Це відео не доступне.
Перепрошуємо.

How Useful Are E-Cores Actually?

Поділитися
Вставка
  • Опубліковано 14 сер 2024
  • In this video I check out to see just how good (or bad) efficiency cores really are.
    ----------------------------
    Follow Me:
    / techevolvedyt
    Buy PCs From Me:
    tevtechstore.com/
    ----------------------------
    Music from freetousemusic...
    Aurora by Luke Bergs
    • ▶️ Cool Beat No Copyri...
    ----------------------------
    Chapters:
    0:00 Intro
    0:22 Intel Processor
    0:48 CPU Benchmarks
    1:51 Temperature Test
    2:37 Gaming Benchmarks
    3:29 Gaming W/ Background Tasks
    3:49 Gaming & Rendering
    4:36 In Conclusion...
    5:14 Outro

КОМЕНТАРІ • 221

  • @tech_evolved
    @tech_evolved  Рік тому +30

    At 0:57 I messed up the single core performance of the P-core/control test. That score was with the P-core clocked at 3.5 GHz. When its clocked at 5.3Ghz its at/above 2000 points. Apologies for the confusion!

    • @Viewer19
      @Viewer19 Рік тому

      You wrre close to proper testing. but a R7 1700X Zen ypu are kidding. The Thread director nin the 12th gen did not multitask well according to Intel as tasks that required mukti threads were often running on the E cores when muktitasking. The 13tyh gen CPUs 13600K abd above got extra E cores to mitigate the deficiencies in TD along with more powerful P cores. One Multi task test does not tell the whole story. Most do not OC and background tasks do factor in and even more when multiyasking. The 1700 is a dead platform. AND AM4 CPUs are priced to sell and easy to sell if movibg to AM5 offsetting the upgrade cost.

    • @tech_evolved
      @tech_evolved  Рік тому +3

      Relax… I was simply stating that it scored the same as a 1700x. Don’t need to get upset

    • @od1sseas663
      @od1sseas663 Рік тому +5

      @@Viewer19 I can't understand what are you even trying to say

    • @kojirofilmz
      @kojirofilmz Рік тому

      Did you buy the CPU just for testing? I'm jealous :(
      Here I am, can only get the i3 12100f out of my money because I'm so poor..

    • @tech_evolved
      @tech_evolved  Рік тому

      @@kojirofilmz I got the 13600k for my main rig. Was an upgrade for me!

  • @JeffWaynee
    @JeffWaynee Рік тому +93

    It's well documented that e-cores are very useful for creative workflows. So if someone will only really be gaming then they can do without them.

    • @Kage0No0Tenshi
      @Kage0No0Tenshi Рік тому +9

      As a gamer I protest, I will not pay full price for a i7 and turn 8 E cores off and only thing I need is 8 P cores.

    • @lalishansh
      @lalishansh Рік тому

      @@Kage0No0Tenshi Xeons are are coming for you

    • @lycanthoss
      @lycanthoss Рік тому +9

      @Kazuto Kirigaya as a gamer you don't need more than 8 P cores, heck even 6 cores is enough. So any other cores can just be E cores, since having more P cores does not benefit you, but having more E cores will benefit people who actually use the CPU fully.

    • @Kage0No0Tenshi
      @Kage0No0Tenshi Рік тому

      @@lycanthoss have 6 cores for my r5-5600x is not egnoth for me, I would be happy to get 12 P cores and disable ht to run 12c/12t xD I am hitting 80% CPU usage in mw2 with my CPU/r-5600x

    • @lycanthoss
      @lycanthoss Рік тому +11

      @Kazuto Kirigaya more cores is not gonna make the game run faster. The reason higher end CPUs perform better is because they clock higher and have bigger caches, not because they have more cores. CPU usage does not dictate how well a CPU is performing in a game.

  • @JohnDoe-ls5xp
    @JohnDoe-ls5xp 8 місяців тому +8

    I been on a 7700k and it preforms well despite its age and being 4 cores so E-Cores out preforming a i7-7700k is crazy. I was led to believe the E-Cores were similar to a celeron or weaker.

  • @dantevortex
    @dantevortex Рік тому +31

    Of all the aspects you studied you forgot to look at the power draw. Which is, the whole use case of E-cores VS P-cores.
    It's not a scam, E-cores are less powerful, but they deliver more power per Watt at the cost of peak power.
    Their purpose is to keep the system alive with light processing tasks, while the P cores are idle. This makes your PC draw a lot less power, while essentially doing the same light application.
    Now, just do the same thing again, all three combinations, and just watch a 1080p UA-cam video, and see how much power they individually draw out of the wall.
    That's going to give some fun numbers.
    Additionally these differences in Cores is also useful in mobile application because battery life is important.
    Most laptops already did this trick for decades using discrete GPU's and the integrated iGPU together and switch off the power hungry dedicated chip when not needed.
    This new E-core vs P-core setup is just the same thing, but extended to the CPU architecture.

