I am glad that I have a 12th gen i5 in my gaming PC that I built in 2022. Been using it for 2 years now and have had no issues at all. The performance has been amazing even under heavy loads.
I think it's finally time for TechTube to teach users about the BIOS. It is the most untapped segment of information and countless users suffer because of it. So many issues could be resolved if people understood what is happening in their bios.
The largest issue is that there are so many names for the same settings, even across the same OEM and chipset. You can tell a person "Enable XMP" but they might have an AMD board and not realize EXPO is basically the same thing.
For my 13900k and 14900k I have done the following: MCE off, PL1 and PL2 limit to 225, limit P-core boost to 5.5 GHz and E-core boost to 4.3GHz, and use balanced power profile in Windows (although I do disable core parking to keep system highly responsive). Oh and just XMP on the RAM. I didn’t change LLC value. I have set voltage offset at a modest -0.015v and set the Core limit to 300 Amps. I have disabled the C6&C7 C states and EIST. Lastly I have locked AVX at 0 offset. I have tested on P95, CB R23 and CB R15. All great and in a mid 20 degree room, no workload exceeds 80c on package or cores. Very happy and benchmarks are very close to where they were before taming these beasts.
U made too many changes just keep the amps at 307 and 253 for pl1 and pl2 thats all u have to do i never had issues since the cpu came out i had it on launch and always configure my bios which many dont do its users fault not intel fault its like a car u have to put the bolts on the wheels dont expect it to work without fine tuning
@@fortnite360HZno it’s intel and the motherboard manufacturers fault. It is completely reasonable for buyers to expect for their products to work stable right outside the box.
Owner of an i7-12700K, on an Asus TUF Gaming Z690-Plus D4, with 32GB of DDR4 at 3600Mhz CL16. No problem at all. And, yes, the Ryzen 5 5600X, that is in my wife's computer looks more snappy at the mouse and keyboard, but that's just this. Even on the internet, the Intel is way more fast!
@@josetrigueiro5978 12th Gen does not have these same reliability issues. It does have the lag maybe because of PCH not on CPU on on mobo which is mentioned in this video. 12th Gen is in fact much more reliable than 13th and 14th Gen regarding stability. Especially DDR4. Now DDR5 XMP on other hand is another story but you are using DDR4 XMP which is smooth and stable on 12th Gen. DDR5 SPD 4800 probably ok but DDR5 XMP on 12th to 14th Gen is terrible on 4 DIMM boards especially Asus from my experience.
@@Wolverine607 I know that. My message was just a response to Savitarax who in his message mentioned the 12th generation as having the same problems as the 13th generation.
@@josetrigueiro5978-_- my initial experience on this cpu has not been that good but it ironed itself out I guess. I’m actually considering am5 next year as I think I’d benefit more from that snappiness
@@josetrigueiro5978 No IT IS Not way more fast(er) ob the Internet 😂😂😂 Tell your wife Not to watch so much porn in the Internet then the PC wont Catch so much mailware
11:39 This is the main reason why learning to overclock and overclocking properly is much more safer for components long term than leaving everything stock without knowing what da hell your motherboard is doing. This is definitely NOT the way things should be but it's been the case since i had used computers. For example my Z370 from ASUS when enabling the XMP configured the VCCSA to 1.35v (overclockers recommend using 1.25v for daily). And this is one example, motherboards have this sort of practices. Vcore, VCCSA, VCCIN, any voltage... you name it. And that's only half the story, i've heard many influencers saying things like "a CPU running at 100 ºC is fine", i can assure you is not, amongs overclockers the safe margin has allways been around 85ºC.
@@Kdog307 See, I put my i9KF into a Corsair 4000d and the same 360AIO that was in a 2015 Full tower from Newegg. After the swap to the 4000d the idle went up 10C instantly almost 38C - 40C on all cores. Wasn't throttling at 5.2GHZ inside the cheap Full tower; where my 13700K sits now; in that old Full tower. It choked that 9900Kf down to 4.4Ghz before i could get it not to throttle in R23. So that case is jinxed for a 360 in the front.... Dumb thing is, that case will really need the front fans as exhaust and using a 240 AIO in the top, and get way more cooling, but your fans in the glass up front will be backward. My 13700K would never stand a chance in the Corsair 4000d.
