Swap 13700k for 7800X3D

[Bobaganoosh]

You can try to turn off the Asus enhancement again, but I don't think you'll see enough improvement unless you improve your cooler situation. I think there's either something wrong with the paste/mount, or the cooler itself might not be adequate. At 253W, any air cooler is going to struggle a bit to keep the temperatures below 100C in Cinebench. Any other load should be fine, but CB is especially hard on CPUs. The only way to really bring that down if you can't improve cooling is to reduce the power, which can be done a number of ways (outlined in all the comment suggestions here, adjusting Asus profiles, power limits, LL, voltage, etc.).

You can buy a cooler upgrade for ~$100 (DRP5 if you want to stay with be quiet!), but if I was spending up to $100, I'd personally go with the $87 280mm Arctic LFIII which is cheaper and outperforms every air cooler out there:

^note this testing is at ~250W sustained with a similar application. The new Dark Rock Elite (~$119, which should perform quite a bit better than the dark rock pro 4) hit 92.4C while the LF3-280 hit 80.1C. GN tested it against the dark rock pro 4 (among others) and it was doing about 10C better at 200W. They didn't test the dark rock pro 4 in their 250W test (which was focused on contact frames).

 

[chrcoluk]

Even when I was running the 3d mark CPU tests that hit my chip harder than games typically will and I was nowhere near 100C. I think something is wrong with your setup if you hitting 100C in games with a respectable air cooler.

Setting VCore manually rather than trusting stock, motherboard or "AI" values has been an option since release.

Experienced builders do this anyway.

FYI you can boost to 5.8 GHz with 1.4 VCore on Raptor Lake so I see no issue "limiting" it to that.

For further reference in this thread, my absolute max vcore on a 13700k is 1.344v, however it only reaches this if I disable all c-states, I did it for science to see how it would peak. (windows lets you kill them on the fly with idle state power schema setting)

With them enabled again vcore is 1.288v.

These are with undervolt applied, but even the VID requests are below 1.4.v.

The highest load on my chip I have seen in a thread heavy gaming type environment is 144W with HTT enabled. Although it is typically way lower and closer to 50W.

Further voltage changes are undervolt for the ring/cache and me overriding the high i/o related voltages ASRock configured for XMP clocks.

 

[Outback Bronze]

now it's even worse

Thats because its max volts went close to 1.5v gees, and this is supposed to be the new microcode bios. OMG..

 

[oxrufiioxo]

Thats because its max volts went close to 1.5v gees, and this is supposed to be the new microcode bios. OMG..

That is just a band aid fix the real fix is suppose to come later from intel.

Edit: It may be out today temps seem the same as older microcode though.

0x129 is the latest microcode.

 

[Zach_01]

Hello Everyone again!

So i tried today's new beta BIOS Microcode update just to see if something would change, and now it's even worse lol it reaches 100º way more faster like in 4/5 seconds, and the only change i did was activate XMP I and set my fans to turbo speed, nothing else, all the rest is in default/intel settings.

View attachment 358008

I would first try to limit VID requests from "IA VR Voltage Limit" (to 1350~1400mV) in BIOS and also shave off at least 50W from power limit until Intel implements its new microcode to new BIOS versions in 1-2weeks from now.

Over 1.3V on full load is a bit stretch while the CPU pulls 250W and be on 100C. 1.2~1.25V is more sane and safe.
Same for single thread and low loads the 1.5V is too much. 1.35~1.4V is better.

Swap 13700k for 7800X3D

Yes lol. he means 50mV, or 0.050V. I did that, and the result was: Cinebench instant crash after 5 seconds lol

www.techpowerup.com

 

[ShrimpBrime]

Hello Everyone again!

So i tried today's new beta BIOS Microcode update just to see if something would change, and now it's even worse lol it reaches 100º way more faster like in 4/5 seconds, and the only change i did was activate XMP I and set my fans to turbo speed, nothing else, all the rest is in default/intel settings.

View attachment 358008

Dude. Guy. Bro.

Again.

Try.

Go to bios.
Set all multiplier "fixed" 40x
All P and E cores.

Hit F10 save restart.

Boot windows.

Run benchmark and show temps.

Please expand HWMonitor so we can see working v-core.

Please take the screen shot in the near end Cinebench, not afterwards.

No idea why you updated bios and thought the cpu would magically be less than 253w capable. This is not how it works. Your cooler is not adequate.

That's all I got for now. If 4ghz runs cool, up the multi to 42x. Do this until you peak 100c during benchmarks.

I've got nothing more to help here. I just hope this eventually works out for you. Be well and good luck.

 

[Bill_Bright]

Assuming the cooler's fan is ramping up to speed as it should, the cooler is fine. But just to verify and appease the naysayers, try removing the case side panel and blast, on high, a desk fan in there. If you still have problems, it is not your CPU cooler. If you no longer have problems, more experimenting is required as it could be your CPU cooler or it could be case cooling, or it could be something else like a corrupt setting or faulty component.

Sure would be nice to try that CPU on a different motherboard.

What are your PSU voltages? You can see them on the same HWiNFO64 screen you used above - just scroll down a little further to your motherboard's section and note the +3.3V, +12V and +5V voltages. Note these sensors are very cheap, low-tech devices so going by sensors is NOT the best, most accurate method for measuring these voltages. But for initial checks, they are good enough. If you see an out-of-tolerance reading, more conclusive PSU testing, or swapping in a known good PSU may be required.

The ATX 2.x Form Factor standard for PSUs requires PSUs maintain output voltage within ±5% of the required specifications. This means for ATX 2.x PSUs, the allowed tolerances are:

12VDC ±5% = 11.4 to 12.6VDC
5VDC ±5% = 4.75 to 5.25VDC
3.3VDC ±5% = 3.14 to 3.47VDC
​

For ATX 3.x PSUs, the tolerance for +12VDC was relaxed a little and is now -7% up to +5% as follows:

12VDC -7%, +5% = 11.16 to 12.6VAC​

 

[Laurijan]

Intel microcode update have started to roll out and mobo makers are putting updates online. Might help to get those temps down due to microcode lowering excessive voltage bumps.

 

[izy]

Seems that some bios updates are up.

 

[Laurijan]

Seems that some bios updates are up.

View attachment 358204

also minimal perf hits and in some cases low gains

 

[Laurijan]

just bear in mind that with the microcode update you still can get a unstable system down the line if its degraded enough. that why intel extended warranty to 5 years in these cases