    • @gruiadevil
      @gruiadevil Рік тому +2

      Desktop users don't care about 5 less Watts being drawn from the wall.
      The overall wattage is horrendous on Intel. The i9 is over 200 W.
      The difference between an i7 and a Ryzen 7 is that in light workloads, the i7 draws 20 W less power. But in higher workloads it draws 100W more.
      So ... by my calculations, it really doesn't make up for it. It's just a marketing gimmick because they're fanbase were jealous looking at AMD since 2017 giving their buyers more cores.
      So Intel gave it's fanbase more cores. And more. And even more.
      The only place where big.little counts is small form factor, laptop, mobile.
      That's where you need all the possible battery lifespan you can get.
      The i5 is dead. The i5 was always a middle ground between i3, which was a budget gaming CPU but that can deliver and the i7 which was a more creator oriented CPU or a CPU that can deliver peak performance, for games who need CPU power.
      The current i5's are not i5's anymore, they're i7's. Their price is of an i7. Their core/thread count is the one of an i7. Their performance is of an i7. Just in name they are called i5.
      Don't worry. AMD is not perfect either. They dropped the budget gaming CPU. Their only offering is a meager Ryzen 5 5500, which brings DDR4 and PCIe 3.0. And it's a super-late newcomer. Considering the 12100F and 13100F were launched long before him.
      So both companies, are making you buy stuff you don't need.
      AMD makes you buy 5600X for PCIe 4.0 and Intel makes you buy efficiency cores which you will never use, if you're not a creator.
      Also the i5 gen 12 costs 2xi3 price, but gives around 50% more performance. Roughly.

    • @svgPhoenix
      @svgPhoenix Рік тому

      @Logojan Gruia I bought a non-x ryzen 5 7600 because it sips power and comes with a cooler. Spent the money that would've gone to a CPU cooler on a better GPU and didn't have to increase my PSU wattage

    • @gruiadevil
      @gruiadevil Рік тому +3

      ​@@svgPhoenix Correct. But you took into account total wattage consumption. Which is 65 versus 105 Watt? Right?
      We were talking about efficiency cores, which do the following thing:
      Even if a CPU total wattage consumption or peak consumption is 210 W; in lighter workloads, they use 5-10 W only.
      Now.
      When you build a system, and you account towards a PSU or GPU, and trying to figure out your total system power consumption; Having light workloads pull 5-10W for the CPU doesn't matter. You need to account for the total 210 W, like you did.
      What I'm saying, is that if you a buy a 10 or 20 Core CPU, you plan on using those cores, doing heavy workloads. That CPU won't spend most of it's time doing light workloads. So having Efficiency Cores in them really makes no difference in a desktop scenario.
      The whole thing is a way to save up money. Those cores didn't cut the bar to be hyper-threaded, so they're single-threaded, to still get some use out of them so they don't bin the wafer. It's the only way they can keep cost down enough to compete with AMD on prices. Otherwise, they would have to put higher prices on their products and nobody would buy them if a 200$ Intel CPU would be outperformed by a 200$ AMD CPU. Right now, they're on par.
      I'm not rooting for AMD. Because if AMD goes on top, just like Intel was for a long time; we're not going to see any major performance and price improvements just like Intel's 2nd to 10th Generation.
      I'm rooting for competition.

    • @tmsphere
      @tmsphere 3 місяці тому

      Who tf buys an i9?? It's niche tier, 13600k is cool as ice and large headroom for oc. Also better and still cheaper than 7600x

  • @edge21str
    @edge21str Рік тому +14

    This e-core performance makes me think they should bring back atom CPUs with low single digit TDPs. Think how long a laptop with that config would last on battery.

    • @tech_evolved
      @tech_evolved  Рік тому +3

      you aren't wrong. E-cores are super power efficient and aren't even that bad if they were a low-end budget cpu. intel needs to make this happen!

    • @minedustry
      @minedustry Рік тому +2

      Look up the N100 cpu

    • @edge21str
      @edge21str Рік тому

      @@minedustry Cool. Gonna look forward to what devices this thing will find itself in.

    • @satsumagt5284
      @satsumagt5284 Рік тому

      Only if they aren’t a turd like the N455 I owned was

  • @MRi-iq5wf
    @MRi-iq5wf 4 місяці тому +1

    I didn't know my new 13th gen i& 1355U is not a deca core. It's basically a gimped quad core lol

  • @jierenzheng7670
    @jierenzheng7670 Рік тому +6

    Seeing the i3-N305 e-core only CPU, would actually probably be enough for most business computers/laptops. From gaming benchmarks, the e-cores actually look not bad too. Now if only Lenovo offer them in business notebooks as the lowest base configuration, would be amazing for both battery life and work.
    I am waiting for i3-N305 benchmarks to see how does it fair against my Thinkpad X280 i7-8650U, the Iris Xe 32EU GPU will probably be double the performance because of both soldered LPDDR5 and architectural improvements, multicore wise, 8 e-cores (8C/8T) is better than 4C/8T in multicore, looking at almost 40-50% improvement.

  • @shanent5793
    @shanent5793 Рік тому +6

    E-cores get their efficiency by increasing latency, which is fine for batch processing but not so good for real-time tasks. It's a matter of leaving out all the stuff that's been developed over the years to increase single core throughput, things like branch prediction, register renaming, SMT, etc.

    • @svgPhoenix
      @svgPhoenix Рік тому +1

      Source?

    • @kojirofilmz
      @kojirofilmz Рік тому +2

      dafuq you talking avout, whres your test?

    • @mikem9536
      @mikem9536 6 місяців тому

      There's even an argument that leaning heavily on e-cores can use more power because they take longer to complete the same task.

    • @davidthacher1397
      @davidthacher1397 Місяць тому

      Dropping L3, would lead to a lot of problems. We are heading to CISC vs Cache. The core counts are basically pointless. They will bottleneck on the memory controller. Some people would have you drive the memory controller so hard you can pipeline stalls. The problem is this drives down the cache size which eventually drives up power consumption relative to actual computation. The cores have to stay engaged during stalls too. We saw massive improvements in 2-4 cores, but I am not sure real world gains are there outside of specific tasks.

  • @vladislavkaras491
    @vladislavkaras491 10 місяців тому +3

    Interesting results!
    Would be interesting those E vs P cores in the laptops as well, just to see how much battery we can spare!
    Thanks for the video!

