My guess would be most customers would not know how to manually setup a cpu in the bios
I don't even know what many of the cpu settings in the bios are myself
Last edited:
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Asus issues BIOS update to solve Intel CPU game crashing.
- Thread starter humbug
- Start date
My guess would be most customers would not know how to manually setup a cpu in the bios
I don't even know what many of the cpu settings in the bios are myself
Soldato
As long as I figured out how to run my 14700k with a undervolt and set the power draw levels to Intel spec, did the same when I had the 12700k, that seems to be as far as I go in the BIOS, oh and XMP for the memory.
There are so many settings, sub menus and and the like, I have not got a clue as to what most of it means.
At one point I watched a few of "skatterbencher" videos, including using subs on occasion. I recognised that he was talking English, but it might as well as been gibberish for what it meant to me.
Then again at the time of watching those I had some thoughts that overclocking might still be relevant for my use, thankfully that seems to not be the case. Altho Asus, in particular, seem to have offered a pre-bake default option if you buy one of their boards.
Last edited:
My 14900k doesn’t like being undervolted,
I have left the PL1 on the Asus default of 253watts and I changed the PL2 from unlimited to 295watts to stop the cpu from hitting 100c with the 330watts or whatever it was drawing when it had the PL2 set to Asus unlimited default setting
I also left the PL2 default asus boost time at 96seconds
This is with the older bios and not the new bios with the changed default safe limits
Last edited:
Soldato
I was surprised to see what a 0.07v undervolt meant in terms of vcore and watts used etc....
Thankfully, no noticed stability issues.
Even tho Gigabytes BIOS had unlimited values for the power limits, iirc, it was only a few games that used the CPU enough where that might have made a difference by limiting it.
Running at 4k with a 4080 it is not usual to be CPU limited.
At least you have been able to make some difference to try and limit the temps
Soldato
I had to do most of this stuff just to keep temps reasonable with the D15S I wanted to run on the 13900K. I already did most of the work, so I haven't bothered with the newer BIOS versions.
The only thing I was missing was the AC LLC = DC LLC. I have now fixed both of those at 0.4.
Operating voltage went up a hair, (-0.010 offset) but sill tops out in the mid 1.3xx range and droops to low 1.2xx / high 1.1vxx under heavy all-core load.
I capped the multiplier at 55 since it never seemed to do any real work at the 58 peak anyway.
Oh, and my Asus Hero board defaulted to "MAX ALL THE THINGS", so my D15S had no chance at default settings lol.
The only thing I was missing was the AC LLC = DC LLC. I have now fixed both of those at 0.4.
Operating voltage went up a hair, (-0.010 offset) but sill tops out in the mid 1.3xx range and droops to low 1.2xx / high 1.1vxx under heavy all-core load.
I capped the multiplier at 55 since it never seemed to do any real work at the 58 peak anyway.
Oh, and my Asus Hero board defaulted to "MAX ALL THE THINGS", so my D15S had no chance at default settings lol.
Last edited:
Associate
- Joined
- 5 Jul 2007
- Posts
- 506
Updated my Asus bios yesterday and applied the Intel baseline. Ram was underclocked to something like 2400 so had to increase back to 3600..... Ran a game for 5 minutes and pop, my old issue of games randomly alt-tab'ing back to the desktop. Never seen any Nvidia video memory errors. Yes Windows 10, all updates applied, latest nvidia driver and no issues in device manager.... Luckily I can game for a few hours on a good day before the weird issue rears it's ugly head.
Was the D15s able to keep up after the tuning ?
Soldato
Yes, but the fan does go to 100% (which doesn't sound bad because it's a big Noctua fan) but the fans on the front of the Fractal North case are tied to the CPU temp too, and they are kind of loud when the CPU is under full load, but not enough to make me further tune things.
Max temp with Topaz Ai rendering 3 videos (100% CPU usage) was 91C in a 24C room.
Associate
I wish I'd stumbled across this and similar information posted elsewhere earlier. It would have saved me a lot of time and money.
This time last year I put together a new PC comprising of an Asus ROG Strix Z790-A Gaming/i9-13900K/Noctua NH-D9L combination. To cut to the chase, Maya and any other software that put a load on the PC caused it to crash to the desktop.
