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.
Do you know what CPU it could be a prototype/earlier version of? Only production with matching core count and L3 I see on Wikipedia is the W7-3445, unless it is 5420+-type with 8 cores disabled (but then, is 5420+ even tiled, or is it a single-die MCC?). I'm trying to figure out if it's an odd exception to the rest that will still work on the Asrock.
don't do that for ES. intel can change mind every day.
especialy at sapphire rapids, the MCC wan't planed at begin, intel decided late because of the power draw of the EMIBs (tile2tile interconnect)
I'm mainly trying to puzzle out compatibility with my Asrock (and because I want to get a better understanding of these things). If the QYFT was workstation-ish, that would explain to me why it's XCC despite having so few cores. Also, it would be interesting to know whether it has 112 PCIE or 80 PCIE lanes, not that that matters for the Asrock.
80lanes.
QYFT is SPR-SP (not WS)
at that time single DIE (MCC) wasn't planned by intel.
but finaly today there are XCC models with only 32 cores.
Xeon Platinum 8444H is XCC and only 16 cores, 32 threads. So yeah, low core count is not guaranteed.
AFAIR, 112 PCIe ES that works anywhere are extremely hard to find. So unless mentioned otherwise it is 80 lane version.
correct. i have 112L ES C0 stepping, but don't work on any retail motherboards (i guess only intel ES board)
I made a MOD microcode for you. This is still in research phase.
I need results for using this MOD microcode by inserting to your BIOS.
Could you test this?
Thank you!
Attached Zip just include .bin microcode file only. I can't upload .bin directly...
I got it...
Thank you for your support!
If I got more time and money, I'll try it on C741 motherboard with MOD microcode!
Thanks
i have already done and much more. its not just the microcode.
ASUS W790-Ace: POST stops with "CD" you can find this in seccore module of BIOS.
not only the microcode. for each cpuid there is a uncore support module. the SPS (like ME but for server) may block it.
i recommend gigabyte, they have almost all native Xeon Max support.
imho it is not worth the hazzle because early stepping is full of bugs.
By the way, I replaced the entire ME region of ASRock W790 with gigabyte's W790 AI TOP, and it progressed to the AA code.
However, perhaps the Xeon MAX ES was defective, and the BIOS screen did not appear...
I've noticed Granite Rapids (P-cores) / Sierra Rapids (E-Cores) Xeon motherboards have started appearing since august for socket LGA4710 and recently coolers for LGA7529, has someone tried any ES on those yet?
Last edited:
the ASRock W790 has not same PCH pin using (sata,pcie,lan,etc) than the GB W790 (PCH straps)
Found an odd one: QYPR, but it's an E1B. I thought QY** was D0/E1A only. Maybe it was one of the first MCC chips before they moved on to the Q0**?
Link: Intel Xeon GOLD 5415+ ES QYPR 8C/16T 2.4G 150W LGA4677 CPU Processor | eBay
Link: Intel Xeon GOLD 5415+ ES QYPR 8C/16T 2.4G 150W LGA4677 CPU Processor | eBay
I'm also curious about that and also for information if R1S parts would be a separate SKU or all 1S LGA4710 GNR would be R1S or any GNR that you'd put into single-socket LGA4710 would act as R1S part as for some of my projects I'd actually prefer single socket machine that have 6+ PCIe Gen5 x16 slots... For a simple reason - my current dual-socket MS73 with 8490H ES setup have weird distribution of PCIe slots across PCIe Root Complexes of both sockets. One PCIe slot is on its own Root Complex, 2 other on second (repeated for 2 sockets) and whenever I put load on second Root Complex, it affects the other PCIe device on it and I would really like to avoid having that.