  • @Z020852
    @Z020852 Рік тому +4

    I'm still at why the hell didn't they just make it possible for Windows to not just have a scheduler that knows the difference between P-Cores and E-Cores, but that on laptops, manually selecing "Eco" Mode (including while plugged in) or unplugging (or the power going out) should disable most of the P-Cores so you essentially run Win11 on one P-Core while all other apps run on the E-cores, making for zero idle power use and waste heat on the other P-Cores. Basically, Eco Mode or an even more extreme low power draw mode should turn your laptop into a less frustrating to use, oversized Netbook. Hell some gaming laptops might even be able to run fine barely spinning the fans if it's just running four E-Cores given shared heatpipes and four heatsinks.
    On desktops...I honestly can't see what the E-Cores are for other than CPU-heavy taks. But really I thought 12th gen mobile would mean "snappier octacore netbook with bigger battery and screen that lasts all-day on low power use." Sadly, for someone who's more likely to need PDF reader, broswer, Word, and Soc Sci Stat hashing apps, or just Word and Excel plus Spotify, AMD is still gonna give me more time away from a wall outlet.

    • @davidthacher1397
      @davidthacher1397 Місяць тому

      I think Intel goofed and made Microsoft play with it. Windows likely will do nothing more than suggest to the processor that I want this core in this power state. Outside of that Windows may scale the number of online processor dynamically against power profile. The scheduler will not really change its operating policy. How the cores fit into the profiles, that's probably anyone's guess.
      Intel implemented SMT(L2)/SMP(L1) with E cores and SMP(L2)/SMT(L3) with P cores. Windows only really supports SMP. So once that L2 domain starts to struggle the E cores will collapse. Windows currently does level of effort SMP. It keeps the logical processors move as best as it can. This is throughput in nature and lowering the power profile may actually increase latency performance. I do not expect them to give up on this unless Intel torches the cache design. What I do not know is how Windows moves threads.
      I would like to see a SMT cluster contract come out. This would allow these E cores to move to the P cores. At this point these are just hyperthreads. I would not remove the hyperthreading from the P core. Each cluster could have 4 to 8 threads on it. This is a lot per L2 domain. I would also implement multiple pipeline depths inside the P core, while keeping the E pipelines the same. The programmer can choose what do. The programmer gets a cluster for a time slice and this includes all the E cores. So no longer would you schedule the E cores in the OS. The OS can still dial them back with the power manager. However this creates headaches as programmers adopt this solution for more performance. Intel has to choose between throughput and latency. Whatever they decide Microsoft will have to live with. Currently the end user does this with the power profile. (Which is a good thing in my opinion.) Overall Microsoft is not likely to do SMT without being forced.
      There is a solution to the latency. Intel can provide a means of ensuring the priority of the cluster contract in L3/L4 domain. This would make life a little harder for the user. They would need to increase and decrease priority of some applications. Technically Windows can do this without Intel in low power configurations. Overall the 8P+16E is only useful in custom software deployments.
      I am really curious what they plan to do with rentable units. That sounds like a front nightmare. I am hoping they are not shooting programmers in the foot with it.

  • @obetemojkardi6473
    @obetemojkardi6473 Рік тому +2

    Dude try clock for clock please. The e cores running at just 4.1ghz while p cores at 5.3, that's almost a 30% uplift in frequency, and cinebench r23 single score shows that there's not even a 30% gap in performance, despite the e cores being so much more smaller and in turn much more power efficient?
    Try with the 13600k e cores vs p cores with hyperthreading disabled and clock for clock vs max oc if possible
    According to my guess, e cores on 13th gen are like 8/9th gen "cores" and at way lower power consumption
    Also, if I could ask, what's the maximum frequency you could get on the e cores with the p cores running at stock settings, or maybe even some underclocking?

    • @tech_evolved
      @tech_evolved  Рік тому

      I messed up the single core performance for the P-cores, sorry if that caused confusion. They are closer to 2000 not 1300. The 1300 score was when I hade the P-core clocked at 3.5Ghz for testing the E-core only scenarios.
      The max I could get on the E-cores "stable" was 4.1Ghz. I was able to get through 1 cinebench run at 4.5Ghz but the system crashed while running longer than that. This trend continued all the way down to 4.2Ghz for some reason even while having all but 1 P-core disabled.
      But yes the E-cores are similar to 8th gen speeds. Definitely worth investing in IMO

  • @manh2704
    @manh2704 Рік тому +1

    DannyzReviews has a video showing how e-cores actually have a decent performance increase when it comes to 1% lows. I wasn’t expecting it to have such an effect in some games but it did. Meanwhile a couple games benchmarks didn’t even differ.

  • @JO_TECH_YTB
    @JO_TECH_YTB 13 годин тому

    what about the power consomption?

  • @ProVishGaming
    @ProVishGaming Рік тому +3

    I think your single core score is wrong for the p cores in cinebench. Shouldnt it be around 1900?

  • @gscurd75
    @gscurd75 Рік тому +2

    With the specials Microcenter has right now, it makes little sense for gamers who live by those stores to buy intel. They throw in RAM for free and building a system with a 7700x ends up being about $50 less than getting a 13600k which is slightly less powerful for games.

    • @tech_evolved
      @tech_evolved  Рік тому

      I wish I lived near a microcenter. Sounds like a good deal

    • @webx135
      @webx135 10 місяців тому +1

      I live near a MicroCenter, and they had a local deal for a 13700k, Z790 motherboard, and 64GB of DDR5-6000 for a little over $500. So that worked out pretty nicely.