As eveything was running at stock it didn't occur to me to check temperatures and I spent months chasing error messages. I swapped the RAM out, changed the graphics card, changed BIOS settings before giving up and buying an iMac.
Roll on a few months and I start to see theads like this one so I updated the BIOS, added a couple more case fans which are all running at 100%, reseated the cooler, underclocked the CPU (it was never overclocked) and loaded up a game. This is the result of running World of Warcraft at the lowest graphics settings.
[EDIT] It appears I can't insert an image. The image showed 24 cores running at 97 deg to 100 deg.
Out of curiosity today I swapped out the processor for an i3-12100F 3.30GHz, reset the motherboard to its default settings and this is the result running the same game at the highest graphics settings.
[EDIT] Again no image. The image showed 4 cores runnin at 65 to 66 deg.
All the software that I'm still licenced to run on a PC no longer crashes to the desktop and this low end processor allows graphics based software to run at much higher settings. I have RMA'd the processor but based on the last year's experience I suspect whatever comes back will be binned. It's a lesson learned, I should have checked the core temperatures which would have saved me a huge amount in wasted time and lost income. In my view there should have been a recall because I do not believe these processors are fit for purpose.
This time last year I put together a new PC comprising of an Asus ROG Strix Z790-A Gaming/i9-13900K/Noctua NH-D9L combination. To cut to the chase, Maya and any other software that put a load on the PC caused it to crash to the desktop.
As eveything was running at stock it didn't occur to me to check temperatures and I spent months chasing error messages. I swapped the RAM out, changed the graphics card, changed BIOS settings before giving up and buying an iMac.
Roll on a few months and I start to see theads like this one so I updated the BIOS, added a couple more case fans which are all running at 100%, reseated the cooler, underclocked the CPU (it was never overclocked) and loaded up a game. This is the result of running World of Warcraft at the lowest graphics settings.
[EDIT] It appears I can't insert an image. The image showed 24 cores running at 97 deg to 100 deg.
Out of curiosity today I swapped out the processor for an i3-12100F 3.30GHz, reset the motherboard to its default settings and this is the result running the same game at the highest graphics settings.
[EDIT] Again no image. The image showed 4 cores runnin at 65 to 66 deg.
All the software that I'm still licenced to run on a PC no longer crashes to the desktop and this low end processor allows graphics based software to run at much higher settings. I have RMA'd the processor but based on the last year's experience I suspect whatever comes back will be binned. It's a lesson learned, I should have checked the core temperatures which would have saved me a huge amount in wasted time and lost income. In my view there should have been a recall because I do not believe these processors are fit for purpose.
Last edited:
@zappaDPJ You sure you can run a 13900K @253w stock and above with that cooler without limiting the power lower than stock as it sounds like thermal issues where kneecapping the performance of the 13900K and causing instabilty once you get that under control if its still happening then look at the possibility of RMA'ing it.
Yeah, there's pretty much zero chance you can run a 13900K with a Noctua NH-D9L in all-core workloads and not experience high temps. It would need to be power limited.
As for World of Warcraft, I don't know, it would depend how much power it is using, but a decent air cooler is usually sufficient. However, running low settings pushes the bottleneck to the CPU and is likely to increase temps and power usage.
The Intel CPU instability problems (that have been in the news, I mean) are not directly related to high temperatures, so high temperatures alone doesn't mean the CPU is faulty.
Last edited:
Associate
Thanks for your replies.
I spent a couple of days running through various suggested BIOS settings from guides I've found on the Net and no matter what, the majority of cores reach 100 degrees C with minutes, sometimes seconds when put under load so I have RMA'd the processor.
For what it's worth, according to Noctua's compatibility centre, the Noctua NH-D9L is compatible with the i9-13900K processor running 'medium turbo/overclocking headroom' so I would have expected it to be sufficient running the processor under-clocked with all fans at max speed. I did try undervolting but still experienced high temperatures and crashes back to the desktop.
To be honest I'm happy with the i3. My PC is now completely stable and runs everything at a satisfactory pace.
Without any power limits enabled (a lot of boards come that way out of the box) it can hit 400+ watts in multithreaded workloads and even at what is normally considered to be "stock", it can hit 300 odd watts.
