| View previous topic :: View next topic |
| Author |
Message |
Neon

Joined: 04 Feb 2008 Posts: 1512 Location: Dallas, Texas, USA
|
Posted: Thu Dec 31, 2009 2:18 pm Post subject: CPUID varies with multiplier setting in AMD K5 |
|
|
I bought a AMD-K5-PR200ABX processor last week. While testing, I found that CPUID changes depending on the multiplier setting. Previously, I had assumed that CPUID is "hard-coded" in the processor such that the same result is returned upon query, independent of the processor speed. This seems to be false, according to these results:
| Code: |
Motherboard setting BIOS report Internal frequency CPUID
2.0x 66 = 133 PR133 100 514
2.5x 66 = 166 PR166 117 524
3.0x 66 = 200 PR200 133 534
2.0x 75 = 150 PR150 112 514
2.5x 75 = 187 125MHz 131 524
2.0x 83 = 166 125MHz 125 514
|
I queried CPUID in Windows 2000 with AMD's CPUInfo utility, CrystalCPUID, CPU-Z, Lavalys Everest; and from Linux at /proc/cpuinfo. All programs agreed on the results.
I also tested a AMD-K5-PR150ABR, which demonstrated the same results, except it cannot boot at either 3.0x 66 or 2.5x 75 MHz. |
|
| Back to top |
|
 |
Neon_WA

Joined: 08 Nov 2008 Posts: 7146 Location: Margaret River, West Australia
|
Posted: Thu Dec 31, 2009 10:02 pm Post subject: |
|
|
Well done at finding this info out  _________________ There are 10 types of people in this world:
those who understand binary and those who don't. ~Author Unknown
http://www.x86-guide.net/Neon-WA/en/collection.html |
|
| Back to top |
|
 |
Neon

Joined: 04 Feb 2008 Posts: 1512 Location: Dallas, Texas, USA
|
Posted: Thu Dec 31, 2009 11:24 pm Post subject: |
|
|
I know that certain default CPUID data can be overwritten by the system BIOS at boot time, but I did not think that the basic family, model, stepping data would get altered. I am not smart enough to know whether this is:
a) just a quirk of my particular motherboard BIOS, or
b) a general property of the K5 processors.
It would be good to hear about others' experiences with these processors. |
|
| Back to top |
|
 |
Elar
Joined: 16 Jul 2003 Posts: 746 Location: Tallinn, Estonia
|
Posted: Fri Jan 01, 2010 5:30 am Post subject: |
|
|
| From K5 datasheet, PR200 not shown, but fits in as model 3 with 2x multiplier and BF pins at "11" |
|
| Back to top |
|
 |
6a6ar09a

Joined: 14 Oct 2009 Posts: 444 Location: Belgrade
|
|
| Back to top |
|
 |
Xaar

Joined: 30 Sep 2009 Posts: 319 Location: Germany
|
Posted: Fri Jan 01, 2010 9:50 am Post subject: |
|
|
Hello!
That's a good Information!
| Neon wrote: | | It would be good to hear about others' experiences with these processors. |
I have similar experiences with my K5 processors. I own: a PR133ABR, a PR150ABQ, a PR150ABR and a PR166ABR and also 4 SSA/5 5k86/K5 processors (whose CPUIDs are 0500h (5k86) or 0501h (K5)).
The PR166ABR's CPUID is detected as 0524h, but my PR150ABR as 0514h while running at 66x1.5 (or even 66x2.0; my boards detect him both as a PR133) and as 0524h while running at 66x2.5 (detected as a PR166). The PR150ABQ only runs at 66x1.5 or 66x2.0 (detected as PR133) and then has a CPUID of 0514h. Also the PR133 (running at 66x2.0 or 66x1.5) has a CPUID of 0514h.
By the way: What would the real setting for a K5-PR150 be? What is he really clocked? There is written on the CPU, that he has a bus frequency of 66 MHz.. _________________ Sorry for my bad English - I'm working on it!
My tradelist ( Updated 2013-12-20 ) |
|
| Back to top |
|
 |
Neon