  • @kojirofilmz
    @kojirofilmz Рік тому +5

    Wow this is really a great test results.. sadly I can only buy an i3100F with no ecores because of my tight budget, T_T
    but still it gives me a better option of price per/performance choice once again. thank you

    • @dd22koopaesverde58
      @dd22koopaesverde58 Рік тому +2

      The i5 12400f is very good hav3 6p cores is cheaper than the i5 12400f whit only 5-10% worse than the 13400f

    • @kojirofilmz
      @kojirofilmz Рік тому +2

      @@dd22koopaesverde58 I mean the i3 12100F is only I can get, pairing it with used gtx1650, but still I'm more than happy with the performance.. i bought it at around 103 usd..

    • @kojirofilmz
      @kojirofilmz Рік тому

      @@dd22koopaesverde58 also with a cheaper h610m, 2x8gb ddr4 3200..

    • @kojirofilmz
      @kojirofilmz Рік тому

      @Ayush Kumar Yes its 12100F..and Im really satisfied ^_^

    • @MLWJ1993
      @MLWJ1993 Рік тому +1

      That CPU is still powered by some very impressive P-cores. They even made quad-core CPU's a relevant purchase option for those on stricter budgets. AMD meanwhile hasn't got all that much to offer in that segment as far as I'm aware.

  • @yosuaalbert34
    @yosuaalbert34 11 місяців тому +1

    Are you running Windows 10? because i heard that Windows 11 are optimized for the Intel Thread Director, so can you re test?
    Maybe results may difference

  • @webx135
    @webx135 10 місяців тому +1

    I think the big thing with E-Cores will be proper scheduling and task assignment. I believe they currently dump things onto the P-Cores and then let it spill over onto the E-Cores. Whereas I think it's in the refresh or the next gen that they'll have management silicon that generally puts tasks onto E-Cores unless they are high-demand.

    • @mikem9536
      @mikem9536 6 місяців тому +1

      My 12700k was doing that by default which was causing dropped frames at 1080p in UA-cam, I was able to change it to the P-cores using Park Control. No more dropped frames.

    • @webx135
      @webx135 6 місяців тому

      @@mikem9536 oh hey I may have to try that out. Maybe do something similar with Parsec to keep it responsive.

  • @mauroavilachable8848
    @mauroavilachable8848 Рік тому +1

    Need a video of only e-cores against the 10th gen intel cpu or 8 e-cores vs i7 -9700 with 8 cores

  • @ah9337
    @ah9337 Місяць тому

    This is one of the best CPU tests on youtube , I liked the mixed test 🔥

  • @annareismith6843
    @annareismith6843 3 місяці тому +2

    Personally, I think it is only a scam. To have something different from AMD yet not be worth much. I have never gone wrong with AMD. It as always been the best for me. I have used Intel on other people's PC's and never bought them to build my own PC. I had a few Intel PC's I got for free and never like them. My first AMD was a 400mhz. Then a single core 1.4ghz, Dual core 2.4ghz, 3.9 to 4.1ghz 4 core APU and now my 12 core 5900X 4.7ghz. I just built a couple mouths ago. I am very happy with it and glad I waited for Zen to get very good. Next will probably be an AMD 9000. I might skip on the 7000's. Though, prices are coming way down on them. I might build one soon if they do more.

  • @antoniocouto7092
    @antoniocouto7092 21 день тому

    All 1700 boards are capable to disable E-core?

  • @dave6800
    @dave6800 Рік тому +1

    Your background task test wasn't fair as you chose the one game that did worse with ecores enabled meaning it wasn't properly optimised for them. Even then while there was still a decrease in avg fps, the 1% lows increased 10%

  • @SteveEricJordan
    @SteveEricJordan 10 місяців тому +1

    this is the best video ever, thank you so much!
    but i would have loved to see the temperature differences when gaming, you only showed us fps.

  • @anuragbishnoi3208
    @anuragbishnoi3208 Рік тому +1

    Such a good , excellent, on point , short video, thank you so much.

  • @rhoads1279
    @rhoads1279 7 місяців тому

    I5 1235u has 2 p cores and 8 e cores.
    I7 1165g7 has 4 p cores 8 threads no e cores.
    Which one is better for learn programming languages

  • @joaorovira
    @joaorovira 29 днів тому

    Amazing work. If you love to do this, I think you shouldn't stop. Thank for sharing your work.
    Keep up 💪

  • @nonaurbizniz7440
    @nonaurbizniz7440 Рік тому +2

    I'm a gamer and wanted pcie 4.0 support so I went with an 11900k since they are so cheap now. Eight strong cores is more than any game needs and I get a better all core oc.

    • @tech_evolved
      @tech_evolved  Рік тому +1

      Even eight cores are sometimes too much haha. But not a bad choice!

    • @Kage0No0Tenshi
      @Kage0No0Tenshi Рік тому +1

      @@tech_evolved well dlss and fsr using CPU to help out GPU what I found out in my benchmarks when I cap fps to same and run native vs dlss balanced and get higher CPU usage on dlss on.
      I got r5-5600x on realise date and used it since that and overclock it then and now runing 4.85Ghz 24/7 but main issue is 6 cores in 3 games I play and need 8 cores at last, have 6 background task like steam, wallpaper engine, msi, icue, hyperx, discord.
      have tested out i7-11700k at stock becuze mb does not support overclock and was slower in one core performance vs my main r5-5600x but my 5600x get ass kicked in multicore.
      8 cores is way to go.
      6 cores sucks
      4 cores is for 4k gaming
      2 cores is for 9000k gaming
      1 core is for Nasa 1961 🙃

    • @tmsphere
      @tmsphere 3 місяці тому

      Yeah, for gaming and minimal background workloads 8 decent cores are more than enough.