A single fan compact cooler (even a Noctua) will only be capable of around 100-150 watts without being either: too hot, or too noisy (or both).
If you want to run long run CPU-intensive workloads (your mention of Maya made me assume so), then without a decent level of power limiting the majority of air coolers are not suitable for these CPUs and certainly not a single fan compact cooler. The highest CPU I'd consider running on a single fan cooler would be a 13600K/14600K.
If you used the dual fan config of the NH-D9L it'd be more capable, but even a peerless assassin/phantom spirit (heavier than the D9L) can't really cool these CPUs effectively if you push them much above 200 watts.
Medium turbo is an admittance the CPU can't be run at stock or close to stock.
That said, even If you have a tiptop high-end cooler, it is still quite possible to instahit 100c with these CPUs, if you run something intensive like Cinebench or Prime.
Not saying you didn't have a duff CPU, just clarifying the reality of working with these CPUs and air coolers.
I had this once and it was a software issue. Unfortunately, I can't remember quite which piece of software it was. Steelseries GG, I think.
Soldato
I installed the 0x129 BIOS on my Z790 Hero today and I like it. I'm running the Intel "extreme" default, but with the same 1.4V VID limit I was using before, but for temps rather than fear of over-voltage.
It comes with lower VID requests, but also a more agressive LLC. So, even with a 1.4 max VID limit, my Vcore is getting much closer to the limit with the new BIOS and the tighter LLC.
This is probably just a coincidence, but the new BIOS also does better with the memory. I can now run the Asus "tweaked" timings. I tried it in the old BIOS and it wouldn't boot. The new BIOS made two passes of memtest before I got impatient and tried rendering some video in Topaz, and that program seems to like the new RAM timings. The rig is now performing better than ever. -probably because of the RAM, but still...
The 1.4 VID hurts in Cinebench R23, (I don't think that benchmark cares about RAM) but I'm still in the mid 35k range.
If I had a bigger case, I might swap the D15S out for a 360mm AIO and remove the VID limit, but I like my Fractal North (non-XL) and the performance is decent as is.
It comes with lower VID requests, but also a more agressive LLC. So, even with a 1.4 max VID limit, my Vcore is getting much closer to the limit with the new BIOS and the tighter LLC.
This is probably just a coincidence, but the new BIOS also does better with the memory. I can now run the Asus "tweaked" timings. I tried it in the old BIOS and it wouldn't boot. The new BIOS made two passes of memtest before I got impatient and tried rendering some video in Topaz, and that program seems to like the new RAM timings. The rig is now performing better than ever. -probably because of the RAM, but still...
The 1.4 VID hurts in Cinebench R23, (I don't think that benchmark cares about RAM) but I'm still in the mid 35k range.
If I had a bigger case, I might swap the D15S out for a 360mm AIO and remove the VID limit, but I like my Fractal North (non-XL) and the performance is decent as is.
Last edited:
Associate
Thanks again for the further advice, it is appreciated. I RMA'd the processor last Friday (received by OCUK yesterday afternoon) and I'm already sitting looking at a new processor less than 24 hours later. That's one quick turn around!
I'll freely admit things have moved on a lot since I last put together a PC over 10 years ago, so there's been a big learning curve for me and I'm certainly going to take onboard the advice given in this thread if and when I do pluck up the courage to pop in this new processor.
Associate
It's usual to deal directly with the supplier of the goods rather than the manufacturer in the case of a fault.
Last edited:
I updated a few days ago. Main thing I noticed was that my system was now stable in all of the scenarios it was previously unstable, so that was nice! But temps were definitely up (Also using Intel extreme, Z790 Dark Hero), vid/vcore up, and scores were a bit down. Around 37k R23 compared to ~38k before (The only difference I saw was in fully multithreaded workloads. No real difference in the main software I use for work or games).
Couldn't resist a bit of tinkering to get the temps down as I still don't have the time to RMA this chip right now, so made some simple adjustments to my ACLL/DCLL and now have vid/vcore/temps all a chunk lower, performance is back up, and it's still stable in all the things it wasn't before. Also enforced a 1.4v IA VR limit. In some ways I wonder if they would even accept an RMA back now given it runs under 'Intel spec', but for piece of mind I'd still like to try and swap it for a new one.