-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
Torture Test System?
- Thread starter jjsimas
- Start date
Sephiroth_IX
Diamond Member
I use prime95, but you should check out our very own RC5 client.
Also, download some huge compressed files (<cough>warez) and execute them to make sure your memory and IDE are running well.
Also, download some huge compressed files (<cough>warez) and execute them to make sure your memory and IDE are running well.
NaughtyusMaximus
Diamond Member
holy cow i cant believe you guys think RC5 is a good stress test
you go "benchmark' a Celeron 700 vs a P3/700
and with anand's great RC5 benchmark you will find there is NO DIFFERENCE
i seriously wonder how good this is a 'benchmark' if it doesnt show any difference between a Celeron700 and a P3/700E
doesnt show difference between a Duron 700 and a Tbird 700
you go "benchmark' a Celeron 700 vs a P3/700
and with anand's great RC5 benchmark you will find there is NO DIFFERENCE
i seriously wonder how good this is a 'benchmark' if it doesnt show any difference between a Celeron700 and a P3/700E
doesnt show difference between a Duron 700 and a Tbird 700
BurntKooshie
Diamond Member
is it the best system stress test? No. Is it a good CPU stress test? Yes.
It uses the integer units heavily (it does no FPU calculations), it fits inside any current MPU's L1 cache, therefore stressing that as well. Will it push your video card to the limit? Heck no! It will put some strain on the CPU, as it is doing work, 100% of the time. In many cases, the CPU is what causes an overclocked system to fail (of course, it could be any of the other overclocked parts of an abused system).
RC5 stresses one thing: the CPU. If your system is unstable doing heavy 3d work (read: Q3 or some other game or whatever), but is STABLE using RC5 (or something like OGR, or gamma flux, or Prime 95, and numerous others) as those are similar in concept to RC5, ie, high CPU usage, low strain on everything else.
Soooo....what is my point? You can use a distributed computing project to try to pinpoint what is causing a system to be unstable. S@H stresses (well, it used to much more than it does now, as it was more inefficient before the most recent client updates) the main memory, and to some extent, the I/O subsystem (read: harddive) as well, due to its frequent swapping od large ammounts of data.
There are obviously other ways to torture test, but RC5 (and other distributed projects), is on that list.
<< i seriously wonder how good this is a 'benchmark' if it doesnt show any difference between a Celeron700 and a P3/700E
doesnt show difference between a Duron 700 and a Tbird 700 >>
Please try and understand what each "benchmark" does. Like I said: RC5, and many others (S@H NOT included) reside nicely inside 32k (and even 16k) of L1 cache. What is the difference between a Culeron? Cache size/associativity, and multiplier lock (which determines the FSB). That's it. The L1 is the same. The functional units are the same. They perform the same per clock. The same holds true for the Duron/Thunderbird, however, the Duron has the same associativity, as it is a different CPU (Culerons are P3's with 1/2 their cache disabled, which causes associativity to be cut in half).
I am glad that you at least looked up the benchmarks, that proves that you tried to do some research to hold up your argument. I like that. I respect that. But, in this case, you are wrong.
BK
[EDIT] added info 'bout the Duron/T-bird
It uses the integer units heavily (it does no FPU calculations), it fits inside any current MPU's L1 cache, therefore stressing that as well. Will it push your video card to the limit? Heck no! It will put some strain on the CPU, as it is doing work, 100% of the time. In many cases, the CPU is what causes an overclocked system to fail (of course, it could be any of the other overclocked parts of an abused system).
RC5 stresses one thing: the CPU. If your system is unstable doing heavy 3d work (read: Q3 or some other game or whatever), but is STABLE using RC5 (or something like OGR, or gamma flux, or Prime 95, and numerous others) as those are similar in concept to RC5, ie, high CPU usage, low strain on everything else.
Soooo....what is my point? You can use a distributed computing project to try to pinpoint what is causing a system to be unstable. S@H stresses (well, it used to much more than it does now, as it was more inefficient before the most recent client updates) the main memory, and to some extent, the I/O subsystem (read: harddive) as well, due to its frequent swapping od large ammounts of data.
There are obviously other ways to torture test, but RC5 (and other distributed projects), is on that list.
