Tool to change CPU string

[xigfrid]

Originally posted by VVJ
2rabeb25:
Of course! XP-Rating wasn’t determined because the program showed 128K L2 cache size. XP-Rating is calculated for the known processor cores: Palomino (256K), T-Bred (256K), Barton (512K) and for Thorton (256K) in the latest version. Does WCPUID show 128K L2 cache size as well? Thanks for cooperation!

Hi VVJ, I just pointed out what happened to rabed25's cpu :

Here i got one friend that had 64K of L2 from a AthlonXP 1700+ overclocked under a Prommy with 2.2V : my friend fried most of the L2 cache with subzero temperatures.

For the PR, i would like to see a benchmark that shows the barton winning other the normal XP with more cache , i got none that confirm the PR.

[rabeb25]

hmm so is that a reason why all my benchmarks as of late suck ass?

[rabeb25]

Originally posted by xigfrid
Hi VVJ, I just pointed out what happened to rabed25's cpu :

Here i got one friend that had 64K of L2 from a AthlonXP 1700+ overclocked under a Prommy with 2.2V : my friend fried most of the L2 cache with subzero temperatures.

For the PR, i would like to see a benchmark that shows the barton winning other the normal XP with more cache , i got none that confirm the PR.

would it happen to look like this..
before was my 2700, now here is another one of my 1700's that do 3ghz in the prommie.

[rabeb25]

here is my 2nd 1700+ that does 3gz in the prommie, finally a non broken one

[VVJ]

2TribesMan:

Maybe I've this problem because I'm running my 2000+ at 2600MHz?

Wow! Then because of this high frequency. I think you know L2 cache running at the full clock speed. Of course, you must not disable ECC checking for high frequencies. Thanks for note!

[NoOne]

Cool program
Ill test it on my 1700+

[]s

[BoGuS]

VVJ:
What happend to your download links ?, or am i missing something ?... the one postet is a beta release ?? and your website saids v3 final ?

nice program though...

BoGuS/

[saaya]

hmmm can i use this tool to make my cpu look like a 166fsb cpu?

[VVJ]

2NoOne:
Hi NoOne!

Cool program

Thanks! But it is no so good as it sounds. I found that CBId still can’t detect multiplier when it is set with jumpers. Other word, the content of the HWCR MSR isn’t changed when you set another multiplier with jumpers. But when you change it in BIOS the content of the HWCR is changed respectively and my program can detect it correctly. For the moment I don’t know how to solve this problem.

[VVJ]

BoGuS

What happend to your download links?

CBId 3.0 Final Release (August 26, 03): http://cbid.at.tut.by/files/cbid302.zip

[VVJ]

saaya

hmmm can i use this tool to make my cpu look like a 166fsb cpu?

No, because CBId isn’t a “hacker” tool.

[STEvil]

Yeah, if you are under 200fsb it doesnt calc right either...


Someone should graph this so you can calc easier 8-)

EDIT

Changing the L1 had no effect for me.

[VVJ]

2STEvil:

Yeah, if you are under 200fsb it doesnt calc right either...

It is unimportant 200MHz or 166MHz, or etc. The program can’t correctly determine the clock ratio on some motherboards. Especially, the problem is occurred on EP-8RDA/+. As for 200MHz please, check out a screenshot at http://pcpitstop.ibforums.com/index....0&t=21932&st=0

[saaya]

i really like the design of the gui btw!

[Spartacus]

Originally posted by VVJ
I found that CBId still can’t detect multiplier when it is set with jumpers. Other word, the content of the HWCR MSR isn’t changed when you set another multiplier with jumpers. But when you change it in BIOS the content of the HWCR is changed respectively and my program can detect it correctly. For the moment I don’t know how to solve this problem.

Just calculate the multiplier instead of reading it. That way you'll get the actual multiplier regardless of jumper settings.

Multiplier=CPUMhz/FSBMhz

Nice program!

