My idea for fixing instability on intel 14th/13th gen K SKUs
Вставка
- Опубліковано 4 тра 2024
- My Patreon: / buildzoid
Teespring: teespring.com/stores/actually...
Bandcamp: machineforscreams.bandcamp.co...
Twitch: / buildzoid
Twitter: / buildzoid1
Facebook: / actuallyhardcoreovercl...
#intel #overclocking #z790 #14900k - Наука та технологія
He just dropped a 2 hours long tutorial for Intel employees
💥💥😂😂💥💥 🎯💯
I can't watch this movie today, let's keep it for tomorrow evening :D
I usually watch through your videos all the way, but a TLDR would be dope in regards to settings. Dope content as per usual 👌
TLDR is if you need a tldr than you are probably better off returning your CPU and going AMD.
@ProIific What are you even going on about...? 🤣 My setup is stable, but I trust his opinion as he prolly has more knowledge on this particular subject than me. Now go AMD fanboy else where.
@@matg1784 If you are a true enthusiast then watch the video. No one is going spoon feed you brother.
@@matg1784 idk, why some people have to act like complete assholes to a perfectly reasonable request, but there is a shorter half-hour version now on the channel.
@@kosmosyche I noticed this late yesterday that he uploaded a shorter version! But yeah, people just want to be dicks online for no reason. Appreciate the heads up brotha.
I enjoy watching this "Short" 1 hour, 45 minute video. I just wish there was a TLDR at some point.. D:
Man… I’ve been running just a Lite Load adjustment on my i9-13900K on an MSI Z699 Edge WiFi DDR4,, this is in Water Cooling Mode 4096w. It has been stable on a 55/44 OC. Caps out at 310-320w (peak) but averages around 288w on CB23 loops. 41000 points, 360 AIO. Temps max out right at 96c.
Lite Load mode worked for me, but what a train wreck this has become. Keep up the good work with the videos.
What Lite load adjustment did you make?
4096w isn't enough for a modern intel cpu
@@TheRufa1990 1 :D
Same, I've been running 5.8p/4.5e/5.0ring since Nov 2022 on a cheap Z790 Tomahawk. I used Lite Load to tune voltage and things have been great for almost 2yrs now.
That’s crazy. I top out at 80°C(+/-3°) with a 240mm on my 14900K 40000 Cinebench on a 10 minute burn in. 54/44
- IccMax is a *peak* current that CPU *should not* exceed so it reduces frequency aggresively, spec says that it is mostly for VRM designers - to select some real-application inductors / decoupling capacitors (because to handle "infinite" currents you need "infinite" capacitors)
- If CPU is hitting any of specified limits (PL1/PL2/IccMax) then it will chose common VID for P/E/Ring - there is no need to limit their ratios to achieve that
- If CPU does not have IccMax/PL1/PL2 set then it will chose highest (-suboptimal) VID
- Ratio between IccMax set and averaged VRM(TDC) current is about 0.62...0.66 depending from load
- Apps with less *peak* currents (like CB15) will be throttled by IccMax less than high-current ones (like CB23) so may consume more power (yes, CB15 > CB23) but once you remove IccMax - it will reverse
The issue is the Ecores will request a higher VID as soon as any of the limits aren't being hit so in some scenarios you'll get the Vcore bouncing between the Pcore and Ecore VID which makes the Pcores run even hotter than if the E-cores just followed P-core vid. Also considering how little the E-cores help performance I see no reason they should ever be able to set the operating voltage for the rest of the chip.
@ActuallyHardcoreOverclocking It depends from fused VIDs, from my experience it is a Cache VID most of the time. Also, if you have current distribution like 10amps P-cores (6.0GHz idle) + 60amps e-cores (moderate load) + 30 amps ring - you will have excessive voltage because that all 100amp current will be applied to AC Loadline to highest VID (P-core @ 6.0GHz) - so it might be not just VID issue but "side-band" (E/Ring) current * AC_loadline + "6.0GHz VID" issue
@@ActuallyHardcoreOverclocking quick question, does setting IccMax in "CORE CURRENT LIMIT(AMPS)" produce different results than setting it in IA VR CONFIG "IA VR CURRENT LIMIT" or does that not even matter? I've been paying attention to these GB videos because basically only you are covering an up to date bios. I'm on the z690 elite ax and for a year and half of owning this thing cant find a single video that truly dives into the bios.