<< i seriously wonder how good this is a 'benchmark' if it doesnt show any difference between a Celeron700 and a P3/700E
doesnt show difference between a Duron 700 and a Tbird 700 >>
Please try and understand what each "benchmark" does. Like I said: RC5, and many others (S@H NOT included) reside nicely inside 32k (and even 16k) of L1 cache. What is the difference between a Culeron? Cache size/associativity, and multiplier lock (which determines the FSB). That's it. The L1 is the same. The functional units are the same. They perform the same per clock. The same holds true for the Duron/Thunderbird, however, the Duron has the same associativity, as it is a different CPU (Culerons are P3's with 1/2 their cache disabled, which causes associativity to be cut in half).
I am glad that you at least looked up the benchmarks, that proves that you tried to do some research to hold up your argument. I like that. I respect that. But, in this case, you are wrong.
BK
[EDIT] added info 'bout the Duron/T-bird
RC5 only stress CPU+ L1 cache
Prime95/Seti stress CPu+L1 +L2 + system ram
i seriously dont know which one is 'better'
if it cannot even stress the L2 - which is considered part of a CPU, how is it a good 'cpu' stresser
you guys recommend RC5 cause you want more people join your team
thatz why u recommend
Prime95/Seti stress CPu+L1 +L2 + system ram
i seriously dont know which one is 'better'
if it cannot even stress the L2 - which is considered part of a CPU, how is it a good 'cpu' stresser
you guys recommend RC5 cause you want more people join your team
thatz why u recommend
BurntKooshie
Diamond Member
it depends on if you are trying to pinpoint the source of the instability or not.
BCM Diagnostics has a built in stress test, works pretty well. I'd disable the annoying parts first, though, like the floppy test. Their site seems to be down right now, but you can find the freeware part on a lot of sites.
BCM home page
BCM home page
yes RC5 ONLY test the CPU+L1 (wow how useful)
while Prime/Seti does CPU+L1+L2+System RAM
so guess which one is it better to use? to test an 'overclocked system'
i bet Prime/Seti is MUCH BETTER to do this task
you talking about pinpointing - that is AFTER u find out there is a problem with your OCed system
you dont 'pin-point' when you start to test stability of your system
while Prime/Seti does CPU+L1+L2+System RAM
so guess which one is it better to use? to test an 'overclocked system'
i bet Prime/Seti is MUCH BETTER to do this task
you talking about pinpointing - that is AFTER u find out there is a problem with your OCed system
you dont 'pin-point' when you start to test stability of your system
BurntKooshie
Diamond Member
I'm not going to get into the scientific method, as I'm sure you understand what it means. Only recently has the L2 been integrated on the CPU. In many cases, people disabled the offdie L2, that way they could see if it was the core itself, or the external ram that was causing it to be the problem.
Do you know that you can still disable the L2 on CPU's with integrated L2 cache? If disabling it makes it more stable, then you know that the L2 is what's causing the instability. If it doesn't do anything, then you know its not the L2.
Do you see how you can use programs to stress different parts of your system? and how they interact with the rest of the system when one variable is changed in a system? This is what allows RC5 to be on the list of things you should potentially use in a "torture test". I am by no means saying its the only one (that would be ludicrous). I hope you understand my point, because alas, I'm tired, still have homework to get done, and don't have time to try and force what I've said upon someone who isn't willing to listen (no accusations - yet. You could prove me wrong and say something completely stupid, but generally speaking, you've at least attempted to bring valid points).
Do you know that you can still disable the L2 on CPU's with integrated L2 cache? If disabling it makes it more stable, then you know that the L2 is what's causing the instability. If it doesn't do anything, then you know its not the L2.
Do you see how you can use programs to stress different parts of your system? and how they interact with the rest of the system when one variable is changed in a system? This is what allows RC5 to be on the list of things you should potentially use in a "torture test". I am by no means saying its the only one (that would be ludicrous). I hope you understand my point, because alas, I'm tired, still have homework to get done, and don't have time to try and force what I've said upon someone who isn't willing to listen (no accusations - yet. You could prove me wrong and say something completely stupid, but generally speaking, you've at least attempted to bring valid points).
