• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Awesome new CPU stress utility needs feedback!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Nanerbeet

Registered
Joined
Jan 28, 2004
Location
Utah
Hey guys,



I wrote this little utility a while ago called Toast which is what I used to test out my overclocking.

I've updated it specifically for Core 2 and I need to get some feedback!! With the testing I've done this little proggy can really heat up the CPU more than OCCT or Orthos or any program I've tested!!


Please download it and compare the temperature to what you're using, tell me if you think it gets hotter. Be careful, it might REALLY get it hot.


A buddy of mine hit 90 degrees C on his Core 2 Quad and the BIOS automatically shut off the machine :)



It runs a very tight SSE loop executing nearly 4 instructions per clock cycle and a lot of vector math. I won't report any errors, it'll just heat up the CPU real hot and crash if your cooling isn't up to the task.


http://damage.vigilantesoftware.com


Like I said, its not done yet, and I've been thinking about integrating a CPU overclocking database to it; the idea is anyone who wants can enable sharing and have it report to the database automatically. Stuff like that.


I don't have real good cooling, so mine runs pretty hot. Running my Core 2 Duo at 3500mhz, 1.525 volts, it jumps from 57 degrees (idle) to 81 degrees in just a few seconds. Then it crashes ;)


hightemp.png
 
Last edited:
I dont trust this on my current hold over computer, not the one in my sig, I post back when I get my main rig running again (currently crosstrading out the cpu)
 
The membership would probably feel more comfortable if you posted source code so that we can verify that you aren't doing anything malicious, particularly as you are a new user here.

Also, I'm moving this over to programming, as you seem to be soliciting feedback on a new benchmark that you're using. The users there will also be a bit savvier in trying out beta software from unknown sources. -- Paul

Users: Please note that I have not verified this software. Anybody who wants to download it, try it, and carefully monitor for spyware / malware / crashes is more than welcome to do so and report back here. -- Paul
 
File is clean, but not real impressed. yawn

Untitled-1.jpg

I got a PM from a member about this, that I was rude. This program provides nothing for me, nor would I use it, so I posted the above. I didn't mean to hurt anyone's feelings. The original post was reported so I figured I'd take a look. I yawned because it as boring, as it I do not need it, I will not use it, and there was nothing that needed to be resolved with the thread itself.

cw
 
Last edited:
sourcey? I recently started using SSE in my programs, and it was ...interesting to say the least. How do you handle it being as a pipeline? As in, what do you do with the rest of your cycles while it pipelines your requests?
Did you ASM the whole thing, or C with inline? Did you look at the -s result?
 
Ill give it try when ive got some better cooling.. im on stock atm and the system fans are s**t also so probably not best to spank the CPU yet...
 
sourcey? I recently started using SSE in my programs, and it was ...interesting to say the least. How do you handle it being as a pipeline? As in, what do you do with the rest of your cycles while it pipelines your requests?
Did you ASM the whole thing, or C with inline? Did you look at the -s result?


I'm using inline __asm code in my C++ programs. Visual Studio 2003.

The Core 2 Duo can handle two SSE instructions per clock cycle along with other ALU instructions. The trick is to write a tight loop and interlace ALU and SSE instructions like this:

Code:
	addps	xmm6, MathData + 48
	mulps	xmm1, MathData + 64
	xor	esi, esi
	fld      qword ptr [edi+10h] 
	fstp     qword ptr [edi+20h]


The instructions need to be aligned on 16-byte boundries and should be using memory operands that hit cache memory. As few bubbles in the pipeline as possible keeps the processor hot.
 
Interesting.

You should consider auto-checking the number of cores and adapting that 4-core processors as well. -- Paul
 
One sec, I'll put up a comparison.

Here is Prime95 stressing all cores, this is once the temps got up to speed, then I started monitoring:

Code:
Session start:    23:24:13, February 3, 2008
CPUID:        0x10676
Processor:    Intel Core 2 Extreme QX9650 (Yorkfield)
Revision:    C0
Time & Date        CPU Speed    
23:24:23 02/03/08    3999.91        65        65        59        61        
23:24:33 02/03/08    3999.91        65        65        59        61        
23:24:43 02/03/08    3999.91        65        65        59        61        
23:24:53 02/03/08    3999.91        63        62        60        61        
23:25:03 02/03/08    3999.91        65        65        59        62        
23:25:13 02/03/08    3999.91        65        65        59        62        
23:25:23 02/03/08    3999.91        65        65        59        62        
23:25:33 02/03/08    3999.91        66        66        60        61        
23:25:43 02/03/08    3999.91        66        66        59        62        
23:25:53 02/03/08    3999.91        65        65        59        62        
23:26:03 02/03/08    3999.91        65        65        60        62        
23:26:13 02/03/08    3999.91        65        65        59        62        
23:26:23 02/03/08    3999.91        65        65        59        62        
23:26:33 02/03/08    3999.91        65        65        59        62        
23:26:43 02/03/08    3999.91        65        65        59        62        
23:26:53 02/03/08    3999.91        65        65        59        61        
23:27:03 02/03/08    3999.91        65        65        59        61        
23:27:13 02/03/08    3999.91        65        65        59        62        
23:27:23 02/03/08    3999.91        65        65        60        62        
23:27:33 02/03/08    3999.91        65        65        59        62        
23:27:43 02/03/08    3999.91        66        66        60        62        
23:27:53 02/03/08    3999.91        66        66        60        62        
23:28:03 02/03/08    3999.91        65        65        60        62        
23:28:13 02/03/08    3999.91        66        66        59        62        
23:28:23 02/03/08    3999.91        65        65        59        62        
23:28:33 02/03/08    3999.91        65        65        59        62        
23:28:43 02/03/08    3999.91        66        66        59        62        
23:28:53 02/03/08    3999.91        64        64        59        62        
23:29:03 02/03/08    3999.91        66        66        60        62        
23:29:13 02/03/08    3999.91        65        65        59        62        
23:29:23 02/03/08    3999.91        65        65        59        61        
23:29:33 02/03/08    3999.91        65        65        59        62        
23:29:43 02/03/08    3999.91        65        65        59        62        
23:29:53 02/03/08    3999.91        65        65        59        62        
23:30:03 02/03/08    3999.91        65        65        59        62        
23:30:13 02/03/08    3999.91        65        65        59        62        
Session end:    23:30:23, February 3, 2008

New program in a sec.
 
Last edited:
WOW, is all I have to say. This puts the HURT on my processor:

Code:
Session start:    23:35:38, February 3, 2008
CPUID:        0x10676
Processor:    Intel Core 2 Extreme QX9650 (Yorkfield)
Revision:    C0
Time & Date        CPU Speed    
23:35:39 02/03/08    3999.91        76        73        69        74        
23:35:40 02/03/08    3999.91        75        74        69        74        
23:35:41 02/03/08    3999.91        76        73        68        74        
23:35:42 02/03/08    3999.91        76        73        69        74        
23:35:43 02/03/08    3999.91        77        73        69        73        
23:35:44 02/03/08    3999.91        76        73        68        74        
23:35:45 02/03/08    3999.91        77        73        67        74        
23:35:46 02/03/08    3999.91        72        73        68        74        
23:35:47 02/03/08    3999.91        76        73        69        74        
23:35:48 02/03/08    3999.91        76        73        69        74        
23:35:49 02/03/08    3999.91        77        74        69        74        
23:35:50 02/03/08    3999.91        76        73        69        73        
23:35:51 02/03/08    3999.91        76        73        69        73        
23:35:52 02/03/08    3999.91        77        74        69        74        
23:35:53 02/03/08    3999.91        76        74        69        73        
23:35:54 02/03/08    3999.91        77        72        69        74        
23:35:55 02/03/08    3999.91        76        74        69        73        
23:35:56 02/03/08    3999.91        76        74        69        73        
23:35:57 02/03/08    3999.91        76        74        69        73        
23:35:58 02/03/08    3999.91        75        73        69        74        
23:35:59 02/03/08    3999.91        76        74        68        74        
23:36:00 02/03/08    3999.91        77        74        68        74        
23:36:01 02/03/08    3999.91        76        73        69        73        
23:36:02 02/03/08    3999.91        76        73        69        74        
23:36:03 02/03/08    3999.91        77        73        68        74                
Session end:    23:36:03, February 3, 2008

Same room temp (probably lower, window is open) and same fan speed...just...WOW :eek:

All it needs a checking feature and I would use it all the time.
 
F? cuz you said room temp, and 70C room temp will kill you lol

btw *ULTRA POKE* for not responding =)

oh if i know you correctly your photochopping a pic.
 
Back