@@_preshuh ICC_MAX is an internal CPU current limit. The VR current limit if is for the VRM. So afaik it doesn't matter unless you have issues with the VRM on your board overheating.
@@_preshuh VR initiated current throtling is very agressive, similar to asserting PROCHOT and not gradual like IccMax/PL1/2 so its basically unusable for interactive daily system, it will slow down way too much.
I hope Derbauer, with his apparent contacts at Intel, will be able to directly ask an Intel engineer about these issues. I’d like to get some real answers.
The problem with that is people in big companies only know what they're working on and the public line.
The info from sources at retailers and OEMs on customer instability issues and RMAs is more trustworthy, it's increasing and they're pissed. Naturally you need to collect from many, not just the ones yelling the loudest.
This is a hot issue and an Intel engineer can either say they know nuffin or put their career on the line.
@@RobBCactive maybe Intel could surprise us all and engage on this issue in good faith. We shall see. My alder lake experience has been fantastic. I live in WA/USA, so it’s some of the cheapest electricity in the country. Rock solid and run 24/7. Very happy with it.
@@Akkbar21 well my electricity is more expensive and I have an x3D CPU which bests Alder Lake and is very efficient, so I am very happy too. I was speaking as someone who was a software engineer in a multi-national making CAD software for chips. We even still had a 12V process that car manufacturers could hook up directly to the battery for their engine management chips without using converters.
@@RobBCactive sounds cool. AM5 x3d is an amazing cpu. Enjoy. At the time I bought my new system, alder lake was a better choice for workloads other than just gaming.
@@Akkbar21 well I am on AM4, ADL never made sense as an upgrade. But you have the most reliable series, at the time on launch it was pretty clear they'd sacrificed on AVX512 and the power usage was a problem.
With RPL I expected them to clock the E cores lower to waste less energy, but they tried to match Zen4 instead and ramped up the power usage even more. It's no surprise that letting mobo firms choose settings would end badly.
Vdrop=R(load line)*ICC, cutting down ICC may help voltage drop under load. I tend to leave load line ac/dc as default and just set LLC to extreme or flat. Incorrect ac/dc load line may cause voltage overshoot.
rip to the hige-end heat-tel
Hey BZ, good thing you gave your viewers some remedies to help out with the Intel 14th and 13th gen K skus. Keep uploading YT content to assist us on issues that we may have troubleshooting glitches in our operating systems.🎉😊😊
BZ, can you pretty please drop a quick video for these settings on your Asus board? I have been playing around with my Rog Strix Z790-E and can follow the Intel documentation you present and translate almost all of what you're doing here from Gigabyte to Asus, but with this stuff the devil is in the details and I am still getting super high voltages/heat. The interplay between IA CEP and LLC/AC_LL/DC_LL is what's getting me, I think. I am getting really high voltages and heat. On a fresh BIOS, I'll set intel defaults, so I have MCE disabled (I change SVID Behavior from Intel Failsafe to Auto).
Should I be using LLC 3 even when it pushes the intel limits? Would LLC4 or LLC 5 be better? Like what is the Asus equivalent of the LLC settings you are using?
Can we just, like, never undervolt again?
/confused (I'll admit it, lol)
I already RMA'd one of these 14900K chips getting caught up in the issues that have come to light, and do not want to lose another chip :)
If anyone can help, I can start posting settings, but I do not want this post to get much longer.
mce = disable enforce all limits;
llc auto;
ac = max specified by intel;
dc = ac;
ia/dc cep on;
svid = intel fail safe (you can try others);
iccmax unlimited bit = off;
201a cache/core limit (or maximum from intel baseline for your cpu);
default powerlimits pl1/2 from intel baseline for your cpu;
intel tdp limits on(or auto for asus tdp profile);
off the top of my head the most important.
tvb optimisations - on;
avx on (you can leave on auto)
You can reduce peak voltage with v/f offset.
