Tbh, certain versions of p95 arent great for ivybridge/haswell/devils canyon. You can easily hit throttling temps at stock nevermind oc'd. Therefore not a great stability indicator.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Skylake CPUs can freeze running Prime95
- Thread starter HazardO
- Start date
Associate
- Joined
- 5 Jul 2014
- Posts
- 231
- Location
- Northumberland, UK
Since a BIOS update fixes skylake its not a massive issue but I'd argue any finger pointing about those CPU generations you mention should be directed to Intel and not P95 itself. My 3570k heats up rather a lot on the latest versions with AVX support - tbh I think the whole TIM issue is to blame for that, Intel went the cheap route and I'm not sure if they even thought about AVX workloads - utilisation of AVX really pumps up the temps on Ivybridge (at least from my experience).
If I were to delid it and sort it out I'm sure I can run much cooler but this CPU I have did not do well in the silicon lottery, it forgot it was on that day!
The voltage needed for > 4.2 makes it too hot to run encoding tasks.
For info I run at 4Ghz and need 1.21v - 4.2Ghz needs 1.30+ my CPU cooler is a NZXT HAVIK 140mm.
As for Prime - I've used it for years that and the program "ycruncher" are my go to's for stability. I also actually use Prime95 for its intended task so one of the few!
PS - FOr a laugh I would love to see what 8Pack could put out on ycruncher see if he could make the leaderboards
http://www.numberworld.org/y-cruncher/#FastestTimes
Since a BIOS update fixes skylake its not a massive issue but I'd argue any finger pointing about those CPU generations you mention should be directed to Intel and not P95 itself. My 3570k heats up rather a lot on the latest versions with AVX support - tbh I think the whole TIM issue is to blame for that, Intel went the cheap route and I'm not sure if they even thought about AVX workloads - utilisation of AVX really pumps up the temps on Ivybridge (at least from my experience).
If I were to delid it and sort it out I'm sure I can run much cooler but this CPU I have did not do well in the silicon lottery, it forgot it was on that day!
The voltage needed for > 4.2 makes it too hot to run encoding tasks.
For info I run at 4Ghz and need 1.21v - 4.2Ghz needs 1.30+ my CPU cooler is a NZXT HAVIK 140mm.
As for Prime - I've used it for years that and the program "ycruncher" are my go to's for stability. I also actually use Prime95 for its intended task so one of the few!
PS - FOr a laugh I would love to see what 8Pack could put out on ycruncher see if he could make the leaderboards
http://www.numberworld.org/y-cruncher/#FastestTimes
TIM isn't the only problem - Skylake has TIM also but 4.6-4.8Ghz overclocks are very common on air even.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Need some benchmarks to see if the fix reduces performance in any situations, but in general Intel problems get brushed under the carpet. AMD TLB bug which they were honest about upfront, told everyone about and pushed out a fix, maybe even before launch(can't quite remember) and it was the end of the world. Haswell(iirc) has a TLB bug and absolutely no one notices. Intel have multiple chipset problems over multiple generations, no one gives a crap. Another CPU bug, a rather big one, goes fairly under the radar with everyone just calling it not a big deal.
I'm under the impression that Intel's big initial fix was to tell everyone try not to use certain instruction types, genius.
Funny how when a bigger richer company has a issue the media downplay it as if it's nothing if it goes reported at all, when it's a competitor of a rich company that problem is talked about as if dooms day everywhere possible.
I'm under the impression that Intel's big initial fix was to tell everyone try not to use certain instruction types, genius.
Funny how when a bigger richer company has a issue the media downplay it as if it's nothing if it goes reported at all, when it's a competitor of a rich company that problem is talked about as if dooms day everywhere possible.
4.7/8 was pretty much standard on devils canyon too on air, (had two that hit theese speeds). But p95 ver 28.5 would send both into the mid 90's at stock under a very capable cooler. Be interesting to see how skylake does running this. I've built a few skylake systems for friends but didn't get much of a chance to play about with them.
