[xeizo]

I'll read it for the 😆 I know Igor will deliver.

For Gaming, that is what AMD said, for workstation the 5900X is faster and now it is a lot cheaper too. The fastest for workstation is 12900K/KF.

 

[tcclaviger]

Lolwut

 

[xeizo]

Lolwut

According to Igors test, that is

5800X3D = great for gaming, efficient at all times
12900K = great for workstation, somewhat power hungry

 

[Kelutrel]

I am a bit sceptic about any advantage that the 12900K may have in games even against a 5900X or a normal 5800X tbh. It may provide 1-2% more fps at 1440p (who plays on a desktop at 720p or 1080p ?) but it uses nearly double the power (probably 20-25% more power at the whole system level). None would notice the difference on that 1% fps, but many may feel the difference in the power bill and thermal requirements. I completely agree with the workstation benchmarks though, but I think that in games AMD always had the upper hand over Intel even without the 5800X3D.

 

[xeizo]

I am a bit sceptic about any advantage that the 12900K may have in games even against a 5900X or a normal 5800X tbh. It may provide 1-2% more fps at 1440p (who plays on a desktop at 720p or 1080p ?) but it uses nearly double the power (probably 20-25% more power at the whole system level). None would notice the difference on that 1% fps, but many may feel the difference in the power bill and thermal requirements. I completely agree with the workstation benchmarks though, but I think that in games AMD always had the upper hand over Intel even without the 5800X3D.

I agree, Intel shines in workstation use rendering/compile etc but it is neither inferior in games. If buying, buy what's priced the best. Intel has one upper hand though, in being able to upgrade to Raptor Lake, Zen 4 requires new motherboard.

 

[J7SC]

...I saw both of Igor's 5800X3 reviews. IMO, in pure gaming only, the 5800X3 seems to be the way to go, but for productivity, gaming and per watt performance, I think the 5950X is the one to go with . I've seen a ton of different workstation app reviews by various testers, and it often comes down to which just productivity apps are included in the 'test-suite'.

The real message though is that with DDR5, faster PCIe4 (and soon PCIe5) storage and GPUs, and faster CPU IPCs, the cache needs to get bigger, a la 5800X3...either Intel Raptor Lake or the subsequent Meteor Lake will add much more cache, and I would assume AMD's Zen4 and up will do the same.

QUICK QUESTION: I finally updated to Bios 3801 (from 3501) and loaded Win 11 Pro last night. Everything works well, but the L3 cache readings are just a touch slower than with Win 10 Pro. I am aware of the early issues of Win 11 and Ryzen, but thought they were fixed by Microsoft via patches (all loaded / up-to-date). Is there anything else by Asus or AMD which is needed as well for Win 11 Pro ? Thanks.

 

[noxious89123]

I get the whea 19 warning not the same thing as whea errors. I didn't see any DPC increase or performance loss.

With that said, I run IF/Mem controller error reporting off/ignore in the bios since my 3900x, not sure if it makes a difference since the error count is frozen on the IF.

If that counter goes up on bugs(whea 19 warnings) it might degrade performance.

When I finish my CO curve, vddg,soc for undervolting 1900IF I will run a good 11h Prime95 blend,large + hci memtest on that 2000IF config and report back.

Hmm, interesting. So should I not be concerned about the WHEA 19 warnings? I would argue that it is still an error, although it is a correctable error, which is probably why Windows only labels it as a warning.

When I'm at Fclk 1900 and memory at 3800 (Uclk 1900MHz), I get roughly 1 WHEA error ever week or two. It's frustrating, because I considered that to be unstable but I've had little luck in finding the precise cause, plus testing any changes takes a long time. I think it might be the memory controller, as when I run 1800/3800 I get no WHEAs (this sets the Uclk 2:1 to 950MHz) and using 1866/3733 (Uclk 1:1 at 1866MHz) is fine. I've also tried 1900/3733 (which has the Uclk at 1866MHz) and this seems to be stable. So I think I've ruled out the instability coming from the Fclk being too high, and instead narrowed it down to the Uclk being too high.

I might just go back to 1900/3800 and start from scratch with my memory OC. Would you all agree that the timings being aggressive could be too much for the IMC, even if the RAM itself is fine?

Yeah, that is true, but some of us peasants can't even boot 1900 with 2 sticks 😭

Have you tried booting 1933? Could be you have the 1900 Fclk "hole". Some people can't boot 1900, but can boot 1933 / 1966 / 2000. Although getting stable above 1900 without WHEAs seems to be rare.

I coan boot Fclk 2000 with very minimal tweaking, but it's also WHEAs out the wazoo...