BUT
you can still end up with a high vcore, and you can further reduce your voltage and throttling by limiting e-cores. It will reduce vid requests and lowers the voltage requirements because they share the same voltage with p-cores. I have noticed that high clocks on e-cores raises voltage on already high clocked p-cores and that in turn increases temps and causes throttling.
If thermal limit is reached and for example you are at a point of throttling already, then the p-cores will get clocks dropped first till e-cores will start dropping clocks. That means, that the voltage will stay the same during this period, because e-cores clocks are unchanged and vid voltage requests are unchanged. That is because core voltage of e-cores on high frequencies can be higher then core voltage of p-cores on high frequencies. So the temperature drop doesnt happen and p-core clocks are already reduced. So you get a further drop of clock speeds.
I dont really know exactly how it works, just guessing, but that is the frequently observed behavior when you overclock e-cores AND p-cores together.
I have given up with all the fine tuning and instead have settled on setting PL1 and PL2 limits to 253w and power limit to 307A.
Very safe option 👍
I don’t really care about Cinebench scores
Gute Besserung!
I'm just going to add a curio here: passing stress tests, but I was getting constant crashes/errors in Premiere and Topaz with my 14900k (253W/400A limit), using the Z790 Master. I did nothing else but change from 'Auto' voltages to adaptive 'Legacy' and manually entered a 0.00 offset, and the crashes are completely gone, with no real change in temps or performance. I'm talking about going from crashing every render to doing 30 in a row with no issues.
No idea why this minor change (which seemingly shouldn't have done anything) brought me back into stability, but it's obvious that some processors are going to require a lot less tweaking than others. Seems like mobo manufacturers just need to bring things up to a sensible baseline - the profiles they released so far are not sensible, particularly from Gigabyte.
yeah I just tried this and cinebench R15 still locks up.
@@ActuallyHardcoreOverclocking I never had any issues with R23/R15 or y-cruncher for instance, just real-world renders where the power limit wasn't clamping down the frequencies as much - running closer to 5.7/4.4. No idea why it fixed my issue.
Why does the voltage offset show as +0.079? Is that a bug?
🙂 cool, thanks for the tips
hello there i am running intel base line Profile Bios ver 3501 on my Z690 maximus Extreme the issue i have is my GPU is Running at x8 instead of x16 and yeah You Do Have idea About i9 12 /13 /14 Gen Pcie Configuration for Gpu And M.2 On Z690 motherboards As i Should Be Using the Top Picex16_1 for Gpu and M.2_2 For SSD Right i am already Doing That For 2 years Now But Now After Bios Update ( I Think Because i haven't Checked it for years ) My Gpu Is Running At x8 instead Of x16 Why is That ? i Have Rog Strix 4090 gaming Non Oc And i just Put The Oc Vbios on it Ver ; 95. 02.18.80.89 And its Been Working So Good After Shutting The PC With No AC Power It Will Switch To x16 Again and will Stays There For many Restarts Until Ac Power Shutdown I think So And yeah i have Disable The PCIE Native Power Management . i need Help I think Maybe For Some Reason It Comes To My Mind Could It Be The Cmos Cell Not Working Properly When There Is No Ac Power i Think So
Thanks for video from the future. Can't wait what we will hear in 2045. :)
nice to meet you
I always look forward to watching your videos.
I run a small youtube channel
I'm making a video about Intel's issues with the 13th and 14th generations.
This video was very helpful
I would like to include a link to this video in the video summary section. Is that okay?
I would appreciate your reply
I look forward to seeing your videos in the future.
Buildzoid, why do you never mention temperature fluctuations an voltage fluctuations in relationship with degradation?
At 9 minutes in video I think the stock loadline is too low. The vdroop is letting the volts drop below stability under heavy load. Too much vdroop and not enough volts at the high power use of high clock full load. I think the fix is higher LLC (and maybe a bit extra volts at top clocks) and reducing the volts to as low as stable with that. And maybe setting a top limit on the volts in the bios.
my friend's 13700k is unstable at default, looking forward to this
My i9 14900k finishing Cinebench R15 without any problems on aorus mobo defaults. So this problem occurs not on every cpu ?