On mine I set the following (on an i7-13700K): - Max TDP of 70'C - PL1 of 153 W - PL2 of 153 W - Disable ASUS Multicore Enhancement, choose Intel Default - Disable Thermal Velocity Boost - SVID to be Auto (Intel Baseline is way too high, about 1.7 V) I'm more interested in a quiet system than high power, nothing I do requires crazy high levels of performance. I tend to run my systems for 20+ years (I still have my P4 Win98 system for retro gaming), so silicon degradation is very important to me
Hey man good video as always! Thanks for the shoutout. Yeah there are just so many variables here. PC gaming/building should not be this complex. I hope Intel can have better luck next generation with their new architecture. As for now, I'm back on AMD's AM5 and I can say that it is much better than it was a year ago. My system feels super snappy. I keep expecting to see some lag or stutter any second lol. But thankfully nothing. Phew. I'll keep you updated! Keep up the good work and so close to 20K let's go!!!!!!!
I had crashes, blue screens, and most of these issues when bench stress testing with other applications. All started with new components, and the new I-9 CPU. Finally figured out that it was the unrealistic BIOS settings by MSI. I contacted them, and ended up in arguments with them because they would not provide the correct settings for the BIOS, for normal usage and OC usage. I was shocked. Got the system all stable and running at a good performance, after changing the voltage and the current to the CPU, and undervoltng. I say the makers of the motherboards are the problem.
1.42 volts on my i7-12700K. At 5GHz, instant 90 degrees on Cinebench with a Deepcool Assassin 3 and Corsair 5000D Airflow. I did -110mv undervolt, now is 1.23 volts max and 67 degrees.
Damn Danny, I didn't know how much I needed this video. I've been chasing problems with with 13700 z790 DD5-7200 4090 system forever. Thanks for the great insight
Same I have been having these issues for months. I have updated to the latest bios with Intel's base limit still hasn't fixed my problem I think my CPU is broken. I will try getting a rma :(
I kept my AM4 mobo and got a 5700X just because of the 65W TDP of that CPU. I try to run my system as efficient as possible. Even though i recommended my wife a laptop with a 12600H because i know that CPU is great. I don't know why intel pushed their high end cpu's so far, having really good gems in the current i5 series.
They are linking the high end user, causing all out power burnout; with the bucket of parts builder that has poopoo for advice of hardware... Makes for a nice ploy to sell more AMD.... 13700K is a cranky cpu to an extent, but you could throw it and a 360AIO in a Corsair 4000d and cook the heck out of it, after it not throttling at all in a 2015 Newegg budget full tower PC case.... Just changing cases will add 10C to idle alone. The glass in the front makes that much difference. You'll cook a 9900KF down to 4.3 Ghz to keep it from throttling in the summer. Or run 5.1 or 5.2Ghz year around in the old cheap 2015 case. I was testing it in a 50C room, that Intel chip wasn't doing one thing to heat that room though. Even added 2 120mm to the rear of the Radiator and didn't change one number on temps.... All a flow issue.
That’s great, if you want to have maximum performance you need to understand what is going on under the hood, not just throwing money into your build. Well done intel!
if i leave my gigabyte bios on stock, i see in hw monitor when running CB23 that VCORE goes above 1.5V, temps low 90s and power draw almost 400 watts. This is after delid too. Going in and adjusting LLC and setting a fixed voltage i was able to run stock speeds at 1.27v with Turbo LLC and vcore stays pegged at that in hwmonitor. So these chips are being massively overvolted by motherboard manufacturers. Another thing to do if you dont want to run a fixed vcore you can set PL1 to 125 and PL2 to 253 watts. Dont need to adjust current limit as itll never hit 307A at stock. Also throw on a -0.075mv undervolt and its rock stable. Max temps in CB now low 70s too with a 360mm aio.
So far I don't have this issue on my setup. I just set my PL1=PL2 to 253W and ICC=307A, Enhanced Turbo=Auto and the rest is custom fan setup curve. I have an MSI MAG Z790 Tomahawk MAX WIFI7 with i9-14900K. Immediately after I finished my build I set my PL1 & PL2 to the Intel power specs and had not had any issues. Running God of War 2018 and finished the entire game with no hickups. If you set the PL1 and PL2 to the default BIOS setting which is an overkill setup, yea the PC will crash and immediately causes the BSOD or sometimes lag and memory error issues. You can set PL1=125w and PL2=253W however this will definitely reduce the performance of the CPU, so you can set it to 253W on both PL1 and PL2 and ICC=307A to run stable. Here's my settings: E-Core Turbo Ratio offset = +1 CPU Ratio Offset when Running AVX = -3 CPU Lite Load = Mode 1 (located in Advance CPU Configuration Settings) CPU Loadline Calibration Control = 5 (located in DigitALL Power Settings) CPU Core Voltage offset = 0.100 PL1 = 253W Duration=56s PL2 = 253W ICC = 307A Hope it helps anyone out there with the same MOBO and CPU. My RAM is a Corsair DDR5 RAM with 7200Mhz. So everything far runs stable. Ran OCCT for stability for 1 hour passed, with Cinebench Multi-core score is 39200, and Single Core of 2k+ I did ran LatencyReport as well and did not return any major issues.
