P4 1.8A / P4T-E Benchmark Results

[Toast]

Greetings,

I have completed the first phase of benchmarking my P4 1.8A on an Asus P4T-E. The effort was constrained to a maximum of 133FSB with a constant vCore of 1.525V across all frequencies. The attached graphs illustrate the results, and provide extrapolations to much higher frequencies. Based on the low voltage need at 133FSB, the cpu appears to have significant potential to be driven much further.

All the benchmark results were obtained through Sandra 2002.

Toast

The following graphs have been updated with the additional information that was graciously provided by rekless and antmanmike. Thanks for the data !

Toast

 

[Toast]

Sandra 2002 CPU Performance Benchamarks

 

[Toast]

Sandra 2002 Memory Benchmarks

 

[Teken]

Could you provide me a link to obtain this Sandra benchmark program...

Regards

Teken

 

[Toast]

Teken,

The following link should help you out.
http://www.sisoftware.demon.co.uk/sandra/

Toast

 

[AntmanMike]

Can u test it 2600+ mhz? I hear that at those speeds, the speed starts to lower, so technically, it should sooner or later turn into a parabola, or something similar (maybe just a curve?). Currently it looks like a linear equation, but i dont think that it is infinitely linear.

 

[Toast]

If the Pentium 4 was capable of operating at ever increasing frequencies, the benchmarks would continue to scale linearly with core frequency. While the Sandra benchmarks appear to be independent responses, other benchmarks are highly coupled between ALU, FPU, and MEM performance with hidden boundary constraints.

Toast

 

[funnyperson1]

yes...only in real world benchies would ytou see the scaling stop...

 

[Toast]

If anyone has data beyond 2.4GHz with the same setup (1.8A/P4T-E), let me know so that I can update the benchmark graph.

Toast

 

[Teken]

Help...

Toast,

Thank you for the supplied link... I unzipped the Sandra file and have 19 files... I opened the obvious set-up exe. and install files with no success... Can you take a few moments and tell me what specific file needs to be opened to execute this program to run and operate...

Regards

Teken<---Who is a little lost right now...

 

[Toast]

Teken,

It sounds like you downloaded the right collections. Once you have extracted all of the files into a folder, run the SETUP.EXE application, and follow the instructions. Make sure you view all the file details to verify the selected file type is an application.

You may want to verify that the version you downloaded is intended for you operating system and configuration.

Toast

 

[rekless]

Hey Toast,

2592Mhz: 4813/3101

and here is 2664Mhz:

 

[rekless]

Mulitmedia:

2592MHz: 9747/11817

and here is 2664MHz:

 

[rekless]

Memory (3x turbo)

2592MHz: 2961/2960

and 2664MHz:

 

[Toast]

thanks dude

Rekless,

Thank you very much for the extra data. Ultimately, I am hoping to generate a simmilar collection for most of the Northwoods. Given the number of overclockers on this forum, there should be ample information to generate the contours.
Thanks,

Toast

 

[rekless]

CORRECTION!

I made a mistake there Toast. Those values were for 2484MHz.

2566MHz: 10032/ 12213

When I did these benchmarks I must have overwritten one of the files cause I no longer have the scrn shot of multi media @ 2592.

I have since reduced my core voltage and FSB. So I cannot run another one @ 144 at this time. But I bet you could tell me what it should be.

Sorry for that screwup.


rek

 

[Toast]

rek,

No problem. I will update the values during the next edit.

Toast

 

[chop]

Mulitmedia:

2592MHz: 9747/11817

and here is 2664MHz:

Hey Reckless, how do you attach your results like that? Do you need the full version of Sisoft to attach the screenshot? Thanks.

 

[rekless]

I just use the prnt scrn function (windows) and then edit in a paint proggy. Has nothing to do with sandra.

This message board has the option of adding an image file to your message.

Thats it, nothing special.

rek

 

[Toast]

refreshed graphs