That's right
Is your CEP settings on or auto?
yes because silicon lottery and lack of proper guidelines
if you do end up having glitchy behavior you should also adjust your settings because last thing you want is a dead CPU from what intel considers "stock operation"
issues surfaces when the cpu starts to degrade, which means not everybody will have the same experience or timing on when their system will start to be unstable
@@emini6 both CEP - auto. PLimits 1(280W) 2(unlimited). During Cinebench R15 CPU draws around 340W on one loop multicore, score around 6k, in R23 around 40k. Seems good?
I use this to curb the thermal meltdown on my X299 systems and I use ASUS mobos only giving them a limit on CEP. My mates 9900K I notice that ASUS bios love to throw a ton of voltage.
會超頻的其實自己調比較好
This is all awesome 👌. However if u must go through this length to tune an Intel chip... for the most average user they have no idea.
I've had manually set all core and voltage since Day 1 on a 13700K +Z-690A pro DDR4 (5.5ghz P core, 1.27Vcore, 1.3SA, 4.2 E cores. Maybe i hit a silicon lottery?). Is this something people could also be looking at instead of letting the system bounce around like mad, or are us static boys in for a reckoning? I mean, I've had no problems for over a year, and temps have been great even under full 253W load....but I may go into bios and check some of these settings; being on a lower end board i imagine most protections were kept in place - i mean the maximum of 253W makes me think so.
either way , 14th gen seems so haphazard, especially the i9's. Just nonsensical chips....
14900k with z790 riptide wifi recommendations ?
I don't have any asrock boards.
i have not seen the tvb settings before now -
- my asus z690-e has these settings ->
TVB Voltage Optimizations [Auto]
Enhanced TVB [Auto]
Overclocking TVB [Auto]
Overclocking TVB Global Temperature Offset Sign [+]
Overclocking TVB Global Temperature Offset Value [Auto]
- after that it has 11 of these settings ( 1 - 11 )
Offset Mode Sign 1 [+]
V/F Point 1 Offset [Auto]
Offset Mode Sign 2 [+]
V/F Point 2 Offset [Auto]
... 3 4 5 6 7 8 9 10 11
i have no idea what those are about
Scotty, give me more power! Are you mad captain, she's going to burn up!
the 14900ks was not even supposed to launch, they launched it just to save face and avoid a cancellation but 1.5 is def not safe. even 1.4 which runs at idle is not good at all, i think default settings should be following stated tdps with 253w pl2.
I'm confused - for the first year or more when these chips came out, everyone, everywhere was "Undervolt! Undervolt!" and I did so ... run 55/43 on all cores with HT disabled, full load temps well under 100. When gaming usually in the low 60's. I had some intermittent issues at 56 but backing down to 55 cured it for me. And, this is with an undervolt. By Aorus Z790 certainly was not "undervolted" when I received it, as my voltage now are considerably lower. Isn't the real issue here that these MB manufacturers essentially overclocked everything by default, out of the box?
Intel allowed them to do so, so it's both of their fault.
can we have a summary of settings done at the end of video or a pinned comment
Tried that, didn't work unfortunately.
Right now I am in a boat where 14900KS can't even boot windows without BSOD / file corruption etc with default MB settings. I need AC/DC just to boot. And it wasn't like that from the beginning but I slowly got there within a month. For now I am using 65/90 AC/DC loadline (it's in between "performance" and "turbo") and "medium" vcore loadline with all the "disabled by default" intel features like TVB, eTVB, 400 current limit, 320w limit etc. My R23 scores are about 38k.
With this method I actually can boot windows and even have 40.5k scores - but most of the games crashes with ACCESS_VIOLATION error. For example cyberpunk crashes immediately after launch. That's fucking sucks. CEP did not fix that either.
I also had the same errors / stability issues on 13900KS even with stock clocks / stock limits. same TVB/eTVB enabled and it was reported numerous times on intel forum yet no reaction whatsoever. Only when nvidia told everyone to go bother intel - they started to move, though first thing they did is blamed motherboard manufacturers.