For dummies that cannot undervolt; or bought a Corsair 4000d and the CFM intake from the front panel is too restricted for 360 radiator flow? Or the people that bought a B660/B760 as a motherboard for an upgrade path? Most people claiming this, has the x3d bug, and really want Intel to give them money back. A 7950x has the same issues with a B650. You tubers just loving clumping User error issues with something that possibly could be a default bios problem, with people that bought a BUNCH of parts.... Processors burning up, yeah voltage. System crashes, usually a lack of voltage or a driver error. Program crashes and blackouts dropping threads, improper undervoltage as well. For myself I need a Class Action, I can use Auto Undervolt in Adrenaline, and it has never successfully kept an undervolt for more than 2 hours without crashing the Adrenaline program. So when glitches have occured in the last 1.3 years of my 13700K's life, it is had Adrenaline crash behind it back to default every time.... So doubtful every issue in 2024 is Intel..... Remember Steve at GN has beat his brains out to get an overclock on every Zen released to date... Usually giving up, he is no longer on the LN2 bandwagon....
The end user should not have to go and setup their BIOS to fix issues caused by the motherboard manufacturer. Like you said, most people will just go in, enable XMP, and call it a day. Why would they even think that an overclock - that they themselves didn’t apply - would be an issue? To make this problem worse is that we, as tech enthusiasts, have been saying for the last few years that OCing isn’t as beneficial anymore as most CPUs will OC themselves. Motherboard manufacturers should have their defaults set to whatever Intel or AMD specify and nothing else. Let users who want to OC do so at their own risk - especially so given that overclocking is not generally covered under CPU warranties.
Not once but twice I’ve gotten unlucky with my i9 14900kf cpu. Blows my mind others are having the same issues I’m having. Especially the unreal engine oodle error. Crazy.
I find it interesting that my manually tuned 13900KS to a 24/7 oc of 5.6ghz/4.5ghz e core in an ITX build has never experienced any of this. This could be a ram issue for some as you highlighted DDR5 is insanely annoying and not ready for the average end user but more so hardcore overclockers. I believe this is 100% a default bios issue or XMP issue. I tested my ASUS Z790I out of the box and it defaulted to 1.5v on the cpu core with vdroop. My cpu requires 1.3V to manually hit 5.6/4.5. My system is almost a year old with zero issues mentioned. Also I think it needs to be very clear that this is board partners releasing these terribly optimized voltages and not intel.
I have the 13900KS After 1 Year of Usage i became all these Issues😢😢😢😢 Even Faceit Anticheat dont work and crashes wit Bsod - KMODE_EXCEPTION_NOT_HANDLED Wolfenstein Youngblood stuttering very much in dark places in game I disabled the Multicore Enhancement but the issues are still there😢 I can't enjoy my Pc since 2 Months😮
Not undervolting so much is one problem. But you throw ANY 360 AIO into a Corsair4000D and that case will raise your idle temps 10C with all the same settings, as in a 2015 Newegg budget Full tower... I had to take a 9900Kf down to 4.3 to get back to back runs of R23 without thermal throttling, when it would run 5.2 in a cheap case with the same 360AIO. B660 and B760's being sold as upgrade paths is another thing causing the program dropouts. Half the bios is locked for the K chips, in bios. Some of those settings go hand in hand with stability. Or even the ability to use XTU effectively. Cache voltage and speed come into play with the K cpus. I'm glad people are getting a new chip, but doubtful their chips will "Make it another day" same user....
I tested with both ddr4 and ddr5, also different mobo, it's intel, I have a 13900kf, worked fine for a couple of months, after that I have been getting constant BSODs, crashes and freezes, I had to downclock to x52 just to get it stable. amazing video, all the errors you spoke about, I had them
If the pc is at defaults... crashing during such things like heavy DX12 shader compilation, out of video memory etc.. all of the bsod's i see people post are clock watchdogs or whea errors... under such situation, this is literally all just due to 100c utilization because the stock cpus/motherboards are running out with ridiculous auto AC_LL values and vcore during load aka heat induced bsod's...
