Man accidentally buys i9-9900KS, release immenient?

[WereCat]

2 minutes ago, porina said:

If you're saying what I think you're saying, there's always going to be a gap between what a manufacturer wants to use to be stable, vs what an overclocker considers stable. While I've not looked at it in detail, the feeling I get is AMD are more generous on voltage, at least on past GPUs.

yes, that gap is what I have in mind 

[Master Disaster]

8 hours ago, Sauron said:

I don't think you'll need a bios update for this... it's just a 9900K... he might get slightly worse power optimizations (if that) but that's it.

In that case it would show as a 9900K in any software as the BIOS wouldn't know it was a different chip and would report it as a 9900K.

 

It shows as a 9900KS ergo he has access to unrelreased microcode. He's an insider leaking early.

[Kilrah]

That's not how it works. The string you see is just what the processor is reporting. Software doesn't need explicit support.

[Flying Sausages]

WDF?! Over 80C temp on Cinebench R20? It seems like Intel uses shit TIM again or high auto voltage.

[QXC]

12 hours ago, leadeater said:

Better option than both of these is just the 3700X and raise the power limit to make it a 3800X, then apply OC if you want.

Has it been confirmed that the 3700X and 3800X hit similar frequencies when you do this? From what I've read/recall, the 3800X is a much better binned die and generally clocked better than the 3700X, but I don't know for sure if that was with similar power limits.

 

[Sauron]

36 minutes ago, Master Disaster said:

In that case it would show as a 9900K in any software as the BIOS wouldn't know it was a different chip and would report it as a 9900K.

Not really, the name is reported by the cpu itself, not guessed by the motherboard based on the specs.

[leadeater]

47 minutes ago, QXC said:

Has it been confirmed that the 3700X and 3800X hit similar frequencies when you do this? From what I've read/recall, the 3800X is a much better binned die and generally clocked better than the 3700X, but I don't know for sure if that was with similar power limits.

See my other post with the SL bin stats, the difference is minimal and while the 3800X can be better the difference between best 3800X and worst 3700X is barely worth caring about. You can also use more voltage to attain a bit high clock, not higher than the highest possible just if your sample isn't that great and would like to hit 4.2 instead of 4.15 etc.

 

That's why it's not worth SL stocking any of those CPUs at all.

[Derangel]

Why are people surprised by the temps? A bone stock 9900K runs hot with Prime 95. A 9900K overclocked to 5Ghz is even hotter. So pre-overlocked 9900K, even one that is presumably binned, will run hot. There is a reason you want to pair the 9900K with a top end air cooler or a 240mm (or larger) AIO if you're going to try and overclock it.

 

A single Cinebench R20 run pushes the max temp of my stock 9900KF to 75c (ambient temp is 24c) and that's with the cores maxing out at only 4.7Ghz. If I OC it to 5+ it will easily hit the 80s.

[Master Disaster]

3 hours ago, Sauron said:

Not really, the name is reported by the cpu itself, not guessed by the motherboard based on the specs.

CPU name is determined by microcode, that's why new CPUs need bios updates.

[Zando_]

8 minutes ago, Master Disaster said:

CPU name is determined by microcode, that's why new CPUs need bios updates.

Ah, is that the reason why? Are there any other reasons or is it just as simple as the name/BIOS actually recognizing the CPU?

[pas008]

hasnt bios support for the ks been around for awhile?

 

 

https://www.asrock.com/MB/Intel/Z390 Extreme4/index.asp#BIOS

 

[Master Disaster]

22 minutes ago, Zando Bob said:

Ah, is that the reason why? Are there any other reasons or is it just as simple as the name/BIOS actually recognizing the CPU?

It depends, I remember back in the Athlon days it was possible to run an unsupported CPU in a socket A board and the board would literally say "Unknown AMD Athlon CPU".

 

As others have said, software dependencies are sometimes determined by microcode, especially with AGESA but I'm not an expert enough to say exactly what.

 

I do know that without the specific KS microcode string the board wouldn't be able to identify the chip.

[Tech_Dreamer]

Lucky Bastard!

[williamcll]

Further news: Tom's hardware has also got themselves a 9900KS

https://www.tomshardware.com/features/intel-special-edition-core-i9-9900ks-benchmarked