  • @davidthacher1397
    @davidthacher1397 Місяць тому

    8 cores delivers 110 extra points while 6 hand out 230. Dropping from 100 to 60Hz due to IO in real world. I suspect those 1 percent lows are synthetic now.

  • @GeoStreber
    @GeoStreber Рік тому +3

    What I'm curious about is virtualization. I'm running a 1240p in my Framework Laptop on Fedora 37, and I wouldn't mind running the VMs I sometimes use exclusively on the E cores.

    • @TheJobieadobe
      @TheJobieadobe 6 місяців тому

      How do you specify which cores your VMs are using?

    • @Navi_Silver
      @Navi_Silver 4 місяці тому +1

      How do you specify cores to use with VMs ?

    • @TheJobieadobe
      @TheJobieadobe 4 місяці тому

      @@Navi_Silver good question

    • @Navi_Silver
      @Navi_Silver 4 місяці тому +1

      @@TheJobieadobe lol, I could say the same thing to you as he didn't seem to answer one year later.

    • @TheJobieadobe
      @TheJobieadobe 4 місяці тому

      @@Navi_Silver in his defense I only asked a month ago but yeah still no answer. So why are you wanting to run VMs on E cores?

  • @pifcudoidef
    @pifcudoidef 5 місяців тому

    fact: 8 ecores from 13700k oc'ed to 4.7ghz, get the same score as an 9900k. The Pcores are much much stronger, the ideea is about the e-cores.

  • @RMKry
    @RMKry 8 місяців тому

    For games only. 1. Disable HT and E-core. 2. Overclock P-cores. 3. Use DDR5 6000 CL30 or better.
    Disable HT +200MHz better oc on P-cores
    Disable E-cores +200MHz better oc on P-cores

  • @minedustry
    @minedustry Рік тому +4

    How good is the performance of e cores in Linux?

  • @user13496
    @user13496 9 місяців тому +1

    It is surprised me that E and P cores can work simultaniously in one app. Before that I thought E cores are almost useless.

    • @davidthacher1397
      @davidthacher1397 Місяць тому

      Software compatibility is pretty much a requirement. The performance can be vary between core types. Priority should be able to manage the rest.

  • @JuliusWise
    @JuliusWise Місяць тому

    Nice video man

  • @a120068020
    @a120068020 3 місяці тому

    Would be nice to have the option of a P core only chip.

  • @lordrushiacanyoupleasesayb3263

    Sir i have a question i5 13500 is 280$ in Bangladesh and the cheapest h610 13thgen mobo is 110$,
    Other hand amd ryzen 5 7600x is 360$ and cheapest am5 mobo is 170$ and 8gb ddr5 ram is 75$ what should i buy i want a pc that can emulate switch ps3 emulator at full speed i dont care about morden aaa game or whatever i just want to play some ps3 and Nintendo switch game at full speed my current config is g4560 😭..... Ok so i am confused my total budget is 500$ i have a ps4 slim and i play most of my game here but i need a strong pc to emulation some ps3 and switch games should i build a ryzen pc or intel i5 13500 pc

    • @tech_evolved
      @tech_evolved  Рік тому +1

      Personally I don’t do emulation but I know the 13500 is a great processor for the price. I’d definitely go with that one if you have budget constraints.

  • @blueyon
    @blueyon 7 місяців тому

    It's not a scam when all CPUs have scores now. The only other option is to buy old CPU's or disable ecore

  • @thetruejay20
    @thetruejay20 6 місяців тому

    Link to that cooler?

  • @fpshooterful
    @fpshooterful Рік тому

    Hi, i am a little bit confused. Are you using 13600k or 13400? Because later in this video i see a 13400 chip.

    • @tech_evolved
      @tech_evolved  Рік тому

      13600k, was using the 13400f as an example

  • @ALPHABYTE94
    @ALPHABYTE94 Рік тому

    What about power draw?

  • @exidrial431
    @exidrial431 Рік тому

    Why did you not test power consumption? Especially Power consumption at idle or near idle is a large area where Intel is ahead of AMD. it's an especially important metric for people inside europe as electricity is by no means cheap there.
    The point of the e-cores is not to be fast, it is to be efficient, to not consume a lot of electricity.

  • @JusstyteN
    @JusstyteN 4 місяці тому

    i wanna see ecores alone power efficiency

  • @BravoSixGoingDark
    @BravoSixGoingDark Рік тому

    On a Intel i5-13400F vs i5-13500 vs i5-13600K // Test in 10 Games benchmark video, a random commenter said E cores causes micro stuttering in games and gave him random BlueScreen in his rig. Can anybody verified this claim? Here is the actual comment of what he said,
    *''Don't do it! I just tried to upgrade to 13500 from 12400F and my games started to stutter, freeze and I even started to get BSOD. I have 128Gb of ram, Z790 Taichi board, Samsung 980 Pro SSD and RTX 4080. Every game ran perfect. I later downgraded to 13400F and it worked great. However, I ended up returning both and kept my 12400F. It is rock stable, easy to cool and easily runs all newest games. I think E-cores are the problem. 13500 has too many E-cores and many games do not know how to handle them. If I was buying a new CPU today, I would go with 12600 (non-K). 6 P-cores at 4.8Ghz or 12400(F) at 4.4Ghz. I don't get BSOD at all with 12400F. All of my drivers and BIOS are updated.''* - From user ''2_UU'' ua-cam.com/video/aLtlFnab17Y/v-deo.html

    • @tech_evolved
      @tech_evolved  Рік тому +1

      That's definitely not a thing. I use an i5 13600k and I haven't noticed any stuttering. E-cores wouldn't cause that unless maybe his OS was out of date or something

    • @BravoSixGoingDark
      @BravoSixGoingDark Рік тому

      @@tech_evolved Interesting... i don't know who to believe anymore lol.