 

[Kelutrel]

QUICK QUESTION: I finally updated to Bios 3801 (from 3501) and loaded Win 11 Pro last night. Everything works well, but the L3 cache readings are just a touch slower than with Win 10 Pro. I am aware of the early issues of Win 11 and Ryzen, but thought they were fixed by Microsoft via patches (all loaded / up-to-date). Is there anything else by Asus or AMD which is needed as well for Win 11 Pro ? Thanks.

I noticed that VBS has an impact on cache and memory latency, and Win11 enables VBS by default I believe, so that may be why you see a tiny bit slower cache and memory values. Tbh I prefer the added security provided by VBS to those few megabytes per second, but your mileage may vary. Fastest way to disable VBS is to disable "SVM Mode" in the cpu options in the BIOS.

 

[tcclaviger]

Hmm, interesting. So should I not be concerned about the WHEA 19 warnings? I would argue that it is still an error, although it is a correctable error, which is probably why Windows only labels it as a warning.

When I'm at Fclk 1900 and memory at 3800 (Uclk 1900MHz), I get roughly 1 WHEA error ever week or two. It's frustrating, because I considered that to be unstable but I've had little luck in finding the precise cause, plus testing any changes takes a long time. I think it might be the memory controller, as when I run 1800/3800 I get no WHEAs (this sets the Uclk 2:1 to 950MHz) and using 1866/3733 (Uclk 1:1 at 1866MHz) is fine. I've also tried 1900/3733 (which has the Uclk at 1866MHz) and this seems to be stable. So I think I've ruled out the instability coming from the Fclk being too high, and instead narrowed it down to the Uclk being too high.

I might just go back to 1900/3800 and start from scratch with my memory OC. Would you all agree that the timings being aggressive could be too much for the IMC, even if the RAM itself is fine?


Have you tried booting 1933? Could be you have the 1900 Fclk "hole". Some people can't boot 1900, but can boot 1933 / 1966 / 2000. Although getting stable above 1900 without WHEAs seems to be rare.

I coan boot Fclk 2000 with very minimal tweaking, but it's also WHEAs out the wazoo...

On all my Zen 2 and 3s, 1900 bclk strap is most temperamental and easiest to cause instability. Settings that work perfectly at 1866 and 1933 sometimes have problems at 1900.

1 or 2 warning 19, the corrected error warning, I wouldn't even pay attention to.

It's not even the speed, it's the strap. I can go 1866 with raised bclk to 1915 FCLK, totally fine by CPU. Go to 1900 strap with same setting won't post. Go to1933 strap totally fine. For me the refusal to fix or even acknowledge it, is the worst fumble by AMD on AM4 platform.

 

[PWn3R]

Hmm, interesting. So should I not be concerned about the WHEA 19 warnings? I would argue that it is still an error, although it is a correctable error, which is probably why Windows only labels it as a warning.

When I'm at Fclk 1900 and memory at 3800 (Uclk 1900MHz), I get roughly 1 WHEA error ever week or two. It's frustrating, because I considered that to be unstable but I've had little luck in finding the precise cause, plus testing any changes takes a long time. I think it might be the memory controller, as when I run 1800/3800 I get no WHEAs (this sets the Uclk 2:1 to 950MHz) and using 1866/3733 (Uclk 1:1 at 1866MHz) is fine. I've also tried 1900/3733 (which has the Uclk at 1866MHz) and this seems to be stable. So I think I've ruled out the instability coming from the Fclk being too high, and instead narrowed it down to the Uclk being too high.

I might just go back to 1900/3800 and start from scratch with my memory OC. Would you all agree that the timings being aggressive could be too much for the IMC, even if the RAM itself is fine?


Have you tried booting 1933? Could be you have the 1900 Fclk "hole". Some people can't boot 1900, but can boot 1933 / 1966 / 2000. Although getting stable above 1900 without WHEAs seems to be rare.

I coan boot Fclk 2000 with very minimal tweaking, but it's also WHEAs out the wazoo...

Yes I have the hole. Sometimes I can get it to boot 1900 after a 1933. Nothing I tried would get 1933 to not have hundreds of wheas per minute.


Sent from my iPhone using Tapatalk

 

[tcclaviger]

Anyone have a workaround for C8E 1900/3800 C14 bug on 1206b where it refuses to use it? C13 works. C15 works. C14 or C13 with GDM enabled - no post.

This is under test because it's weird and I don't trust it.

 

[tcclaviger]

deleted

 

[TMavica]

Anyone have a workaround for C8E 1900/3800 C14 bug on 1206b where it refuses to use it? C13 works. C15 works. C14 or C13 with GDM enabled - no post.