I have my i9 13900K since january 2023.I start having crashes in the beginning of 2024.Most of the games were crashing,but all the stress tests and benchmarks were running just fine.I just disable MCE and for now it's better.I pray that it will hold on for few years more,but after it will start to make problems again,i'm going back to AMD.Maybe i will try to RMA before that,but where i'm from,RMA is very difficult to solve.
after 4 months later my I9 13900k became unstable, crashing from games and after effect renders and later i solved it by changing my asus proart z790 motherboard power settings close to intel power settings and now it is stable: pl1 253w pl2 253w 340amps
My early production 13900K was experiencing issues with stuttering, crashes, large power draws at idle and, the most annoying issue is the USB ports randomly drop and reconnect. I tried 3 different motherboards, 4 different RAM brands, 3 different power supplies, BIOS revisions and settings, etc. I went back and forth with Intel and they agreed to RMA the CPU. I updated the motherboard BIOS, applied the latest ME firmware and recommended BIOS settings from Intel and Initially, the problem seemed like it was solved. Now, the USB connection issues are back and some slight stuttering is happening again. On top of all of this, the motherboard onboard 2.5Gb NIC's have always had disconnect issues (which I solved by installing a 10Gb Marvel NIC) and with all of these issues compounding, I'm done with Intel. I'll be switching to AMD when the 9950X3D comes out next month.
I am running a 12900ks and happy to say, I'm having none of these issues, it is happy to pull 320w+ with no issues, and gives me 5.2ghz ''all core'' (5.2ghz 8xPcore/4ghz 8xEcore) in game or stress test, and 5.5 single core no problems on the 2 cores that reach it lol(4+5), well except temps but, an under volt helps, and come on its a 12900ks able pull over 300w, its going to get hot lol, it also has to run 2 GPUs as I have an ARC card dedicated for OBS(streaming and recording) and an RTX for Gaming. The only ''very rare'' crashing I have, is probably more because I daily drive my RTX card with an OC of +20% on the die and +15% on the Vram lol.
I purchased a 13600kf last year after seeing the price to performance. I manually undervolted and clocked my CPU as follows. Amps - 250, Voltage short term - 160 and voltage long term - 160. I did this after reading an article claiming that based on a calculation, I cant remember the exact but, its like amp x volts = power. Feel free to correct. In it he was speaking with Intel. He was claiming that a 13600kf should only need 175 or 200 amps. It may have even been in one of Intel's specification sheets. After reading that I tried using 175 and then 200 Amps in my bios. ASUS. 175 and 200 would cause my CPU to throttle even at the rated 181 volts. Without going through all the different scenarios I tried I will say that based on temps and performance. The 250 amps and 160 volts short and long term resulted in a Cinebench score of 24250. With the CPU staying at around 70 Celsius.. I would randomly check the Cinebench scores after bios changes. Last month while checking the score dropped to 22200. Now is a 2000 drop acceptable for 11 months of use? I then tried to turn up amps and volts, overclock etc. This time the pc was crashing. I also tried turning up my ram speed. I was able to overclock 6400 to 6800 no problem. Now not even 6600. It is causing issues now, that it didn't before. Everything I just said is a long winded way of saying. THESE CPUS ARE DEGRADING. I am unable to set up my pc as I once did and recreate the same results.
Im glad this issue is coming to light. A few weeks after 14th gen cpu came out, i had a 14900k and it was having all the same issues that is spoke about. I ended up returning it and buying a 13700k. It’s been running stable on stock settings compared to the 14900k.
I have ran into game crashes on helldivers 2 with Ryzen 7700X with stable overclock and 64GB memory. Can confirm there is a stability issue in the game itself
I have a 13700k. I was running an all P core 5.5 oc all E cores parked for two weeks with no issues. Then one day I get constant game crashes half way through ranked COD matches. It was a little frustrating to say the least. I had to downclock the cores to 5.2 for stability. I was not running crazy voltage 1.36-1.4
Intels "performance+efficiency cores" is such bullshit... Imagine buying a car advertised as having 400 power, but only 50 of them are horsepowers and the rest are duckpower.
My Asus Gaming Wifi Z690 killed my 13900k in a little over a year. I had Asus MC Enhancement off, cooled by a 360 mm AIO, and temps were never crazy or above normal. A week ago, I shut down my computer and it never fired back up again, with a red CPU light on. I got the Z790 Gaming Wifi II and 64 GB of RAM now, and the 13900K refused to work on my brand new motherboard. Bought the 14700k, and the system fired right up without issue. I am still very upset that a one year old, $600 CPU is dead and I was not overclocking it, kept it always cool and dusted off, in good cases.