  • @bluegizmo1983
    @bluegizmo1983 Рік тому

    It isn't a matter of the E cores being useless. They obviously contribute to the overall performance. The issue is they could've just as easily left out all the E cores and added two or four more P cores and had even MORE performance than the extra 6 to 8 E cores adds. But instead they choose to add more slower cores because they think core count is all people notice.

    • @tech_evolved
      @tech_evolved  Рік тому

      I'm not sure on the manufacturing process of P vs E but if I had to guess the E cores are a lot cheaper since they are pretty weak in comparison. I do think if intel were to release an 8 P core CPU at the same price as the i5 series it would throw the market in a good way but I don't see any incentive for them to do so unfortunately

    • @od1sseas663
      @od1sseas663 11 місяців тому +2

      It's because 4 E-Cores offer 60% more MT performance for the same die area as 1 P-Core. It isn't marketing, it's efficiency

    • @evilnaruto9229
      @evilnaruto9229 7 місяців тому

      But that’s 4 e cores they could have added atleast 2 p cores and destroyed amd@@od1sseas663

  • @mastfamastfa1256
    @mastfamastfa1256 Рік тому

    i5-13600K or i7-12700k ? (I need it for editing)
    *i7 has 8P cores while the i5 has 6P cores !

    • @tech_evolved
      @tech_evolved  Рік тому

      Tech notice has a great video on this. Check him out

  • @shubhamtripathi2270
    @shubhamtripathi2270 10 місяців тому

    a very perfect video

  • @atom608
    @atom608 Рік тому +2

    So basically E cores are completly useless for gamers and if its anything multi core then its worth a buy? coming from someone with a 13400F and a 3060ti

    • @tech_evolved
      @tech_evolved  Рік тому +1

      Yeah basically. Honestly throughout my testing it made me curious just how many core gamers actually need. I will definitely testing that in the near future.

    • @atom608
      @atom608 Рік тому

      @@tech_evolved Kinda regret my decision buying this almost wished I got a 5700x for the same price haha but good video mate!

    • @cuongtang9539
      @cuongtang9539 Рік тому

      @@atom608 No brah, the e-cores take care of discord etc. its a beast, you dont have to regret it. I have the same cpu its sooo good.

    • @mushbeck
      @mushbeck Рік тому +1

      ​@@atom608 not sure why youd regret it. You are not penalised when playing games and benefit when doing multi core work loads. Assuming you didnt pay a high price compared to the 5700x than you have a really balanced system.

    • @ilovehotdogs125790
      @ilovehotdogs125790 Рік тому

      @@atom608 the 13400 p-cores are also faster than the 5700x. Even though there is only 6 of them, they are 6 faster cores. For gaming it’s better

  • @zhila5958
    @zhila5958 Рік тому

    i needed this video. now i can go sleep in peace haha

  • @theanglerfish
    @theanglerfish Рік тому

    but what is important for gamers is cache and IPC and intel wins in IPC but amd in cache and big cache can make bigger difference because IPC differs only with few %

  • @monyetcerdik9177
    @monyetcerdik9177 Рік тому

    yeah, i don't really need to play games with the background running blender whats? just sell those to design animators or content creators, and gamers don't really need e-core just add p-core lol

  • @russelrenznerosa8969
    @russelrenznerosa8969 Рік тому +8

    AMD is going Big.little route soon. I hope it's enough to regain the better CPU for render.

    • @tringuyen7519
      @tringuyen7519 Рік тому

      Zen 4c isn’t an “E-core”. It’s Zen 3 migrated from 7nm to 5nm and shrunken to fit 16 cores into a single chiplet. It supports multithreading and will work seamlessly with windows!

    • @theanglerfish
      @theanglerfish Рік тому

      @@tringuyen7519 umm...no there are two ccds next to each other as we saw with x3d version extra v-cache have only one ccd...why?🤷‍♂maybe for overclocking or productivity side of things?! maybe?! for me it´s a bit strange even IHS design is ... stupid💁‍♂

    • @marvcyber815
      @marvcyber815 Рік тому

      @@tringuyen7519 I share the same thoughts too, even with zen 5 their "E-Cores" are still roughly speaking "P-Cores" but with less cache that looks eerily similar(at least to me) the laptop version of the cores themselves
      What I would like to see is for AMD to stick some "real" E-Cores into IO die itself so it can achieve some really low idle power consumption, that would be super good on their future chiplet laptop CPUs

  • @raindrops3953
    @raindrops3953 Рік тому +1

    This video really helped thanks ~

  • @AndreZoiaum
    @AndreZoiaum Рік тому +1

    lol useless e-cores, could use their silicon space for more real cores. also the scheduler is crap 3 gens later after their first appearance

  • @CrocoDylianVT
    @CrocoDylianVT Рік тому

    I mean it's kinda weird to have big.LITTLE CPUs, considering that's something that came from phones

  • @dianaalyssa8726
    @dianaalyssa8726 10 місяців тому

    TY

  • @doomsday5286
    @doomsday5286 Рік тому

    Why OVERCLOCK the cpu ? so many benchmarks shows poorer results with OC, next time do on stock if you want real results.(in gaming at least)

    • @tech_evolved
      @tech_evolved  Рік тому

      what benchmarks are showing poorer results OC? Most all I've seen benefit in gaming from it.

  • @eilegz
    @eilegz Рік тому

    The thing its that intel need windows 11, amd option can work well on windows 10.

    • @tech_evolved
      @tech_evolved  Рік тому +1

      Intel works fine on windows 10. They fixed that issue some time ago.