[Scyphe]

a small feature request: to get rid of all fake screenshots, how about the same kind of checksum for the results. Thus, any faked values could easily be disregarded.

Isn't it possible to calculate PR-rating from frequencies alone?

Currently, BIOS etc. calculate model of chip just by the multiplier (at least on my A7N8X), ie. 12.5x = 2800+ no matter what FSB I set. The system then tells me that my faster-than-an-actual-3200+-barton is a 2800+ (while it's an overclockec 2500+ Barton).

Like making a small table of Frequencies and corresponding PR-ratings. It would need different tables for different models, but you understand what I mean. That would be a more appropriate approach. There's nothing that says it HAS to be calculated from multiplier*fsb since most people overclocking their systems don't use the standard 133/166/200Mhz FSB.

A Barton running @ 2.2Ghz = 3200+, 2.3Ghz = 3400+ etc.

[MrLavender]

Originally posted by Scyphe
Isn't it possible to calculate PR-rating from frequencies alone?

No. Both higher fsb and larger cache size will give a higher PR rating for the same MHz.

E.g. when the tbred b's first came out there were 2 variants of the 2600+, 12.5x166.66.. (2083MHz) and 16x133.33.. (2133MHz).

So (according to AMD) the difference between a 133 and 166 fsb is 50MHz of cpu speed, which is about 75 PR points.

Also compare the 2600+ (12.5x166) and the 2800+ Barton (12.5x166). Exactly the same fsb and MHz, but the extra cache is worth 200 PR points.

The trouble is, AMD screwed with the ratings when the 3000+ came out. That runs at 13x166 which is the same as the 2700+. So by rights it should be a 2900+, not a 3000+!

It all gets a bit too murky.

BTW, the general PR formula for a 133 fsb tbred b is Rating = 3 x Frequency/2 – 500. There are a lot of articles about PR Ratings at X-bit labs here, the first one gives what is supposed to be the new formula for calculating Athlon64 PR Ratings.

[VVJ]

Hi all!
Central Brain Identifier 3.50 has been already done. I’m looking for beta testers who have the mobile AMD processors: Athlon 4, mobile Duron, Athlon XP-M/XP-LV. I have realized VCC_Core detection for all mobile AMD processors, but I have no chance to test. If anybody would like to help me, please let me know by email jungle@mail15.com. Thanks!
P.S. The unofficial download link for beta testers is http://cbid.at.tut.by/files/cbid350b.zip
CBIdCfg.ini file has two new parameters.

[VVJ]

Daved+, ReelMonza, well this bug has been already fixed! But I haven't updated CBId yet.

[ReelMonza]

OK, thanks, I'll be looking forward to the next release !

Another thing, why is the Performance State Block Header/Performance State Tables all greyed out ?

[VVJ]

Hello there!

I need some volunteers to officially prove my assumptions regarding how to change a HyperTransport bus frequency and its width using VIA AMD64 specific chipsets. Well, I didn’t use any tricks to get a successful result, I only reread the HyperTransport bus Specification once again and made some conclusions. So, I’ve checked updated CBId 8.0t with my test ASUS K8V-MX K8M800+VT8237R Plus and the result is wonderful! No any reboots or freezes for me! The frequency and the bus width are changed even better than I need :-) Please, check it up!

Here is a screenshot to find appropriate options:

[VVJ]

Willis, Manchester is a city of Great Britain. Please, give me a CPUID(1) signature of your processor or its OPN number.

[Scyphe]

My Venice (ADA3200DAA4BW) shows DH-E6. And Willis, you must be talking about Winchester, not Manchester.. lol

[VVJ]

Just tested with Epox EP-9HEAI K8T890+VT8237R. Damn, works great too!

Scyphe, his post has confused me, because...

[lawrywild]

My Venice (ADA3200DAA4BW) shows DH-E6. And Willis, you must be talking about Winchester, not Manchester.. lol

nooo he's not lol