Wish I’d seen all these issues 3 weeks ago when I built my system, been having so many issues that I think is the 14900k. Switching it in bios to to tower cooler mode (was on water cooling by default for some reason) seems to have solved them 😅
Looks like i made a good choice last year when i built my first all AMD system. The only "issue" i have had was the long boot time, but that got fixed after a few BIOS updates. Other than that my PC has been working flawlessly.
I just came yesterday from work... And started to experience crashes on callipsto protocol, battlefield v and 2042...sniper elite 5 too... I feel SCAMED 👺
good marketing strategy, release CPUs heavily overclocked from the factory, so you get the good looking score numbers at the release, get more hype, more people buying it, after few months those CPUs are getting unstable from running so hard so you gotta lower the clocks, loosing performance lmao
Which is a real bummer because like many others I expect my systems to run a very long time. After they are no longer suitable for gaming or other such intensive tasks I often retire them to "light duty" or give them away to the old folks in my family. Either way, I am in the very process of RMA'ing a 14900K right now. I must say, so far Intel has treated me very well. Just as mentioned in this video, my 14900K was really snappy but after like 2.5 months I started getting BSOD. My main rig is sitting here next to me as my 14900K is en route to Kentucky for a replacement. Relevant system info for those interested: Proc: 14900K Mobo: Asus ROG STRIX Z790-E Gaming Wi-Fi Cooler: EK Nucleus CR360 Dark AIO. Ambient hovers around 22°. - Purchased the Proc new, boxed from Newegg in early Feb of 2024. - Initial testing and benchmarking using the usual software was good. Thermals seemed average for the chip considering my setup. - Knowing Asus MCE is the devil, I disabled on initial bios setup. - No manual overclocking - Undervolting performed via LLC (My rig is not up now so my numbers may not be exact, but I believe I was at AC 0.49 and DC 1.12) - BIOS Version: Problems started with Version 2002. Problems persisted with Version 2102. - Problems persisted with everything stock, MCE settings made no difference. ---> Now here's when I proved it was the 14900K and not my other hardware/settings. My other rig has an Asus TUF z790 Plus Wifi running a 13700K. System is 24/7 stable no matter what you throw at it and has been since last July when I built it. I simply swapped the procs between my two systems and the very same issues I experienced with the 14900K on the ROG STRIX system presented themselves on the TUF system. After the swap, the ROG Strix system ran the 13700K without a hitch. In short, the problem was the 14900K, sure as the summer rain. Explaining this to Intel is when they approved my RMA with ease. 👉 IMPORTANT NOTE TO READERS: AN UPDATED BIOS (Version 2202) HAS BEEN RELEASED, DATED APRIL 19TH. I CANNOT VOUCH FOR THE BIOS, BUT IT SEEKS TO ADDRESS THE VERY ISSUE MENTIONED IN THIS AND SO MANY OTHER VIDEOS. COPYPASTED FROM ASUS SUPPORT FOR THE MOBO MENTIONED ABOVE : "The update introduces the Intel Baseline Profile option, allowing users to revert to Intel factory default settings for basic functionality, lower power limits, and improving stability in certain games." 🙋♀UPDATE YOUR BIOS, FOLKS🙋♀ Oh, final observation. I know SP score can be a bit iffy, but, for what it's worth, my particular proc was pretty much trash tier out the box with a SP of 90.
@@COMMANDandConquer199 and who approved that? INTEL. If these are MB issues it should have happened a long time ago and to every manufacturer, the point is intel gave them the go signal to allow this so they are responsible as well.
@@nielsenrainier7710 No, Intel did not approve those. They power and clock settings are literally being set to unlimited out of the box on these boards. Intel would have to be retarded to approve that.
@@COMMANDandConquer199 wow someone WHO eat shit for real, Intel IS scaming and doing that they are Marketing Ther own CPU with the clocks now people spend 2x the amount of Money for Intel get 2x - 3x the Watt usage and less Performance then on AMD, they diserve that for Sure because they Support the Bad product in all cost. First you Had to buy a CPU frame because Intel fked IT Up now you have to clock IT down b cause Intel fked IT up awesome and the CPU still cost too much and eat too much Power awesome
Disabling hyperthreading increases the 14900K's and 14900KS's power efficiency, performance is moderately less than Hyperthreading enabled (24 Core CPU No HT VS 32 Core CPU With HT), more performance per watt and overclocking headroom with 25% to 50% better thermals.
