Intel Burn is the Best OC Test SW!

I just tried Intel Burn and wow. In just 5 minutes I found that my system
had many CPU errors. I ran tests like Memtest86 and Prime95 for hours and
they did not find any problems. The system seemed fine for months. Once I
raised the CPU voltage 0.05 everything was fixed. It makes sense that Intel
would make the best CPU error detection software.
http://www.ultimate-filez.com/?page=downloads
Using anything else is just a waste of time.

"Kent_Diego" <> wrote in message
news:JarRk.1897$...
>I just tried Intel Burn and wow. In just 5 minutes I found that my system
>had many CPU errors. I ran tests like Memtest86 and Prime95 for hours and
>they did not find any problems. The system seemed fine for months. Once I
>raised the CPU voltage 0.05 everything was fixed. It makes sense that Intel
>would make the best CPU error detection software.
> http://www.ultimate-filez.com/?page=downloads
> Using anything else is just a waste of time.

If your system is running Prime95 and Memtest86 for hours and hours without
errors, and this test is finding them immediately while no others are, it
should tell you that Intel Burn is overly sensitive and not finding errors
of any real consequence. Using it is more of a waste of time.

"Augustus" <no_one@no_where.net> wrote in message
news:P_uRk.1304$xJ3.863@edtnps83...
>
> "Kent_Diego" <> wrote in message
> news:JarRk.1897$...
>>I just tried Intel Burn and wow. In just 5 minutes I found that my system
>>had many CPU errors. I ran tests like Memtest86 and Prime95 for hours and
>>they did not find any problems. The system seemed fine for months. Once I
>>raised the CPU voltage 0.05 everything was fixed. It makes sense that
>>Intel would make the best CPU error detection software.
>> http://www.ultimate-filez.com/?page=downloads
>> Using anything else is just a waste of time.
>
> If your system is running Prime95 and Memtest86 for hours and hours
> without
> errors, and this test is finding them immediately while no others are, it
> should tell you that Intel Burn is overly sensitive and not finding errors
> of any real consequence. Using it is more of a waste of time.

Follow up: My system is an E8400 running at 3.8Ghz / WinXP32 SP3....a
Gigabyte EP35-DSR3 with a C0 Wolfdale E8400 and a matched set (2x2Gig) 4 Gig
OCZ Platinum DDR 1066. The system runs with the FSB set a 422, with no other
BIOS setting tweaks for memory or voltage. That's it. It runs 24/7, and has
done runs of 5-7 days (when I'm on holiday usually) on Prime95, Orthos and
Memtest. It has never generated a single error during any of these runs. It
doesn't locked up, crash or reboot. I ran Intel Burn....and it instantly did
a hard reboot. I tried it again at 3.6Ghz, 3.4Ghz, 3.3Ghz, etc....same
thing. It generated errors at 3.2Ghz and was error free at 3.1 Ghz. Boosting
core voltage to the point where it and generating unacceptable heat levels
(with decent Scythe hs/f and Antec P180 case with great air flow in a 22C
room ) allowed error free operation at just under 3.3Ghz. So by this stress
test, I should drop my overclock by 600Mhz and have the E8400 run 5-7
degrees C hotter at 3.2Ghz than it does running error free at 3.8Ghz
according to every other stress test.

.....
> ... So by this stress test, I should drop my overclock by 600Mhz and have
> the E8400 run 5-7 degrees C hotter at 3.2Ghz than it does running error
> free at 3.8Ghz according to every other stress test.
I would. It is not that you are not error free at 3.8 GHz, they just occur
(very) infrequently. Some day every system will crash and you are left with
the question; is it was from the overclock or some other reason? If you can
go 20 minutes with Intel Burn then you know you have a stable overclock.
Another reason is I have seen systems that appeared OK, but really had RAM
instabilities, corrupt the hard drive. So some day when your system refuses
to boot you will always be asking yourself if it was the unstable overclock?

"Phil Weldon" <> wrote in message
news: m...
> 'Augustus'
>> If your system is running Prime95 and Memtest86 for hours and hours
>> without
>> errors, and this test is finding them immediately while no others are, it
>> should tell you that Intel Burn is overly sensitive and not finding
>> errors
>> of any real consequence. Using it is more of a waste of time.
>
> Since a 0.05 voltage increase was all it took for 'Kent_Diego' to get an
> error free run with Intel Burn, I'd say Intel Burn did a very good job.
> And is MUCH more useful than Prime95 and Memtest86 for overclocking
> purposes. Intel Burn would seem to be much more sensitive to overclocking
> errors and especially the effect of heat, EXACTLY what is needed as a
> guide when making changes to parameters and cooling to increase
> overclocking. After all, a sensitive, small grained test for the smallest
> changes in overclocking is the goal, NOT proper calculations or even less,
> memory parity checks. I'd say your comment on Intel Burn is exactly
> backwards. It is not 'errors of real consequence' that are the most
> useful in a fine grained approach to overclocking, but rather any errors,
> and most especially errors of the least consequence that are most useful.
>
> Phil Weldon
> (who may have to skip the 'Nehalem' because he just bought a Meade LX200
> GPS 12" telescope - on the other hand, it will be USB'ed to my T7400
> notebook for control and digital image transfer which may require
> semi-heroic digital manipulation, so...)
>
> But I'll never leave a.c.h.o; the level of participation here has about
> the highest tone and deepest knowledge base of any Usenet group I've
> een. - despite Augustus' momentary lapse B^)
>

I like it too. It does exactly what I would want when overclocking since my
goal is to go as high as possible without "any" errors. I will still
probably do a Nehalem build in early '09. I hinted about doing some
astronomy to my wife and she didn't make any negative comments so I am
looking around at what I may need. It is almost Greek to me at this point,
but I will see how your setup goes and let you lead the way.......:-).


Ed

> But I'll never leave a.c.h.o; the level of participation here has about
> the highest tone and deepest knowledge base of any Usenet group I've
> een. - despite Augustus' momentary lapse B^)

My lapse, if I may elabortae, was simply this: If my system is being 100%
stable running the games and software that i use, and not generating errors
at 3.8Ghz with any of the tried and true standard o/c stress tests done over
rather lengthy time periods, why would I wish to raise the CPU voltage and
drop the FSB and speed based on the results of this one stress test which
seemingly (to me, anyway, but feel free to enlighten me) does not relate to
the stable o/c operation of my current system or software. The heat
generated on my E8400 dramatically increases with increasing voltage. This
test on my system indicates continual errors in operation with less than a
10% o/c and fair amount of vcore boost. Yet the system is error free, cooler
and stable at much higher speeds and stock vcore on EVERY other stress test
and the apps/games I run. What am I not getting here? Seriously......