Phenom II X3-720BE List Capable to running 4 cores or not

Printable View

Show 100 post(s) from this thread on one page

03-02-2009, 04:02 AM
zorz

AMD Phenom II X3 720 BlackEdittion BOX, 0851 FPAW
Biostar TA790GXB A2+, BIOS 78DDA113.BST, Upload Date: 2009-01-13

http://www.benchmark.rs/forum/attach...1&d=1235989670 and http://www.benchmark.rs/forum/attach...1&d=1235989686

All Cores Successifuly Unlocked with little o/c without any voltage raising (for me, enough)!!!

"And wolfs feed up, and sheeps on the number."

;)
03-02-2009, 05:51 AM
fanello

Quote:

Originally Posted by Benny Lodewijk

* 0904 BPMW-------> Boot OK, detect as Phenom X4 and then stop, debug card shown 83 code.

My 0904 BPMW is running fine on 4 cores with Asus M3a78-t @ Bios mod
03-02-2009, 07:43 AM
Apoptosis

Quote:

Originally Posted by fanello

My 0904 BPMW is running fine on 4 cores with Asus M3a78-t @ Bios mod

What BIOS are you running and what settings did you have to change to get this to work on the M3A78-T? I have the same board and had no luck.
03-02-2009, 12:36 PM
dehx

Quote:

Originally Posted by fanello

My 0904 BPMW is running fine on 4 cores with Asus M3a78-t @ Bios mod

You can get the BIOS here..

http://www.xtremesystems.org/forums/...4&postcount=32
03-02-2009, 01:07 PM
socrilles

If I'm remembering correctly the last three numbers in the serial number (or last row on the chip) represent how close to the center of the silicon disc the chip was made, and previously people correlated OCs with how low that number was. I'm wondering if people could post this number in additon to the standard week and batch to see if it correlates with ability to unlock the fourth core. Of course I might have just imagined all this as well. However PCriminal chip which does well is a low number 066 I believe.
03-02-2009, 01:10 PM
Halk

What happens is this.

AMD design a product, the 4 core Phenom II. They also need a lower tier, cheaper product. They start making the 4 core Phenom II and when they test the CPUs if one of the cores cannot cope then it gets branded as a 3 core processor.

That's all nice and simple, and by that logic an X3 should rarely work as an X4, and perhaps it may need one of the cores to be downclocked.

However it's not just as simple as that. When AMD has too many X4s and too few X3s they will simply rebrand X4s as X3s. When all the single core Opteron 939 goodness was going on that's what had happened. AMD had rebranded superior cores they couldn't sell at superior prices as something lower.

There's also another possible outcome. AMD has too few X4s and too many X3s. All they can do there is to try to enhance the manufacturing process, but in the meantime the X3 price drops and/or the X4 price goes up.

And what does all that mean for people who are trying to get a bargain and pay an X3 price and get an X4 product? It's either down to a lot of research about steppings, weeks and other numbers (see the excellent thread Chew* made), or it's just pure luck.
03-02-2009, 03:37 PM
socrilles

Quote:

Originally Posted by Halk

What happens is this.

AMD design a product, the 4 core Phenom II. They also need a lower tier, cheaper product. They start making the 4 core Phenom II and when they test the CPUs if one of the cores cannot cope then it gets branded as a 3 core processor.

That's all nice and simple, and by that logic an X3 should rarely work as an X4, and perhaps it may need one of the cores to be downclocked.

However it's not just as simple as that. When AMD has too many X4s and too few X3s they will simply rebrand X4s as X3s. When all the single core Opteron 939 goodness was going on that's what had happened. AMD had rebranded superior cores they couldn't sell at superior prices as something lower.

There's also another possible outcome. AMD has too few X4s and too many X3s. All they can do there is to try to enhance the manufacturing process, but in the meantime the X3 price drops and/or the X4 price goes up.

And what does all that mean for people who are trying to get a bargain and pay an X3 price and get an X4 product? It's either down to a lot of research about steppings, weeks and other numbers (see the excellent thread Chew* made), or it's just pure luck.

I'm not sure what you were trying to get at, I think most of us knew that there is only two possibilities (and its probably a combination of these possibilities) one is AMD is binning chips to increase profits (ie all cores are good, but to have a diverse product lineup they disable some of the quad cores), and the other is that the yield of quad cores is weak and many have a defunct core and are sold as tri-cores, the question is what is going to correlate with success when unlocking cores which is what I was trying to get at, and I only know of three possible indicators - the week the chip was made, the stepping and those last three numbers, so far people have been posting the week and batch numbers but not the last three numbers.
03-02-2009, 04:49 PM
ah_khoo

can i ask a favor from all 0904bpmw users - to do a prime test with 8-4096k & w/o run fft in place? i'll get system lock up rite away after i click d start button but i can survive prime test with 8-512k, run fft in place (usually known as small fft a.k.a cpu test) bout 40 mins with mine. thanks in advance. :up:
03-02-2009, 09:42 PM
mngdew

Quote:

Originally Posted by ah_khoo

can i ask a favor from all 0904bpmw users - to do a prime test with 8-4096k & w/o run fft in place? i'll get system lock up rite away after i click d start button but i can survive prime test with 8-512k, run fft in place (usually known as small fft a.k.a cpu test) bout 40 mins with mine. thanks in advance. :up:

I've tried two 0904BPMW and one 0904CPCW. They all lock up my system immediately when I tried to run OCCT.:shrug:
03-02-2009, 10:14 PM
ah_khoo

Quote:

Originally Posted by mngdew

I've tried two 0904BPMW and one 0904CPCW. They all lock up my system immediately when I tried to run OCCT.:shrug:

it's becoz that's a blend test i supposed. try to run p95 w/ 8-512k, run fft in place then i'm pretty sure u can last for sometimes. :)
03-02-2009, 10:15 PM
Shrekogre

Quote:

Originally Posted by ah_khoo

can i ask a favor from all 0904bpmw users - to do a prime test with 8-4096k & w/o run fft in place? i'll get system lock up rite away after i click d start button but i can survive prime test with 8-512k, run fft in place (usually known as small fft a.k.a cpu test) bout 40 mins with mine. thanks in advance. :up:

mine runs fine. did u raise ur nb clocks too? is this on the dfi or biostar?
03-02-2009, 10:20 PM
ah_khoo

Quote:

Originally Posted by Shrekogre

mine runs fine. did u raise ur nb clocks too? is this on the dfi or biostar?

yours blend test ok too? no, i leave nb & ht related thingy to auto except d voltages & d cpu multiplier... :)

i only have 790fxb w/ me now. ;)
03-02-2009, 11:13 PM
mngdew

Quote:

Originally Posted by ah_khoo

it's becoz that's a blend test i supposed. try to run p95 w/ 8-512k, run fft in place then i'm pretty sure u can last for sometimes. :)

But then my system is not really stable. I don't want some rigged test that will give me a fake result. I want blind test for my system stability.:yepp:

You see my sig below. This system setup is OCCT, Linkpak and Prime 95 torture test stable for 24 hours, and I expect the same from AMD.
03-02-2009, 11:22 PM
ah_khoo

Quote:

Originally Posted by mngdew

But then my system is not really stable. I don't want some rigged test that will give me a fake result. I want blind test for my system stability.:yepp:

You see my sig below. This system setup is OCCT, Linkpak and Prime 95 torture test stable for 24 hours, and I expect the same from AMD.

well, if stability is d main concern, d 4th core shouldn't be unlocked. take it as it is - tri-core. one will appreciate d oc offered by phII more if he/she was with agena b4. :D
03-03-2009, 12:21 AM
fanello

Quote:

Originally Posted by ah_khoo

can i ask a favor from all 0904bpmw users - to do a prime test with 8-4096k & w/o run fft in place? i'll get system lock up rite away after i click d start button but i can survive prime test with 8-512k, run fft in place (usually known as small fft a.k.a cpu test) bout 40 mins with mine. thanks in advance. :up:

I've got the same problem, I don't know what the problem is.
03-03-2009, 12:31 AM
ah_khoo

Quote:

Originally Posted by fanello

I've got the same problem, I don't know what the problem is.

guess that's what they call crippled core. but i'm not giving up yet. :D
03-03-2009, 08:37 AM
fanello

Quote:

Originally Posted by ah_khoo

guess that's what they call crippled core. but i'm not giving up yet. :D

Turn down the Multi to 5 or 6 and then test again. If the same error occurs, it's probably defect.
03-03-2009, 08:56 PM
socrilles

0904 EPMW 533

Installing windows 7 on quad core 720 BE at 3.4 ghz right now :) - of course this is no indication of stability
03-03-2009, 09:32 PM
ah_khoo

Quote:

Originally Posted by fanello

Turn down the Multi to 5 or 6 and then test again. If the same error occurs, it's probably defect.

what amazed me was i was able to blend test stable core 3 (4th core) for hours. could it be d unlocked core is not core 3? :confused:
03-03-2009, 09:52 PM
largon

Quote:

Originally Posted by ah_khoo

what amazed me was i was able to blend test stable core 3 (4th core) for hours. could it be d unlocked core is not core 3?

The disabled core can be any of the four. It could be any of the core designations in Task Manager eg. ("CPU 0/1/2/3") once it is enabled.
03-03-2009, 09:58 PM
ah_khoo

Quote:

Originally Posted by largon

The disabled core can be any of the four. It could be any of the core designations in Task Manager eg. ("CPU 0/1/2/3") once it is enabled.

so it will change everytime i reboot my pc? if that's d case i can only trust AOD (same as what set in bios) to determine which is the crippled core. :shakes:
03-03-2009, 11:15 PM
largon

Change everytime you reboot?
No...

That would mean OS would randomly number the cores, which is not the case.
03-03-2009, 11:33 PM
ah_khoo

Quote:

Originally Posted by largon

Change everytime you reboot?
No...

That would mean OS would randomly number the cores, which is not the case.

i think one of d 4 cores need to be set diff value of acc. :)

what i'll do is set acc 0% to all core, prime (small fft test) w/ stock clock & voltage by runnin 4 instances of orthos (w/ each instance assigned to one particular cpu) or i need to run only one instance of orthos only??? d crippled core would be d one which fail d test. :yepp:

d thing is, how do i know d crippled core, say we name it as core X (in task manager), is labelled as which core in bios? :shrug:
03-04-2009, 12:26 AM
largon

Since you are interested in finding what core, if any, is unstable you, of course, need to run 4 stress test threads each with affinity to a single core. The one that (likely) fails is the crippled core, this core number is the same as in bios, as long as your bios counts them as 0, 1, 2 and 3.

The thing I don't get is why do you even care about the number of the crippled core?
03-04-2009, 12:37 AM
ah_khoo

Quote:

Originally Posted by largon

Since you are interested in finding what core, if any, is unstable you, of course, need to run 4 stress test threads each with affinity to a single core. The one that (likely) fails is the crippled core, this core number is the same as in bios, as long as your bios counts them as 0, 1, 2 and 3.

The thing I don't get is why do you even care about the number of the crippled core?

pardon me for my bad english mate, i never meant d bold part. :)

will try your suggestion tonite, thx... :up:

Show 100 post(s) from this thread on one page