Printable View
May I know if the figure(s) in the "Results" column matter when it comes to stability? i.e. I'm able to complete 25runs of IBT using different OC settings, but sometimes I get 2.something instead of the more "common" 3.something in the results column. Does that make a difference?
Also, since I can complete IBT on a few different set of OC settings, I should use the one that takes the shortest time per run & give the highest Gflops right?
Updated to v2.51
- Linpack binaries updated + patched for AMD CPUs
- Official support for Six-Core CPUs and Sandy Bridge Architecture
- Added new algorithm to detect calculation errors faster
- Fixed a bug that caused blank results to be valid
Hopefully this update helps address issues and concerns raised by you guys. Haven't had time to maintain the program until now. I appreciate the support and encouragement that you guys have given me :)
Looking great! Thanks for all your great work, about everything you do is amazing! :)Quote:
Originally Posted by AgentGOD
+1 :up: - who want's to run 24hrs of Prime95 when you can run IBT for 5 or 10 passes and have the same confidence in your overclock!Quote:
Originally Posted by AgentGOD
Cheers and keep up the good work!
Great job buddy. Can't wait to try this out on Sandy when it gets here Wednesday.Quote:
Originally Posted by AgentGOD
I can tell you 24 hours of Prime95 does NOT equal 10 runs of linpack, I don't know where that came from. Its more like 24 hours of Prime95 (blend using all memory customly) is more equal to 10 hours of linpack all memory regardless of max RAM.Quote:
Originally Posted by cbjaust
Threads on auto is still broken for i7 920 and probably all other 4 core HT CPUs, all threads jump around 40%~60% use I still have to manually set it to 8 threads, as well xtreme stress mode near instantly freezes the system overclocked, stock, underclocked, every system its been tried on. Other than that its fine I'm starting to use it more over LinX it has slightly higher GFlops with slightly less time.
The "Auto" mode goes by what Intel's Linpack decides to do. From what you describe, it sounds like it's defaulting to loading up threads only up to the number of physical cores.Quote:
Originally Posted by GigaByte
It's not frozen. It may appear that way because Linpack is given the highest priority in Xtreme Stress Mode.Quote:
Originally Posted by GigaByte
Well, thank you.Quote:
Originally Posted by GigaByte
So is it 8 threads or 16 threads for core i7 2600K now?
4 cores, 8 threads ;)
Thx I cant seem to hit a 5ghz OC without crazy voltage, my max seem to be 4.8 before I hit the wall. Not even stable at 1.52v in bios for 5ghz on maximus iv extreme mobo w/1303 bios.
4.8 is stable at 1.42
I'm having a bizzare problem that seems to have started out of the blue. As of a week ago I could run any varrient of linpack ( old or newer supporting avx intructions ) at max memory. Now when I try the executable instantly crashes (ie linpack_xeon64.exe or linpack64.exe) I can set any value at or below 4gb without issue and can run hundreds of passes without error.
The bizzare thing is this is occuring at stock cpu / memory clocks, overclocked and even underclocked. I've tried switching the dimms, made sure the pins were clean / dust free , no cigar. I've ran memtest on them for multiple passes no issue. And now the strange part, I can run 2 different varrients of linpack through different gui with 4 threads each and affinity set by exe ( 1 in ibt and another in linx ) with memory set close to 90% between the 2 ( 3072 per instance ) and no problem at all. Everything I've tried indicates a software issue when attempting to address more than 4gb to a single executable and for the life of me I can't figure out what it could be.
Heres an example of what I did :
http://img813.imageshack.us/img813/7...x3072mb.th.jpg
And here is IBT with just 1024MB problem size :
http://img718.imageshack.us/img718/9...x1024mb.th.jpg
I've ran 1024 300 passes no problem and 4096 a good 50 passes. Before this issue surfaced I had been able to run roughly 7000 for 40 passes.
If anyone has any suggestions as to what might be causing this I'm all ears. I'm still fairly confident it isn't a hardware problem based on my findings however so any software suggestions are welcomed.