  • @storyanaksekolah2
    @storyanaksekolah2 Рік тому

    i read power consuption review from techpowerup and its confusing how the E cores drew so much power. the E cores of 12900K drew 118 watt on multithreaded test. even the ryzen 9 7900 only drew 150watt. AMD cores on 7000 series non X model are more efficient than intel called E cores.

    • @theanglerfish
      @theanglerfish Рік тому

      umm 5nm process is more efficient than 10nm but yea it´s "e" in quotes and ...e-cores are not multithreaded or i am wrong?

    • @storyanaksekolah2
      @storyanaksekolah2 Рік тому

      @@theanglerfish "multithreadhed" not "hyperthreaded"

    • @selcukkolunsag5304
      @selcukkolunsag5304 Рік тому

      that is wrong test for e core they r for efficiency test them on idle they consume 50 watt less than amd ones on idle if you are 24 hours multi thread amd more efficient if u re normal user intel definetly

    • @storyanaksekolah2
      @storyanaksekolah2 Рік тому +1

      @@selcukkolunsag5304 you need 16 cores for idle 😂?

    • @tmsphere
      @tmsphere 3 місяці тому

      I mean you're comparing grapes to raisins here but ok..

  • @akteadinda2922
    @akteadinda2922 8 місяців тому

    the most shocking for me is that i3, i5, and i7 u series have the same 2 Pcores....wtf.
    i3 1215u has 2P+4E
    i5 1245u has 2P+8E
    i7 1265u has 2P+8E
    and all three of them has similar single core performance (of course, all have the same 2 Pcores)
    i7 only advantages is bigger cache than i3, and lil higher max boost freq . and thats about it.
    i think 4 Ecores is enough, it is for slim laptops anyway. what we need is more Pcores. so i7 u series is clearly stupid choice. id go for i3 if i dont need iris XE graphic

    • @syarifairlangga4608
      @syarifairlangga4608 7 місяців тому

      Lol true.. now the different in P-xore is in the U-P-H-HX naming

  • @Shuttterbugg
    @Shuttterbugg 3 місяці тому

    Everyone is goijg the pc ecore route even apples new m4 is both. Watch amds new procesors gsurentee they will be cuz its proven. I mean they have lpecs now. Like 8 p 8 e and 4 lpe cores

  • @Piketom1
    @Piketom1 Рік тому

    On a desktop, no, they are not worth it. On a laptop, the answer is a little more complicated. If the task scheduler is able to properly utilize the e-cores, then a big/little architecture makes a lot of sense. On a desktop, just give us more performance cores. Power consumption is irrelevant on a desktop.

    • @tech_evolved
      @tech_evolved  Рік тому +2

      I think the point you miss with "just give us more P-cores" is that the alternative to what intel is doing is just having 6 P-cores in the case of an i5. They are literally giving free performance for multitasking for the same price. Because lets be honest, Intel would still charge ~$300 for the i5 13600kf even if it was only P-cores.

  • @karu3067
    @karu3067 Рік тому +1

    It would be great if intel plans on introducing the e-core to the i3 while still having 4 p-core, maybe we can see more benefits for it compared to i5.

    • @tech_evolved
      @tech_evolved  Рік тому +2

      I plan on testing that hypothetical situation soon! Would be interesting to see

    • @karu3067
      @karu3067 Рік тому

      @@tech_evolved Looking forward to that man

    • @JeffWaynee
      @JeffWaynee Рік тому +1

      The 13100 is literally just a re-badged 12100 with a slight overclock. So I wouldn't count on it.

    • @khumanthempoireiton5036
      @khumanthempoireiton5036 11 місяців тому

      They did but only on laptop processor

  • @quoctan1964
    @quoctan1964 Рік тому

    Well... I thought the processor would be blue eyes white dragon @@

  • @marsflee3815
    @marsflee3815 Рік тому

    The last thing you said is the, i5-15600k? That’s what it sounded like. (15th generation is out?)

  • @guitaristkuro8898
    @guitaristkuro8898 Рік тому +2

    Biggest game changer for me is the much higher power draw over AMD. That sealed no Intel for me. I only game so the creative advantages aren’t for me either.

    • @tech_evolved
      @tech_evolved  Рік тому

      Makes sense. As long as your buying for your needs

    • @isipisilemonsqueezy
      @isipisilemonsqueezy Рік тому +1

      Sorry to say but, Intel is more efficient in gaming tasks :D than amd :D abbout 10-30 watts

    • @dracer35
      @dracer35 Рік тому +1

      The problem is that the power draw figures everyone always talks about is at 100% load. If your computer is not running full throttle 100% of the time while its turned on, then those numbers you see are not realistic. This guy shows that intel is actually more power efficient because even though it uses more power at 100% load, its getting the job done faster and then dropping back down to low power faster using less overall power than an AMD CPU using less max power but taking longer and using more energy in the long run. Plus idle power draw is a lot worse on AMD also. I doubt you will spend the few minutes to see what im talking about but if you actually did want to learn, here's the video. ua-cam.com/video/JHWxAdKK4Xg/v-deo.html

    • @tech_evolved
      @tech_evolved  Рік тому

      @@dracer35 not sure if this was intended towards me or not, but thank you for the information! Definitely nice to have this info/sources

    • @dracer35
      @dracer35 Рік тому

      @@tech_evolved Sorry no, it was intended towards Guitarist Kuro because he stated no Intel for him because of higher power draw. There is a lot of misinformation about power draw and it bugs me. But nobody seems to listen. They just run a blender render for a few minutes and blurt out like a dumb caveman "Intel power number higher so Intel bad".
      People will probably think I'm just an Intel fanboy but in reality I've been team AMD most my life because the value was better to me even though the performance was worse in many cases. My first build was with an AMD Athlon 64 X2. (Yes back when dual core CPU's were the big new thing) the last AMD CPU was the Ryzen 9 5900x 12 core CPU. Recently I switched to Intel to try the hybrid core architecture and it has been great. For the games I play, I get more fps and use less power than my old Ryzen and productivity tasks are much faster now. Yes I see power spikes to some high numbers rarely but overall it's quite efficient and I don't mind leaving it idle for a while since it's low power idle is so much better than Ryzen.
      Sorry for the long reply but it annoys me that people are spreading false information.