BurntKooshie
Diamond Member
wow, you keep editing your posts 😛 Makes me look foolish when I'm replying to something that is no longer there😀. Most people, to make sure they stay credible around here, add something along the lines of "[EDIT] - WHATEVER WAS EDITED HERE" so that we don't get trolls around here. Just my opinion in this regard.
But, considering your now altered stance, I would agree with you. RC5 isn't the best system test, but I never said it was. I said it was a good CPU test, and is the CPU not a part of the system? I don't understand your zealotry against something like RC5. Not my loss.
But, considering your now altered stance, I would agree with you. RC5 isn't the best system test, but I never said it was. I said it was a good CPU test, and is the CPU not a part of the system? I don't understand your zealotry against something like RC5. Not my loss.
if the L2 is actually failing but your CPU is not
you really should not be disabling that L2 since the performance you loose is much greater than the extra Mhz you gain due to the modern CPU architecture L2 is really 'part' of the CPU
this renders a NON-L2(RC5) cache test program useless
you really should not be disabling that L2 since the performance you loose is much greater than the extra Mhz you gain due to the modern CPU architecture L2 is really 'part' of the CPU
this renders a NON-L2(RC5) cache test program useless
BurntKooshie
Diamond Member
Why're we crazy about RC5? I'll tell you why:
1) We've used it to track stolen computers (no joke. if you want the whole story, PM me or something).
2) We've used it to detect viruses (no joke. if you want the whole story, PM me or something).
3) We use it to see if the CPU is the limiting factor in overclocking (hymmm....we've discussed this one here).
4) It keeps the temperature of the CPU constant (shouldnt' normally be a problem, but in some situations, ie, if the environment is too cold, heating up / cooling down can be bad...I'll leave it to you to figure out why).
5) If you are going to be wasting energy anyway (read: not using a program like rain, cpuidle, or running an OS with the HTL instruction built it), you might as well have that energy do something - ie, search for the optimal golumb ruler, find the optimal storage medium for nuclear waste, etc.
6) It does, to a small degree, show the relative performance potential within a given architecture. I say to a small degree, because only those things which can fit into the L1 cache can truly show off the architecture (not even being in the L2 can do that, as that is something that is much more easily modified).
Goodnight, I've wasted too much time already on this, I still have programming, and Pro engineering work to get done.
BK
[EDIT] Ok, so I lied. I'm not going to bed.
<< you really should not be disabling that L2 since the performance you loose is much greater than the extra Mhz you gain due to the modern CPU architecture L2 is really 'part' of the CPU >>
Wow, we agree on something else. You're right, but some people still like to know what's causing it. I've known people who have done it, just to check. And they weren't RC5 fanatics.
1) We've used it to track stolen computers (no joke. if you want the whole story, PM me or something).
2) We've used it to detect viruses (no joke. if you want the whole story, PM me or something).
3) We use it to see if the CPU is the limiting factor in overclocking (hymmm....we've discussed this one here).
4) It keeps the temperature of the CPU constant (shouldnt' normally be a problem, but in some situations, ie, if the environment is too cold, heating up / cooling down can be bad...I'll leave it to you to figure out why).
5) If you are going to be wasting energy anyway (read: not using a program like rain, cpuidle, or running an OS with the HTL instruction built it), you might as well have that energy do something - ie, search for the optimal golumb ruler, find the optimal storage medium for nuclear waste, etc.
6) It does, to a small degree, show the relative performance potential within a given architecture. I say to a small degree, because only those things which can fit into the L1 cache can truly show off the architecture (not even being in the L2 can do that, as that is something that is much more easily modified).
Goodnight, I've wasted too much time already on this, I still have programming, and Pro engineering work to get done.
BK
[EDIT] Ok, so I lied. I'm not going to bed.
<< you really should not be disabling that L2 since the performance you loose is much greater than the extra Mhz you gain due to the modern CPU architecture L2 is really 'part' of the CPU >>
Wow, we agree on something else. You're right, but some people still like to know what's causing it. I've known people who have done it, just to check. And they weren't RC5 fanatics.
CPU Stability Test Here.
Try it. I'll bet many of you who think your systems are stable wont get thru the 12 hours of continuous running required to generate the certificate. 😉
Try it. I'll bet many of you who think your systems are stable wont get thru the 12 hours of continuous running required to generate the certificate. 😉
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