That doesn't mean his 'opinion' is always right though does it.
Soldato
- Joined
- 14 Nov 2012
- Posts
- 17,955
- Location
- Close to Swindon, but not Swindon
Soldato
- Joined
- 5 Dec 2010
- Posts
- 3,212
- Location
- Solihull
My weapon of choice is OCCT cpu test - large data set with error checking turned on.
15 mins of that and I can be sure of stability.
I have used Prime for many years though and it is very good.
I'm with zoomie in that any less that 100% isn't good enough. It needs to hold up well in any situation.
15 mins of that and I can be sure of stability.
I have used Prime for many years though and it is very good.
I'm with zoomie in that any less that 100% isn't good enough. It needs to hold up well in any situation.
I've heard the latest couple of versions stress so hard they cause throttling. Not as useful from an overclocking perspective, but then it primarily exists to find prime numbers!
Yeah. Was thinking just re-run geekbench 3 before/after, unless there's something better anybody can suggest here?
Yes indeed. Which means Prime95 is quite an appropriate tool for doing that. As a way test whether or not your cooling solution is going to be thermally stable / keep up under all possible load conditions. That is a concern if you intend to ever do long-running heavy compilations. Or other kinds of CPU intensive calculations such as simulation work. I.e. stressful over significant durations which are not disk-bound and require you to leave the machine unattended. Which is not any concern for gamers. Unless the game is seriously buggly (like Batman / whatever). In which case it will be unplayable anyway maxing out the CPU like that.
Anyway the point of this topic is supposed to be about this Skylake CPU bug and not the relative merits of various OC tools. Its my OPINION (not confirmed yet) that this is indeed the cause of a freezing crash which can occur at any time. Which would mean it is not purely inside of the Prime95 tool only. Just that prime 95 happens to be a tool which can help users determine if they are affected.
Only after microcode update arrives can this be confirmed. Hmm lets see... Havent received microcode yet from MB BIOS vendor yet. Havent checked my windows 10 recent KBs either. But would not surprised if they provied through Microsoft also. Since it is a significant CPU bug which affects system stability and therefore probably meets the necessary requirement of being a critical update.
Last edited:
Except mathematicians and those that understand the usefulness of large prime numbers.
Intel's super computer, scientific, academic and financial customers will be very interested in these kind of bugs. The kind of people who place single orders that run into the millions.
Tried to use it on mine last week, exactly this issue... But as people rightly say I don't think its a great measure of stability for general use. Runs every other stress test fine, some very heavy use for weeks with things overclocked tells me its stable for the tasks at hand.
If its not prime stable, its not a stable overclock... But i guess i can post pictures of cpuz showing 5ghz and say woohoo i got 5ghz when even the smallest amount of load would crash it.
Just because prime95 will cause huge temperatures doesn't mean it useless today... If anything thats perfect to see if you able to cool your beasty overclocks. Whats the point of having a huge overclock if its unstable at 100% load?
Just because prime95 will cause huge temperatures doesn't mean it useless today... If anything thats perfect to see if you able to cool your beasty overclocks. Whats the point of having a huge overclock if its unstable at 100% load?
There are plenty of stress testing programs which run my 6700k at 100% with no problems for hours at great temps, Prime95 crashes in minutes, or at least it did before this 'update' i've not checked since (don't care). The temps are not really an issue as they sit mostly under 80c but it still blue screens. I'm sorry but nobody here is stupid enough to simply look at the speed listed in CPU-Z and think its stable, thats just silly talk we are all aware of putting the CPU under more than 2% load.
Prime is not really a comparative benchmark of a 'stable' overclock, the fact my overclock 4.6ghz has been running perfectly fine with heavy gaming, encoding, editing for weeks tells me its stable... So really what do I care if a program for which I have no use crashes?
I don't think there is any argument here that Prime is THE platform for stress testing, more people rightly pointing out that it has other functions.
Funny how no Skylake user complained so far about crashes when doing encryption, video rendering, or any other task apart from this one guy who works with prime 95.