I have a dual-core cpu (as listed below in my sig) and I'm trying to get two instances of Prime95 going at the same time, one for each core. I don't know if I'm doing something wrong, or something is possibly defected on my system (oh god I hope not) or maybe something isn't suppose to read right.
Well here is how I'm going about it. I unzip it once to my C: drive then again to my F: drive (got my XP install drive (the raptor) and a 250GB SATA storage drive). I place shortcuts of each on my desktop, one on the left side another on the right so I don't get them confused.
I'll open the Prime95 shortcut on the left of the desktop (the one for the C: drive version), in the program I'll set the Affinity to "CPU 0" THEN in Task Manager select it and set the Affinity also to "CPU 0". Then I open the Prime95 on the right side of my desktop (the one for the F: Drive version), again in the program I'll set the Affinity to "CPU 1" now THEN open Task Manager again select this new one and set the Affinity there to "CPU 1" too (and I know I didn't select the first instance again because it still has CPU 0 and 1 checked).
Next I'll start the torture test on the first one (CPU 0) and then start the torture test on the second one (CPU 1). Now here is where it doesn't seem normal. If I open Task Manager under the Performance tab or even my Asus Probe 2 it will show CPU 0 at 100% and CPU 1 at about 0%-6%??? And it'll do this if I do it the other way too, if I start the torture test on CPU1 first then CPU0 then CPU1 will be 100% and CPU0 0-6%. And so under Task Manager it'll say my Total system usage is only at 50-51%.
This doesn't seem right to me, so I'm wondering what is wrong, or what am I doing wrong, or does it always just report like this, or maybe even is something broke? I did notice today (set it up to run while I'm at work right now) that if I set the Priority on the 2nd instance (CPU1) to High it does flucuate the CPU1 usage from 0-6% up to 18-20% but still not 100% total usage of both cores.
Maybe it's just me, I just assumed running two instances of this program, one for each cpu would take up close to 100% of each cpu's resources
Oh and I had run one instance for each core seperately and get no erros after 4hrs on either CPU.
Currently running both at the same time while I'm at work to see what happens.
I'm hoping to hear from some of you that have done this before, as I've seen the numerous screenshots of two Prime95 running with Smart Guardian on top of it. See how it worked for you guys
Unfortunately, i'm not in front of my machine at home but what I did was, I right clicked on the Prime95 Icon and created a copy of it to the desktop. On the original one, I right clicked and added -A0 ( that's an A and a zero) to the end of the target line. Then on the copy, I did the same thing except added -A1.
I opened up both of them and changed affinity to core 0 and the other to core 1. You also want to see the RAM settings too so each instance will only use so much RAM. For example, I have 1 Gig of RAM. So I assigned both instances 400 Megs of RAM each so all of my RAM isn't overloaded.
This works great for me. Once you get one instance going, the task manager should show 50% CPU usage. Once both of them are going, you'll see 100%
How do I set the amount of Ram to use for each process? Maybe that is what I'm doing wrong. That would probably make sense, though I can't believe on instance would use all 2GB of Ram I have. The only thing I know that does that is BF2
Anyway, I'll try that, but let me know how to do that :wink:
Yup that did the trick, I set each one to 900MB of ram, ran both at the same time for over 9hrs with no problems, not a big surprise since it is all stock speeds.
I do have one more question, under torture test which of the 3 options is the better one to run to test stablity? Small FFTs, In-place large FFTs, or Blend? I ran the 2nd one as it says maximum heat, that was something I wanted to see what it was.
Below pic of all the stats and stuff with temps. Again nothing fancy just practicing posting specs for when I actually do OC. Just noticed I took this after I stoped the test, the temp under load after 9hrs was 45C but the chipset was bout the same.