Hardware Canucks

Hardware Canucks (http://www.hardwarecanucks.com/forum/)
-   Overclocking, Tweaking and Benchmarking (http://www.hardwarecanucks.com/forum/overclocking-tweaking-benchmarking/)
-   -   Intel Burntest Query (http://www.hardwarecanucks.com/forum/overclocking-tweaking-benchmarking/12596-intel-burntest-query.html)

rjbarker December 6, 2008 10:35 PM

Intel Burntest Query
 
Hey guys...quick question......my previous OC's over this past year have all been proven stable by very long runs of Prime95, both Small FTT's and Blended as well as running Cycles (usually 3) of Memtest86. By very long Cycles of Prime95 I usually go 12 - as much as 18 Hrs.
Now I was reading in the Evga Forum where folks are now using Intel Burn Test as its much faster for detecting instability...so I download it and run it...only to find what I thought were Stable OC's to be unstable (according to Intel Burn Test).
Have any of you guys been using this method? Here was the skinny with my 1800 FSB 4.05 OC
1. Prime 95 Stable Small FTT - 6 hrs (got lazy) - no errors
2. Blended 6 hrs - no errors
3. OCCT - 3 Cycles one hour each - no errors
4. Memtest86 - 5 Cycles no errors

Boom--Intel Burntest shows instability.:doh:

enaberif December 6, 2008 10:40 PM

I don't trust IBT and neither do a lot of people on xtreme forums.

I haven't personally used it but from what some people say it works the cpu to hard.

rjbarker December 6, 2008 11:44 PM

You got that right...it only takes 300 seconds to run one Cycle, so it is quite fast, however, my Core Temps were up to 65c vs 56c in Prime95....I was just kind of like "WTF ?" when it failed all of my OC settings that I had been using over the past 10 months.......
Just thought I'd ask as some of the more "extreme" guys (Evga Forum) are using it in lieu of traditional and swear by it...in hindsight to I did update my BIOS last week and that could be affecting the OC stability as well.."I knew I should have left it alone " !

MpG December 7, 2008 12:20 AM

It's a good temperature failsafe - you can just about assume that whatever temperature it makes your processor hit, you'll never see that in RL use. But yeah, there are a lot of people who don't trust it. If you're in a hurry to find instability, OCCT might be a better bet. It doesn't get quite as hot as Prime95, but it's usually pretty fast at finding errors, and there's been enough work on it that most of the bugs have been ironed out by now.

DK2 December 7, 2008 06:07 AM

From what I can tell Burntest is only about testing various math functions, It just does a lot more than Prime95.
Frankly, calculating Prime numbers, or PI is relatively easy, even to thousands of digits.
I think people get sucked into the fact that because this would be very hard
for a human to do it must also be hard for a CPU. Its just not the case, these programs run
through a few basic math instructions and nothing more, very easy for computers.

I have seen someone say that Intel uses Burntest before shipping, donít believe it!
Intelís QC would freak if anyone at Intel suggested they do that.
There are over one thousand instructions in the x86 language, and each one will
turn on and off various transistors in the CPU according to their function,
and most programs only use at best 10% of the full instruction set when running.
So I think itís a good bet they have their own test suites to do just that, and also
sample testing using random instruction test suites.

In the end I would say the increased heat put you over the edge.

sswilson December 7, 2008 09:18 AM

Ultimately it comes down to what testing you find is required to ensure a stable OC for your needs. I've never used this program as I've found Prime + OCCT + 3DMark testing works well for me. Once folding starts, that's the real test... :)

3.0charlie December 7, 2008 10:24 AM

Quote:

Originally Posted by sswilson (Post 123208)
Ultimately it comes down to what testing you find is required to ensure a stable OC for your needs. I've never used this program as I've found Prime + OCCT + 3DMark testing works well for me. Once folding starts, that's the real test... :)

I second the use of OCCT (30 minutes for quick tests) and then crunching using BOINC.

BALISTX December 7, 2008 10:36 AM

BIONC does stress for sure. I was running it using 50% CPU usage and my CPU temps on the this laptop got to 75C

sswilson December 7, 2008 10:41 AM

Heh... I gave up folding on my laptop. Too much stress for too little points. :)

BALISTX December 7, 2008 10:59 AM

It's the only comp that I got right now I can fold with. I'll be using BIONC on my brother-in-laws to test it.


All times are GMT -7. The time now is 03:34 AM.