EDIT: There appears to be an exact amount where things crash. With the linx binaries I can't set a problem size any larger than 4111MB and with the BurnTest binaries I can't set a size larger than 4295. In both cases 1MB more results in an instant crash. Looking at memory usage in the task manager they were both addressing around 4230MB to 4230MB with a very similar commit charge at the time (5505-5510) I ran mem test another 5 full passes for comformation and no errors... This crap is werid.
EDIT2 : I cross referenced the application exception codes and no suprise they match...
Intel burn test (linpack64.exe) – EC c0000005 – EO 000000000032bf0a
Ychruncher (x64 ACX) – EC c0000005 – EO 0000000000001f80
LinX (linpack_xeon64.exe) – EC c0000005 – EO 0000000000002da1
I then figured why not try these in safe mode ( had it in my head that wouldn't work - suffered a medication reaction and was in the hospital over the weekend so still not all here cognitively ) and to my surprise they all worked fine addressing the full amount of memory. Now I'm down to trying to figure out what driver / application is at fault. The only thing I recall changing is going from the forware 268.55 to the 270.61s ( again ) Who knows... I'll add more when I find the culprit.
SOLUTION : Found the culprit... Comodo Firewall. It recieved an update recently which fits my timeline for being able to run said applications trouble free beforehand. Trying to figure out if I can work around it or if it must be a software fix on their end. I'm also running Comodo on my file server so I will be sure to test this out on it as well *** Tested on other machine and comfirmed, Comodo is causing applications wanting to address more than 4gb to crash. Uninstalling Comodo solves the issue, still playing around with things to see if a work around is possible.
Glad you figured out what was causing Linpack to fail ;)
Mine's reporting wrong memory? Less than installed and wont run. 2GB installed, reports around 500MB but varies a little. I5 2500K and P8P67 Pro...
It's reporting free memory amount. ;)Quote:
Originally Posted by peterpiper
Thx zalbard. Forgot I posted here!
Thanks for this awesome app.
I just used it to check stability on my new AMD chip. Hopefully this means that my system will now be F@H 24/7 stable. Keep up the great work! :up:
I've just ran IBT for 5 Passes @ Maximum stress level. It passed, but my GFlops were only @ 40. I reran it now with just Standard Stress level, but my Gflops are all over the place:
1st pass: 32.2105
2nd pass: 48.3728
3rd pass: 36.4371
4th pass: 37.1269
5th pass: 45.2307
This was with IBT 2.51 loading 16 threads.
The results are coming in appropriately, but I'm unsure whether this is stable or not?
Specs are correct in sig.
Follow-up question, is this a good stress-test to determine RAM stability when running the Stress level on Maximum?
EDIT - Here's a screenshot:
http://img651.imageshack.us/img651/2905/unledupa.th.jpg
Uploaded with ImageShack.us
EDIT 2: I forgot to note that both C1E & Intel Speed Step are disabled in the BIOS. So I don't believe any throttling is happening. My temperatures for each core (according to Realtemp) are 81-82c on the 1st core, 79-80c on the rest.
Updated to v2.52
- Linpack binaries updated (7-29) + patched for AMD CPUs
- Branding update for Xtreme Gaming Studio
Awesome! thank you AgentGOD! IBT is the best to keep my machine 100% stable! Q9650 @ 4ghz here!Quote:
Originally Posted by AgentGOD
How much memory should be used for the test and how many loops?
Quote:
Originally Posted by shimmishim
from the OP/1st post, you got to scroll down and it will say....
2. Use the most available RAM possible (IntelBurnTest can now do this on
its own automatically). The more memory it uses, the more accurate.
However, most people can use "Standard" mode as it should be sufficient.
3. Select the run # (should use at least 5 at minimum, no more than 20).
I have a Llano mobile chip.
This cpu allows me to change cpu multiplier, while everything else remains stock.
I want to use IBT to test cpu only, minimal ram.
If I set custom to 128 MB, and let it run for several hours, will that give me a good cpu-only test ?
I did a short run of 20 times, and it did peg the cpu at 100% load, so running for a few hours should be good ?
Thanks. I didn't make it quite that far down. :)
Quote:
Originally Posted by Gomnadz
lol, done that before....... :D
Quote:
Originally Posted by shimmishim