Joined: 04 Feb 2008 Posts: 1512 Location: Dallas, Texas, USA
|
Posted: Fri Jan 01, 2010 11:56 am Post subject: |
|
|
OK, good! Your results for PR150ABR agree with my results.
The PR150 is a strange animal, and apparently AMD did not publicly document its specification. There is conflicting information on websites. Some sites claim that PR150 settings are 2.5x 60MHz = 150MHz, with internal frequency 105MHz. This is logical because it would complete a set of 3 analogous pairs of products that use 60MHz/66MHz bus speeds: PR90/PR100, PR120/PR133, and PR150/PR166.
However, as you noted, the PR150 (all that I have seen) are marked 66MHz bus. So ???
I speculate:
a) From a manufacturing point of view, PR150 is identical to PR166, and it performs the same;
b) From a product point of view, customers who have motherboard with maximum bus speed 60MHz will want PR150 because it is the fastest AMD upgrade possible. |
|
| Back to top |
|
 |
Neon

Joined: 04 Feb 2008 Posts: 1512 Location: Dallas, Texas, USA
|
Posted: Fri Jan 01, 2010 2:26 pm Post subject: |
|
|
I just completed some additional testing with K5-PR166ABR and K5-PR133ABR.
For the PR166ABR, the results are exactly the same as above for the PR200ABX. This includes motherboard setting 3.0x 66 = 200 (internal frequency 133MHz, a 16MHz overclock of nominal 117MHz). At this setting, PR166 has CPUID 534.
For the PR133ABR, at motherboard setting 2.5x 66 = 166 (internal frequency 117, a 17MHz overclock of nominal 100MHz, woo-hoo!). At this setting, PR133 has CPUID 524.
In summary, for K5-PR200ABX, PR166ABR, PR150ABR, and PR133ABR processors, the CPUID value returned depends only on the motherboard multiplier setting:
2.0x --> CPUID 514
2.5x --> CPUID 524
3.0x --> CPUID 534 |
|
| Back to top |
|
 |
Xaar

Joined: 30 Sep 2009 Posts: 319 Location: Germany
|
Posted: Fri Jan 01, 2010 2:54 pm Post subject: |
|
|
| Neon wrote: | | a) From a manufacturing point of view, PR150 is identical to PR166, and it performs the same; |
Well, I read some articles, where there is written that the PR150 is nearly identical to the PR133 but a little bit faster.
While reading the data sheet of the K5 processor, I found out, that the logical level on the pins BF0 and BF1 during the reset is responsible for on which multiplier the cpu runs. The following states and multipliers are mentioned on the data sheet:
| Code: | BF1-BF0 - Multiplier
1-1 - 1.5
1-0 - 1.5
0-0 - 1.75
0-1 - reserved <- that one could be the multiplier of the PR200 |
It depents also on these logic levels which Model ID the processor has (see list posted by Elar). For the reserved one the Model ID could be 3.
I found the datasheet here: Link _________________ Sorry for my bad English - I'm working on it!
My tradelist ( Updated 2013-12-20 ) |
|
| Back to top |
|
 |
gshv

Joined: 01 Feb 2003 Posts: 7898 Location: Fairfax, VA USA
|
|
| Back to top |
|
 |
Neon

Joined: 04 Feb 2008 Posts: 1512 Location: Dallas, Texas, USA
|
Posted: Fri Jan 01, 2010 5:28 pm Post subject: |
|
|
The datasheets explain a lot - I should probably learn to read before posting my "discoveries". If I understand correctly, the motherboard multiplier determines the signals to model 1, 2, or 3 CPU pins BF0 and BF1. This determines the processor internal multiplier, and that determines the CPUID model value.
| Code: |
Motherboard BF1-BF0 Multiplier CPUID P-rating CPU frequency
1.5x 1-1 1.5x 51x PR120 90
PR133 100
2.0x 1-0 1.5x 51x PR120 90
PR133 100
2.5x 0-0 1.75x 52x PR150 105
PR166 117
3.0x 0-1 2.0x 53x PR200 133
|
|
|
| Back to top |
|
 |
|
|
|
|
|
You can post new topics in this forum You can reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum
|
|
Powered by phpBB © 2001 phpBB Group
|