Me, the main problem that manifests itself is that the mouse movement sometimes freezes, and in other cases it disconnects for a few seconds all USB connection to the computer and it is as if it restarts after a few seconds. I don't know if it is related to the i9 13900K CPU that I have in ASUS Rog Maximus Z790 Hero (latest BIOS 2301 from May 30, 2024) with Intel profiles, without any OC, and RTX Suprim X 3080 Ti, Corsair RM1000i source. Other times, games like Call Of Duty after a few minutes of playing it closes by itself for no reason and with normal temperatures (complete EKW custom liquid cooling of radiators and CPU block) without going over 75 degrees. In previous Bios (eg Bios 2002) the initial compilation of shaders sometimes crashed. With everything that is happening, could it be a CPU problem?
100% yes. Hand it over under warranty before BSODs start appearing endlessly like mine. There are more and more reports that CPU 13700k, 13900k, 14700k, 149000k are dying in large numbers.
I have the same issue with the mouse stutter right after entering my password. Never had any crashes but temps through the roof and I’ve been trying to lower the temps . I did 253 253 307 -.045 and I get decent temps. I turned off turbo boost and only lose about 20-30 dps and the temps are mad low but what ever. Also my power bill is insane
Because mobo makers aren’t locking down the specs out of a desire to get high bench scores. It’s not a CPU thing it’s an industry thing around overclocking hardware.
This last pc I built, I actually went Intel for the first time. Been using amd since the late 90s. No real issues with amd specifically but trying something new, and they were a better bang for buck for the money. Makes me glad at the time that I went with the i5 12600kf instead of with a newer generation and went with ddr4. Intel really needs to fix this pronto. Might be going back to amd down the road as I rarely ever had issues. This time it mainly came down to total platform price and the fact Intel is an American company that had a good reputation.
@@drunkdriverat the time I was looking at the budget I think. I’d gotten the 12600kf a $140 on sale and a z690 on sale for 90 something plus able to use cheaper ddr4 ram. A bit of savings on the platform cost vs am5 at the time. Not totally unhappy with the Intel build but am still a fan as well. I did also purchase a 6800xt to go in that new Intel box as well:).
Been having 4 seconds freeze in CS2 for the past two months, changing the bios setting to intel guideline did not help. I am changing the performance core ratio from 55 to 50 now, and lets see if the issue still occurs. Fingers crossed.
Nice Vid! Have you noticed the increase of voltage in the last apex BIOS version that you mentioned? Im one of those who don’t want to run XMP because the SPD writing don’t consider voltages for the CPU side. So, even an external controller must have the right voltage. So, board parters set SA in auto and even worse, CPUVDDQ2 at very high levels like 1.35, which is wrong. You don’t want what a 8200/8400 run XMP at 1.4v VDD/VDDQ. There’s no understanding or knowledge in the community about DDR5 and how it trains and work. Many Techtubers not only do not know, but they misinform the community, such as Buildzoid. I'm sorry he got into this topic. Instead of reading data sheets, he could have learned with Futurplus or also with some BIOSdev, but he does not manage relationships in his timings, he only works with (low) registers. It is not easy, but it is possible to reduce the voltages (lower) than an XMP profile, and not have stability problems. Finally, many users use fixed voltages with LLC6, limiting the core clocks, without allowing droop to exist, but for light loads they have excess current. The CPU must have its boost capacity, and depending on the quality of the silicon, limit the PLs, ICCMAX, IVR to avoid degradation.
My gosh, you reach those framerates in Hogsmead? I think you just destroyed a 7800X3D at stock, what a beast of a chip, easily more than 10 years of a lifetime for the CPU.
This might be a dumb question but I don’t notice anything like this on the i5 13600k cpu is it mainly i9 and some i7. As I’m listening I have done a neg offset of 73 on the cpu is that maybe why it’s not happening to me? Highest clock I allow it to hit is 5.1 I don’t OC to anything higher running ddr5 6400 cl32.
there's definitely a latency issue with those intel chips, i upgraded from an old i3 and i saw no difference in the usability, it feels just as slow, the only real difference is in games and when editing, but in usability is the same or slower.
Question, does this affect the Non K models as well? Cause what I've been reading it's all for K's. Apprently my 13700 does not exhibit this behavior for now.
Any reports on the gen12 ones having issues? I've got a gen12 12900KS which seems fine, but the 14900K I just got, for sure had some issues. I pulled it put my KS back in today. I needed to swap my MB anyways, so just bought a new cpu to go with it. I like my KS.