  • @bergamMNE
    @bergamMNE Рік тому

    E-cores in Audio workload mean nothing.

  • @littlebuch
    @littlebuch Рік тому

    All I'm gathering from this video is how good E-cores are despite what he is saying...

  • @sobatmedhok1306
    @sobatmedhok1306 Рік тому

    The problem of e core is windows scheduler can't fully utilize it so it ended up no effect....
    So maybe Intel should design dedicated hardware for scheduler instead of reliying on firmware....

  • @jasonking1284
    @jasonking1284 Рік тому

    I think it's Intel who need to catch up. All P Cores like AMD will ALWAYS be better than less P Cores + E Cores.....

    • @tech_evolved
      @tech_evolved  Рік тому

      But they aren't less P cores... If you look at actual platform cost instead of just surface level CPU costs you'd realize this.

    • @MLWJ1993
      @MLWJ1993 Рік тому

      So you'd rather have either a much bigger CPU (physically) or much less cores in total.
      You really need to keep in mind that you only get a single P-core for every 4 E-cores...

  • @ALPHABYTE94
    @ALPHABYTE94 Рік тому +1

    4E cores are in place of 1P Core
    Multithreades performance more else 2E cores = 1P cores

  • @Madness801
    @Madness801 Рік тому

    Well intel still beats 90% of amd in benches

  • @ryanvtec3885
    @ryanvtec3885 Рік тому

    8 p cores ftw

  • @adriancoanda9227
    @adriancoanda9227 Рік тому

    Intel are much more stable and you ca throw every ram configuration

  • @masterdftw4983
    @masterdftw4983 Рік тому

    P cores for heavy tasks E cores for lighter i.e. Back ground tasks. E cores do help when gaming so the back ground tasks do steal reasources from P cores.

    • @storyanaksekolah2
      @storyanaksekolah2 Рік тому

      if e cores for light task it should be no more than 4 cores. also the E cores eating so much power for called as "efficient". look at apple, they only put 2 e cores because it really works for efficient task like UI, background notification etc. the number of E cores and it power consumption clearly build to crush AMD on core count and marketing purpose

    • @masterdftw4983
      @masterdftw4983 Рік тому

      @@storyanaksekolah2 I agree but im not the engineer 😁

  • @GodsSoldier2
    @GodsSoldier2 10 місяців тому

    E-Cores are TERRIBLE FOR GAMING! If all you do on your pc is game, TURN E-CORES OFF!!!

  • @adamricecracker7128
    @adamricecracker7128 10 місяців тому

    Discord call white youtube playing , twitch stream running in the Background while Gaming ???? WHHOOOOOAAA thats crazy .
    but Honestly with my Intel Xeon 1231v3 from 10 Years ago and a gtx 1060 i watch Twitch on my second monitor, while i Game Rocket League or else, while i record my Game WHILE using HANDBRAKE! This and even more is something i wanna see be tested if we talking 2023 Year CPUs multicore /multythreading.
    My old Trash CPU can do more in the background while Gaming than that what you tested.
    So if there would be a difference in performance with some youtube and twitch stream in the background i would CRY in your case.
    its like testing my cars fuel consumption with and without a Snickers bar on the passenger seat., you need to stress the COres until you see at what point it does make a difference.

    • @tech_evolved
      @tech_evolved  10 місяців тому

      I ran a render while gaming... what more could you want?

  • @framedthunder6436
    @framedthunder6436 Рік тому

    For desktop they are really useless

    • @tech_evolved
      @tech_evolved  Рік тому

      because...?

    • @ryry187
      @ryry187 Рік тому

      @@tech_evolved because he wants a threadripper

    • @od1sseas663
      @od1sseas663 Рік тому

      @@tech_evolved Because he has amd stocks

  • @uhohwhy
    @uhohwhy Рік тому +1

    Very useful, shame on AMD.

  • @Mr11ESSE111
    @Mr11ESSE111 Рік тому

    they are scam especially in laptops

    • @od1sseas663
      @od1sseas663 Рік тому

      Nope

    • @Mr11ESSE111
      @Mr11ESSE111 Рік тому

      @@od1sseas663 yes they are, performance are garbage for so much cores

    • @od1sseas663
      @od1sseas663 Рік тому +1

      @@Mr11ESSE111 it's not garbage lmfao. 4 E cores use the same die area as 1 P core, yet they offer ~70% more multitreaded performance. Don't say stupid stuff

    • @Mr11ESSE111
      @Mr11ESSE111 Рік тому

      @@od1sseas663 they offer shit,they are barrelly better then 6800h and eats more power

    • @od1sseas663
      @od1sseas663 Рік тому

      @@Mr11ESSE111 it easily destroys 7600x

  • @gonzalomartinez9740
    @gonzalomartinez9740 День тому

    Trash core.

  • @thickdaddymukbang
    @thickdaddymukbang 3 місяці тому

    Why would amd use e-cores when they have efficient 16+ core cpus? Heterogeneous architectures are only good for mobile devices that require low power, not a desktop with a psu! Simply underclocking and undervolting a performance cpu will still be much faster than an e-core. I dont see how AMD is leaving money on the table at all. The tests clearly showed how useless e cores are!