For now whole situation around 13th/14th gen makes me think of buying AMD... I never had so much troubles with my systems in any time in the past, this is ridiculous.
you have XMP disabled for testing this right? How much do you need to drop the P-core ratio before you stop getting the ACCESS_VIOLATION?
@@ActuallyHardcoreOverclocking no, for extreme LLC I didn't disable XMP. Though I tried to disable it in the past with zero effect. Just to confirm that it's not memory / system agent etc.
I haven't tried to drop clocks with "default" voltage, for now I am staying at "Medium LLC" and 65/90 AC/DC LL, clocks dropped to 60/58. Plus eTVB drops clocks for all core load further to 56 and with such high LLC CPU consumes 360ish watts. Or with 320w limit (which has no effect on ACCESS_VIOLATION) - drops further to 53.
Current limit (right now 400), even 307, also didn't improve crashes. Neither does CEP.
I also know that simply removing 62 also doesn't help, still getting 0xc0000005 with 59/59. So that's why I dropped to 60/58.
So the problem is that even if system is stable under full load or one core load or even if system is stable under full load with limited cooling (pumps running at minimum rpm) and thermal throttling - it doesn't mean it will be stable in games. I think the worst offender in my case is Armored Core 6 right now: it loves to crash when you transit from hangar to loading screen via cutscene: simple camera switch in the same rendered hangar, mech starts to move and briefing monologue starts. But from cpu load standpoint it doesn't make any sense.
So probably there is something extremely wrong happening with idle / low load voltages and light load jumping around the cores.
As for 13900ks - I haven't managed to fix it properly, still had errors with stock clocks, tvb/etvb enabled, 60/90 AC/DC LL ("performance") and hoped to fix it with 14900ks. Ha-ha.
But at least with "default" bios 13900ks didn't have issues booting windows. And it lived for a year at least like that, while right now it feels like my 14900ks just dying already.
Additional frustration is that gigabyte still doesn't show biscuits for 14900ks so I don't even know should I blame silicone quality or what.
by the way - ActuallyHardcoreOverclocking - if you hold the shift key down when selecting restart windows - it will go into recovery and there you can tell it to load the bios when it boots - that way you don't have to spam the f2 or del button to get cmos to load - you can also make a cmd file with
shutdown -r -f -o -t 0
in it - that will restart and boot into recovery - i think if you don't have recovery set up - it will just go to the bios instead of giving you the menu to make choices - try it and see if it helps
a couple of years ago - i built this machine with asus rog strix z690-e mb - 360 aio cooling and 13900k cpu - several months ago - i put in a 14900k
- all this time - i have been ignorantly wondering why i have been having more problems in the last two years than the previous 40 years combined - yes i started with computers way BEFORE windows - even before DOS ( actually in 1976 ) - then i found out about the cpu problems and i think that must be at least 50 percent of the problem - i think that windows 10 and 11 have their own troubles too ...
- with the 1300 - i had a lot less trouble with oc - with the 1400 i finally just set the p cores to 55 and the e cores to 42 and current and powers to
CPU Core/Cache Current Limit Max. [307.00]
Long Duration Package Power Limit [253]
Package Power Time Window [Auto]
Short Duration Package Power Limit [253]
-
i also set the offset voltage and adaptive voltage to negative values
Actual VRM Core Voltage [Offset Mode]
- Offset Mode Sign [-]
- CPU Core Voltage Offset [0.04500]
Global Core SVID Voltage [Adaptive Mode]
- Offset Mode Sign [-]
- Additional Turbo Mode CPU Core Voltage [Auto]
- Offset Voltage [0.10000]
-
and CPU Core Auto Voltage Cap [1.30000]
-
it runs great - but i only get about 37,000 on cb23 - however my temps stay below 75 c - and my vcore stays below 1.3 and the wattage doesn't go above the limit set of 253 or 307 watts
the only game i play anymore is call of duty -
i am using a 2560 x 1440 screen - rtx 3090 gpu - i get 200 - 300 fps in the game - which is fine - during the game i hardly ever draw over 200 watts cpu - and temps stay below 65
i asked chatlbgtxyzump about the vcore for the i9-14900k and it said
1.3 and lower is good
1.3 to 1.4 is high
1.4 to 1.5 is higher
1.5 and over is bad
so not very definitive - just as is intel information
What happened to if you can cool it properly run the clock. This rule is universal and hasn't gone away. Forget everything else if you can keep it below 80C under full load that's the safest clock your CPU can run all core. P Core and E Core enabled.