This is under test because it's weird and I don't trust it.
View attachment 2558309

I am using C8E and 0606 bios, you should able to boot, by adding 56 to AddCmdSetup, and those Clkdrvstr to 40 20 24 24, you should try 1T too

 

[tcclaviger]

So it's working now but the 57:3 strap is a bit of an ass. tRDWR/tWRRD was the culprit. These timings work all the way up to 3941 at 1t.

 

[safedisk]

ROG CROSSHAIR VIII Series 4201 BETA BIOS

1. Update AGESA version to ComboV2PI 1207

ROG CROSSHAIR VIII HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII HERO WIFI BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-WIFI-ASUS-4201.7z

ROG CROSSHAIR VIII FORMULA BETA BIOS 4201
ROG-CROSSHAIR-VIII-FORMULA-ASUS-4201.7z

ROG CROSSHAIR VIII IMPACT BETA BIOS 4201
ROG-CROSSHAIR-VIII-IMPACT-ASUS-4201.7z

ROG CROSSHAIR VIII DARK HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-DARK-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII EXTREME BETA BIOS 0801
ROG-CROSSHAIR-VIII-EXTREME-ASUS-0801.7z

 

[GRABibus]

View attachment 2558416

ROG CROSSHAIR VIII Series 4201 BETA BIOS

1. Update AGESA version to ComboV2PI 1207

ROG CROSSHAIR VIII HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII HERO WIFI BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-WIFI-ASUS-4201.7z

ROG CROSSHAIR VIII FORMULA BETA BIOS 4201
ROG-CROSSHAIR-VIII-FORMULA-ASUS-4201.7z

ROG CROSSHAIR VIII IMPACT BETA BIOS 4201
ROG-CROSSHAIR-VIII-IMPACT-ASUS-4201.7z

ROG CROSSHAIR VIII DARK HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-DARK-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII EXTREME BETA BIOS 0801
ROG-CROSSHAIR-VIII-EXTREME-ASUS-0801.7z

thanks man 👍

 

[Daylight_Invader]

View attachment 2558416

ROG CROSSHAIR VIII Series 4201 BETA BIOS

1. Update AGESA version to ComboV2PI 1207

ROG CROSSHAIR VIII HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII HERO WIFI BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-WIFI-ASUS-4201.7z

ROG CROSSHAIR VIII FORMULA BETA BIOS 4201
ROG-CROSSHAIR-VIII-FORMULA-ASUS-4201.7z

ROG CROSSHAIR VIII IMPACT BETA BIOS 4201
ROG-CROSSHAIR-VIII-IMPACT-ASUS-4201.7z

ROG CROSSHAIR VIII DARK HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-DARK-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII EXTREME BETA BIOS 0801
ROG-CROSSHAIR-VIII-EXTREME-ASUS-0801.7z

Sadly this new Beta does not work at all for me if I boot into DOCP and also use the SOC offset of +0.00625 which I have needed ever since the 3601 BIOS.

 

[enzu4l]

ROG CROSSHAIR VIII Series 4201 BETA BIOS

1. Update AGESA version to ComboV2PI 1207

ROG CROSSHAIR VIII HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII HERO WIFI BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-WIFI-ASUS-4201.7z

ROG CROSSHAIR VIII FORMULA BETA BIOS 4201
ROG-CROSSHAIR-VIII-FORMULA-ASUS-4201.7z

ROG CROSSHAIR VIII IMPACT BETA BIOS 4201
ROG-CROSSHAIR-VIII-IMPACT-ASUS-4201.7z

ROG CROSSHAIR VIII DARK HERO BETA BIOS 4201
ROG-CROSSHAIR-VIII-DARK-HERO-ASUS-4201.7z

ROG CROSSHAIR VIII EXTREME BETA BIOS 0801
ROG-CROSSHAIR-VIII-EXTREME-ASUS-0801.7z

Will it fix the TPM Lag?

 

[GRABibus]

Will it fix the TPM Lag?

Try it 😊

 

[Blackfyre]

View attachment 2558416

ROG CROSSHAIR VIII Series 4201 BETA BIOS

1. Update AGESA version to ComboV2PI 1207
ROG CROSSHAIR VIII HERO WIFI BETA BIOS 4201
ROG-CROSSHAIR-VIII-HERO-WIFI-ASUS-4201.7z

Thank you. Updated and all is working well here so far. Same RAM/CPU overclock from before applied too.

Have not had TPM issue since January when I upgraded to a dTPM instead of continuing with fTPM. But I am sure those who have been waiting will report if this BIOS will finally resolve that issue for them.