16:42)
4. Apple MacBook Pro, 13-inch, M1, 2020 (16:32)
5. Samsung Galaxy Book Flex 15 (15:44)
6. MacBook Air with M1 (14:41)
7. LG Gram 14 2-in-1 (2020) (14:00)
8. Dell Latitude 7400 (13:23)
9. HP Spectre x360, 13-inch, Late 2019 (13:20)
10. Dell Latitude 7400 2-in-1, (13:08)
Good to see Apple finally catch up...lol