ES Xeon Discussion
- Thread starter britinpdx
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
It would be interesting to look at the tests of the Cinebench and Aida 64 CPU Queen
if it work i will post test results
im ordering this from smicro : SYS-221H-TNR | 2U | SuperServer | Products | Supermicro
I don’t quite understand, you mean, SM sells ES/QS of CPU
Last edited:
look at the CPU-Z screenshot AVID-MEDIA has posted:
its the last digit of the cpuid:
stepping D0 806F3 (step 3)
stepping E0 806F4 (step 4)
Before I bought an QWAT-CPU,from a person from this forum
I send message to SM tech-support.
If you can sell me or at least send me a version of theES/QS CPU as a gift, they answered me, we do not sell qualified samples, especially earlier steppings
no. where do you got this ? he want to buy the server from SM and run ES on it.
checked the link ???
Just keep in mind, that unless ES/QS have same stepping as production version of CPU nobody will be able to guarantee you that it won't have grave bugs or that it would work reliably for any amount of time. E0 and E5 are differs by 5 minor revisions and there were bugs fixed between them, just nobody(?) figured out what they were or those who did have NDA signed with some one that prevents them from talking about that (in most big tech companies you probably would have people who seen the list of fixed bugs)
nobody expects early engineering samples to be free of bugs,while prod.units still have some.
If no one tries, no one can know.
those who buy such cannot afford a production unit.
Yeah, I probably misunderstood what he meant. Sorry.
But basically yeah, I agree with what you said, just still worth reminding so people shouldn't try to use it in production or if they are - they should prepare for all kind of problems. Also I still wonder if people already found some behavioral difference between D0, E0, E3 and E5 steppings?
understand, such thing's like 1044 days AMD EPYC timer bug LOL. sorry, couldn't resist.
Your. Or like Avoton's AVR54 bug. Just with early ES - more than usual, probably
P.S. I also should get my Gigabyte MS73 and 8490 ESs next week, and we'll see if NH-U12S would fit it
remember TSX bug in prod. Xeon E5 Haswell-EP ?Your. Or like Avoton's AVR54 bug. Just with early ES - more than usual, probably
desktop don't support TSX thought. used and found the bug 3 month after launch, then disabled by MCU.
Then I misremember that. I though there was some desktop or laptop CPU that also got it disabled because of the bug.
Intel spent a year fixing bugs from E0 to E5. I'm guessing it's mostly accelerator bugs.
The price of prod units is catastrophic. but If ES is used, and any problems arise you will ask yourself: Is that an ES problem? No one can answer you, unless you can debug with the prod unit.
I have an E0 ES, but I can't evaluate it because I don't have a prod unit.
The price of prod units is catastrophic. but If ES is used, and any problems arise you will ask yourself: Is that an ES problem? No one can answer you, unless you can debug with the prod unit.
I have an E0 ES, but I can't evaluate it because I don't have a prod unit.