The moment intel decided that efficiency and architectural innovation is not the answer, these problems are bound to happen.
Yeah, I'll just stick with AMD for now 🤣
new bios update F12d fixes everything
I think 13700k is 'safe'. Please confirm.
My 13700k needs 0,020V extra in Intel XTU to be stable at underclocking, standard speed and overclocking (Gigabyte motherboard).
my friend's 13700k is unstable at default msi settings
I have a 12900k and either I suck at tuning or I really lost the silicon lottery. On a 30 minute cinebench run my hotttest core gets to 89 and the rest somewehere from around low to mid 80's. I can not run prime 95 small fft without throttling unless i disable e cores. Just impossible. (Motherboard is an evga z690 dark). Im using a lian li 360mm aio and a thermal grizzly contact frame.
Assuming Gigabytes DC is correct for auto was a bad move its just an arbitrary value of 0.9 that's locked in for any load line setting when it should be 1.1
Hopefully keep the performance and its something intel/board partners can bite off. Sorry, copy. Too many diss videos today.
doesn't work for me
some details would be nice.
Same settings as you. I have a 13900 KF and Z790 Elite AX But i only get 19000 points in cinebnech 23.
And if I set it to 370 amps I have to give it 25 volts more to make it stable
@@ActuallyHardcoreOverclockingFound a solution. Gigabyte has a new bios F11b. they fixed it
@@onlinegaming4826 I'd guess the F11g BIOS actually just disabled CEP. Because the performance drop happens when CEP thinks the CPU isn't getting enough voltage.
casual user: 1.40v * 300 amps is too much for a cpu
cpu: nope it is whatever watts i say so
voltage regulator: hi
motherboard vrms: hi
decoupling capacitors: yeyyeyeyiii
cpu: 250w
casual user: 🤔
Interesting content from Buildzoid, but dog… you need an editor to distill some of this stuff down to its valuable info, theories and conclusions. I don’t have hours to watch this stuff and pay attention.
The PDF name is bait
Easiest way? Switch to 7800 X3D-
Can you upload a shorter version? Or can you write down the point of this video? It's too long for me. Thanx bro
sure random youtube commenter! that ever work or?
Watch on 2x
This is BZ; the video is as long as it needs to be. Even Tech Jesus failed to impose their will on this issue. 🤣🤣🤣What hope does A. Rando have?
There is a new shorter version have arrived….chill bros. He is a nice guy. I knew that.
Pretty funny set of comments in the end, peace!
13700k on a z690m pg riptide here ...
never a single issue on stability, cheap thermalright cooler (and contact frame) , good temps and never a single problem. Same with all 12 13 and 14 gen builds for clients ... 0 problems.
It’s mostly i9 that is having these issues, not i7
turns out that when your CPU only turbos upto 5.4 ST and 5.3 MT it tends to be more stable than CPUs that boost to 5.8-6.2GHz. Imagine that.
My 12700k has been rock solid as well with no PL at default Asus MCE. No surprise there.
The issues are on 13900k/14700k and 14900K/KS
@@ActuallyHardcoreOverclocking My totally unrelated 12600K is also working fine! Can you believe it!
Are you serious ?
Why all the messing around all your going to do is confuse alot of people Just set the LLC to Standard, AC to 0.6/0.7, DC to 1.1 and your done, I mean it's a fun video for me but alot of people dont get it
because if you just raise the AC_LL to fix stability the core voltages go up and if the chips are degrading that's gonna kill them faster.