I have a 14900KS, and have had much issues. So far only some games are having really bad issues with the high clock rate. So to fix this I put in intel factory settings for power and a down clock to 5.3 on my performance cores. Seems to have done the trick and temperatures are below 70c at all time. Also doesn't affect my performances in gaming that much. The memory issue was stopped by turning on re-size bar. G-sync was also crashing some games so I turned it off in the global settings. I was also changing XMP profile to lower than my rated 6000. I ended up removing a 32gb stick, in case it was creating instability. The latest Nvidia driver also created crashes so I reverted back to the one before.
@@jjlw2378 I have 20 days left to return the PC and get a new one. I'm waiting to see if a new BIOS update supporting the KS is getting released for my MOBO. If not I'll ship it back.
There was no awareness about this. But now it’s hopefully well known. I couldn’t believe it messed up 3 different times for me and no one on the internet related with me until i tested different clockspeeds, ddr4 ram, and voltages. Im 5.2 and undervolt using Lite load modification. And im ok on 14700开was on 2 separate 13900ks before i did this
Don't buy the 14900k, it's garbage. I have a big problem with this CPU that you can't imagine and I bought it on July 10, 2024. It is quite difficult to deal with the temperatures, lowering it to 100w gives good performance but guess what blue screen when playing.
I've been watching this story with interest, I run a 13600kf but it's been undervolted since day 1, I do a lot of sim racing and get extremely consistent frame rates and timings, I've certainly never noticed frame drops or stuttering and I haven't had any crashes, it is a very stable system at either 5.5GHz all P cores (most used OC) 5.8GHz all P cores or even 6.0GHz on two P cores and 5.8GHz on the rest. I think Intel or mobo manufacturers were a little optimistic with the voltages.
I've had freezes lately but its related to the trash IME drivers from Intel. I've had freezes from 1 second to 20 seconds after logon. I've installed different IME drivers and it highly affects the stutters. Downclocking or adjusting voltage doesnt affect the stutters, but sometimes it does because its a bit random how the IME drivers loads at logon I think. Pch chip temp is also a problem, they often have to small heatsinks creating dropouts on usb etc. The undervolt the pch voltage to get away with a too small heatsink. When the pch chip dumps all the heat into the mb the ram signals is also affected. The heat affects the pcb traces increasing the resistance so your ram oc is no longer stable. Try to keep the Pch chip under 65C.
I purchased my 14700k in October of 2023. I don’t run it overclocked and started experiencing crashes and blue screens about 2 months ago in productivity applications. I’m currently in the process of getting it RMA’d. Nothing I do in the bios fixes it, the ram wasn’t the issue, however I haven’t tried underclocking, under volting or overvolting. The CPU should not be crashing at its base spec. It would crash even with the default PL1 and PL2 implemented by Intel. I’m very disappointed..
Heat is the enemy of CPUs/electronics. If you can properly cool it, run that clock speed. Power limits and all else aside, if you can cool your CPU to 85C or less under an all-core load at a given clock speed, run it at that clock speed. Use a VF curve offset not a static core voltage. These CPUs need to run as they were designed to. Just not at TJMax.
Got a 14700k, for a few months it worked fine with extreme over clocking on prime z790p. Recently games would crash, but not the OS. Now I can only run on stock settings.. using a kracken 240, push pull fan setup, also had the overclock limiter set to 90c. My only guess is that the cpu fried.. also it passes cinebench just fine. Running 1000w psu... I just don't get it 😕
So Intel has pushed the 1945 von Newmann architecture beyond its limits. Maybe it's time to look at a modern (1960s) architecture like the direct execution stack matching (like the GA-144).
Very strange behavior of the processor, interesting that most of them work without problems for 2-3 months, then the problems start, that is, something happens, it changes within this period of 3 months.
@@Michal_BauerMy i7-12700K was running at 1.42 volts at auto settings on my MSI motherboard. On Cinebench, it reached instant 240 watts and 90 degrees. I did a -110mv adaptative undervolt through XTU, now it's 67 degrees max and 158 watts Same perfomance scores.
I have a z790 14900k. Asus put out a mobo bios update the other day with an "intel baseline" setting. Did that then set xmp and AI overclocking. Still hits 1980 in Cinebench 24 (was getting 2170ish with MCE enabled). Xmp at 6400mt . Temps don't hit peaks higher than 87 and sit around 74c in cinebench. Boosts 2 cores to 6200 then sits at around 4800-5000 all core to get that score. Still much faster than my old 12700k, but clearly not as good as the mobo settings made people think.
