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.
Well, apparently I was wrong. I received some additional screenshots from the Japan seller. Not exactly what was requested, but more details for sure:
I'm pretty sure prime95 is very highly not recommended to use anymore. From what I've read and what intel says is that it damages the CPUs now. It didn't use to be this way but some time a few years ago something changed with their architecture that when using prime95 it permanently damages the CPU. I think it has to deal with causing such extreme temps that the chips or silicone break down.Man, now I'm so confused....
So what is Friz chess bench doing that Prime95 is not?
Sent from my iPhone using Tapatalk
I agree with this and Linpack also.
Both of these are fine for a few minutes I think, for a simple stability test. But I have seen people run both of these for very long periods, then the exact opposite of what they are trying to do, show stability of the system starts to happen. The CPU degrades and starts blue screening, this can carry over to normal system operation when slight loads cause a blue screen.
Massive prime number benchmarks searching for extremely large prime numbers run for long periods is extremely hard on CPU's. I believe in the CPU the circuits that get used for these benches get over worked and like yobigd20 said, causes way to much heat in very small areas causing degrading of that area. I actually have a long story about this I could post.
I have seen this happen with CPU's running at stock speeds and especially with OC'd systems. Most of my testing was back with 3960x's and dual cpu systems from that time period. Stability would suffer resulting in backing off of settings, then stability suffers again etc.
This is one of the main reasons I do not OC anymore, I have seen to many CPU's degrade over time. Even Phase Change and Water Cooling can not help this effect, it just takes longer for stability issues to start showing up, but they will.
Do you have any Intel sources which adresses this problem or is it something that is generally accepted as a truth?
My experriences with electromigration is almost purely due to extreme voltage which produces more heat than what is possible do dissipate quickly enough. Todays CPUs with adaptive voltage, frequencies and thermal throttling have I not seen or even expectt tis to happen. I has mostly regarded it as rants on forums. Wrong or right, I do not know, but some hard facts would be informative.
Sadly I do not have any Intel info that address this.
Most of my experience was with overclocked systems so yes voltages were raised for this. And again my experiences were with mostly 3960x's and various other desktop cpu's.
I used to build a fair amount of systems for HFT guys, they will disable HT and get 4.8ghz out of these 3960x's. They loved them and some are even still running today after years of operations.
I had another system that was phase changed cooled with a 990x in a data center, fist thing they did was start running Linpack on it and they forgot they had it running, a few days later the CPU had degraded and eventually replaced.
Even with E5-2687W with a 103 BCLK on dual proc rigs I saw strange things running prime number crunchers for very long periods.
Most of my experience was with overclocked systems so yes voltages were raised for this. And again my experiences were with mostly 3960x's and various other desktop cpu's.
I used to build a fair amount of systems for HFT guys, they will disable HT and get 4.8ghz out of these 3960x's. They loved them and some are even still running today after years of operations.
I had another system that was phase changed cooled with a 990x in a data center, fist thing they did was start running Linpack on it and they forgot they had it running, a few days later the CPU had degraded and eventually replaced.
Even with E5-2687W with a 103 BCLK on dual proc rigs I saw strange things running prime number crunchers for very long periods.
there are A LOT of end users that have the failures, so thats proof enough, (easy to search the web for all those threads on lots of various forums), but I've never seen anything "official" from Intel other than small statements like this: Prime95 and Haswell that admit theres an issue, but they never say anything else or go into detail. I think they are just trying to make sure the issue doesn't get blown up in the lime light as it could hurt their business.
Intel has identified an issue that potentially affects the 6th Gen Intel® Core™ family of products. This issue only occurs under certain complex workload conditions, like those that may be encountered when running applications like Prime95. We have not experienced any issues with the 4th Generation Intel® Core™ i5 Processors and Prime95.
Just bear in mind that Intel does not validate third party tools as per diagnosing our products. I invite you running the Intel® Processor Diagnostic Tool
I highlight that last sentence because AFAIC that's them saying "yep, it'll destroy the product, and we dont want to get sued so don't use it, and only use our own test tool which won't break our own product".Just bear in mind that Intel does not validate third party tools as per diagnosing our products. I invite you running the Intel® Processor Diagnostic Tool
these are sceenshots of a Xeon sold as E5-2630L V4 1.7 GHZ ES2
its not exact like the prod. 2630L v4 (1,8GHZ),
the author of HWinfo told me this is an intermediate ES without a official name.
Last edited:
First of all, thanks for all of your information throughout the thread. Please let me know of your opinion.
I have Asrock x99 Extreme4 BIOS 3.40 (most current). And I just purchased E5 2695v3 QFQG. I wonder if they are compatible.
ASRock > X99 Extreme4
Based on the link above my motherboard is supposed to support both C0 and C1 stepping but I am not sure what QFQG stepping is... hopefully it is C0 at least.
http://www.cpu-world.com/sspec/QG/QG7R.html
This webpage only shows QG7R as C0; there is no information about QFQG
Do you think QFQG and QG7R are very different from each other in terms of stepping? Thanks ahead for your help
p.s Do Asrock drop the compatibility for previous ES cpu? For example, maybe older BIOS versions could run engineering samples of v3 but no longer support them with newer BIOS. My 3.40 version seems too new to run v3.
Thank you so much. That helped alot. I really really appreciate it. Hopefully my bios support it.
Why did you not go for a E5 2690 V4 instead? It runs cooler, is faster and have the same price.
NO.
E3 xeon are only (SP)single processor
there are no E5-1xxx (SP) cpu too.
maybe it is not worth because they are rel. cheap. as used/retail.
Last edited:
There are E5-16xx (SP) processors. I've even seen a couple ES E5-16xx floating around, but very few.
Question about ram to go with ES Xeons.
Bought the Asrock X99 WS, but now the E5-2686V3 sold out on ebay again, so I'm trying to figure out what ram to get.
I'd like to get Reg ECC 16Gb on each memory stick, but only one option is listed on official support list. Has anyone used anything else on this motherboard?
Also, anyone know any other way to buy that cpu besides ebay? (I'm located in EU).
I see plenty of them on taobao :|
Bought the Asrock X99 WS, but now the E5-2686V3 sold out on ebay again, so I'm trying to figure out what ram to get.
I'd like to get Reg ECC 16Gb on each memory stick, but only one option is listed on official support list. Has anyone used anything else on this motherboard?
Also, anyone know any other way to buy that cpu besides ebay? (I'm located in EU).
I see plenty of them on taobao :|