@@ActuallyHardcoreOverclocking i see your point but the reason for your instability is the stupidly low AC_LL and not degradation that Gigabyte never should have set on 14th gen
12900K they set 0.7, 13900K they Set 0.6, 14900K they set 0.4 most chips simply arnt going to be stable at 0.4 my 13900K Certainly wouldnt be 0.6 is as low as it can go stable so the propper fix is just to bring it more inline with intel AC_LL Spec
Take LLC Standard,, AC_LL 0.6/0.7, DC_LL 1.1 for a ride and see how you like it without messing with anything else
Intel give the board manufacturers tools to check socket impedance on a board design so they can set AC_LL lower than 1.1 mOhm's by validating the design if a board is set for 0.7 mOhm's for 12th gen it should be the same for 13th and 14th as the presumption is that board was validated by Gigabyte for 0.7 mOhm's but the reality i think they are just punching in random values until instability on their sample and doing zero validation at the socket
all Gigabyte Z690/Z790's are 0.7 0.6 0.4 Depending on the gen cpu, this alone says it hasnt been validated as every board model isnt going to use the same values
Thank you for the hint. Set AC DC LL manually and could move from LLC 6 to Level 1 even with a 50mv UV. Single cores hitting thermal limit is gone. Performance is the same. Perfectly ❤
The solution: 7800X3D
Not everyone just plays games.
@@skywalkies77 Agreed, but The 8 core 7800X3D is still plenty fast for apps. If you need more, 7950X3D. Plus you get Gen 5 NVMe and future cpu upgrade in socket capability. No 300W cooling and endless tinkering for stability required.
@@jrherita I like ram faster than 6000mhz and I need Quicksync for my exporting.
I haven’t had a single issue with my 14900k.
Edit : Also the AM5 platform is only likely to be around till next year so it’s hardly future-proof. No hardware is.
@@skywalkies77 1 year is still longer than LGA 1700. Also no AVX512 on 12th-14th gen. I’ve used Intel my entire life, my last AMD system before the 7800X3D was (no joke) an AMD 5x86-133 OC’d to 160 MHz from 1995-1997. But the power consumption of the current i9’s for the performance you get is really insane and not worth it. It’s 3-5% faster in apps than 7950X3D for 3 times the power and a dated platform. Nah thanks.
Don't buy one in the first place!?
U talk to much and give no real answers… ur just a blah blah blah!
how you minimize degradation, buy a 7800X3D or a 7950X3D and be done with it.
Fucking simple.
Either 7800x3d or 7950x.
@@vincentvega3093 I play games so the X3D 7950 would be what i would pick but either or at that point honestly.
@@vincentvega3093 7950x3d adjusts now to use right ccd, just need to alt G quickly ensure gamebar is detecting the game, if not tick the box that forces it and boom works as good or better than 7800x3d
My idea for fixing Intel instability - Buy AMD which also fixes the inefficiency problem too
Package power = 340 watts. lmfao! Amd 7800x3d package power = 90 watts
no
#
Imagine being midwit enough to compare a low voltage and controlled chip like the X3D to a chip with way more cores and aggressive power limits... Not even same category. Hell 14900KS and to a stretch, the 14900K are moronic CPU's for 99% of users. It's like complaining that a Ford GT's 10 cylinders consume too much gas compared to a Civic Type R when doing 120km/hr on the highway.
@@Greez1337 you telling me that an 8 core CPU tends to consume overall less power than a 24 core CPU? mindblowing. What's next? that running 5.7Ghz needs significantly more power than running 5Ghz. Some people needs fact checking: they probably be surprised of what an i9 can pull off with 90W-125W, surprised that it can run it's 24 cores 4.5/3.8Ghz on very heavy loads and it can hit its turbo boost 2.0 advertised frequencies in games and not even hit 90W power draw.
The best thing to fix Intel instability is to not get one for starters. Intel is done and people have to move over.
That very unbiast profile picture is a really stong argument
@Takashita_Sukakoki hahaha. His autism is showing.
Ah yes, "No this big company is ded, please send your money to other uncaring faceless corp"
Dude forgets 5 minutes ago when AMD chips were literally blowing up.
Not to mention literally no one buys AMD anything. Its not even a competitor. Check AMDs graphics card sales, too. Its not as stable as Intel full stop.
My idea is safe and effective...use AMD.
no