I am glad i researched my 14900k though kinda regett getting it but to late. Anyways i found out to fix the bios fir heat snd stability. It takes a cenibench multicore test just fine.
Could this be why the first descendant game shits itself with high pitched audio errors and constant crashing, while it was somewhat more stable on my 8 year old pc?
my PC just crashed playing on the Division 2 was fast traveling and froze, shut down, went to Bios, and would not want to come out of bios. Had to power down and came back on to windows. Never happened B4, have a i7-12700K - MSI MAG Z690 Tomahawk be quiet - Dark Rock Pro 5 Quiet - WiFi DDR4 - 32GB Ram - RM750x PSU - RX7800XT This i built it Dec 4, 2023, just today 4/22/2024 and yes the studdering on some games, looping, Buffering, Loading, on browsers, youtube, i thought was very strange... i Have not overclocked, only XMP.
Nice video! At 11:02 yeah about that part not all motherboards of the same brand (cough Gigabyte cough*) runs the same way following a tutorial, for example the Aorus Z390 Pro Wifi which I own won't get along with the Aorus Z390 Master and another thing the bios upgrade always changes the menu and ways the motherboard will funtion, it could be a downgrade or upgrade of the modes from their list, the point is that after following the tutorial it won't start loading windows so by the time someone with knowledge has the guts to make a Z390 pro wifi version then I'll take the step to OC my rig. For now I'll keep it at the original stock setting and not waste my money down the drain! Cheers!
I got a headache because of 14900K CPU Unstable, after restarting and restarting with different BSOD, at first I thought was because of RAM, SSD, or my 4090. But everything is fine
I dont have any issue i have a 13900k everybody needs to change in there bios add the amps at 307 for the cpu and 253 watts for the wattage and u should not have any problems
I still get paranoid about the AM5 burning cpus too, it's not just Intel. I have my own 7800X3D with a tiny bit less soc voltage than recommended just in case, lel
I have an i9 13900KF, z790 MB. I get the out of video memory. In the BIOS, I have to select "Sync all cores" for both E and P cores. Any other setting and I get a hard freeze when idle at the desktop. This makes the CPU run about 10 deg hotter, at 46deg cel. For some games like Hogwarts, I have to quickly set the affinity to run on 4 cores only. If I dont then Hogwarts crashes when it begins to download the shaders. Everything is stable now, just have the slightly hotter idle temps and having to set affinity when launching some games.
It is ridiculous that normal users should still be forced to go into a thing like the BIOS and change things, or that MB manufacturers can make changes. Look at Apple's M-chips: They use way less power and are so much more efficient and stable. Intel pushing the very limits of their architecture?
I knew these high ass temperatures was going to be problematic. Egos at play here. Silicon quality isn't cutting it. As far as the power limits that shouldn't be a problem the Motherboard would allow those limits, The CPU could never draw that kind of power. The Silicon simply sucks for the clock speeds. Reserved for the premium KS.
The thing is what i am gonna do with the CPU if I get a new I9 13k it will get broken in 2 month so I cant do that. Get a Lower CPU? Well I payed for a high end CPU and still want one in my rig. And to buy a AMD CPU would mean also to change the Motherboard... What a Dilemma
I have an ASUS Z-690 E WIFI motherboard with an i9 13900 K CPU and have had no problems so far. I am running stock settings and have run some intensive benchmarks with no problems so far. Maybe I'm just lucky?
As weird as this sounds, ever since I built my 13900k z790 ddr4 4090fe, I feel like my gaming experience feels bad. Feels slow and almost laggy. Im gaming 240fps @ 1440p and it feels so weird. I'm not sure why. Maybe this will help
Try disabling hyperthreading in the bios. I disabled it because I noticed in certain games there was weird lagging like every 10 or so seconds it would stutter and disabling the hyperthreading fixed it and it is butter smooth now. This was with a i7-12700k.
display control logic in the middle e cores also in the middle inter connect pass for fabric so technically everything can talk to everything , quad channel memory boom even IGPU's will work good
I just Built a PC with i9-14900k and am getting blue screen. Fixed the CPU problem by manually setting Amperage an max Watts draw in MB. But never in my life dis i know the CPU supports a max Ram frequency like you put in table 8:53. So my CPU supports max 5600, my mother board max 7200, and i purchased 7000. All in all usless ram and motherboard frequency max when the CPU does not support it.
