AMD Phenoms have a hardware bug! :(

2007-12-04 06:43:18

#1

GC_PaNzerFIN: Work and study @ Technical Uni; +528|6891|Finland

whole article here -> http://www.techreport.com/discussions.x/13724

This is seriously bad thing. I would suggest anyone who is thinking about buying a phenom to wait for the new revision to come out.

"To recap, the erratum is a chip-level issue involving the TLB logic for the L3 cache that can cause system hangs in specific circumstances. AMD has a fix for the problem in the works, but it degrades performance. AMD has stated publicly that the workaround can lower performance by as much as 10%, although one source characterized the performance hit to TR as 10-20%."

"Instead, AMD is preparing a hardware fix in the next revision of the chip, dubbed B3. Future revisions of the Phenom, including the planned Phenom 9700 model at 2.4GHz and the 9900 at 2.6GHz, will include the fix. AMD plans to replace the current Phenom 9500 and 9600 models with new 9550 and 9650 models, based on the B3 chip, as well. Saucier's best estimate for the arrival of B3 chips is "mid to late Q1" next year."

rumors say that AMD knew about the hardware bug BEFORE the launch and they still did it...

3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8

2007-12-04 06:53:49

#2

Freezer7Pro: I don't come here a lot anymore.; +1,447|6674|Winland

This is one of the reasons I don't like AMD...

The idea of any hi-fi system is to reproduce the source material as faithfully as possible, and to deliberately add distortion to everything you hear (due to amplifier deficiencies) because it sounds 'nice' is simply not high fidelity. If that is what you want to hear then there is no problem with that, but by adding so much additional material (by way of harmonics and intermodulation) you have a tailored sound system, not a hi-fi. - Rod Elliot, ESP

2007-12-04 06:55:21

#3

Bull3t: stephen brule; +83|6779

So much for a rival for the Q6600!

I was counting on AMD!

I am slowly but surely leaning over to Intel..

2007-12-04 06:56:31

#4

Stormscythe: Aiming for the head; +88|7026|EUtopia | Austria

Uhm... That's why they didn't launch 9700 and 9900. so... yeah, they knew about the bug before the launch, that's why they only launched the 9600 ^^

2007-12-04 06:58:33

#5

jsnipy: ...; +3,277|6999|...

so sad, they definately did not need this

2007-12-04 07:04:52

#6

Gawwad: My way or Haddaway!; +212|7162|Espoo, Finland

I hope they'll release a competitive product soon, or the prices of good processors might start rising even more...

Would love to see such competition that RAM manufacturers have.

2007-12-04 07:08:12

#7

max: Vela Incident; +1,652|7044|NYC / Hamburg

huggles q6600

_{once upon a midnight dreary, while i pron surfed, weak and weary, over many a strange and spurious site of ' hot xxx galore'. While i clicked my fav'rite bookmark, suddenly there came a warning, and my heart was filled with mourning, mourning for my dear amour, " 'Tis not possible!", i muttered, " give me back my free hardcore!"..... quoth the server, 404.}

2007-12-04 07:14:35

#8

mikkel: Member; +383|7078

This is just cheap journalism. You find very few CPUs that don't have engineering samples with bugs, and it's not too uncommon for processors to have bugs in their first steppings.

AMD clarified when information on this bug was released that it only happens under extremely high loads, and that they have been unable to recreate it at all in a normal user environment. I really wouldn't put too much into it. Get your processors before the fix!

Last edited by mikkel (2007-12-04 07:17:00)

2007-12-04 07:18:00

#9

GC_PaNzerFIN: Work and study @ Technical Uni; +528|6891|Finland

mikkel wrote:
This is just cheap journalism. You find very few CPUs that don't have engineering samples with bugs, and it's not too uncommon for processors to have bugs in their first steppings.

AMD clarified when information on this bug was released that it only happens under extremely high loads, and that they have been unable to recreate it at all in a normal user environment. I really wouldn't put too much into it. Get your processors before the fix!

I would. The fix drops performance about/more than 10%.... and that info is from amd....

Last edited by GC_PaNzerFIN (2007-12-04 07:19:33)

3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8

2007-12-04 07:19:42

#10

Bull3t: stephen brule; +83|6779

jsnipy wrote:
so sad, they definately did not need this

*9500

2007-12-04 07:21:46

#11

Ender2309: has joined the GOP; +470|7048|USA

GC_PaNzerFIN wrote:
mikkel wrote:
This is just cheap journalism. You find very few CPUs that don't have engineering samples with bugs, and it's not too uncommon for processors to have bugs in their first steppings.

AMD clarified when information on this bug was released that it only happens under extremely high loads, and that they have been unable to recreate it at all in a normal user environment. I really wouldn't put too much into it. Get your processors before the fix!
I would. The fix drops performance about/more than 10%.... and that info is from amd....

you don't have to get the fix (which is software). the newer chips don't have the problem via a workaround, if i read correctly.

2007-12-04 07:29:01

#12

GC_PaNzerFIN: Work and study @ Technical Uni; +528|6891|Finland

Ender2309 wrote:
you don't have to get the fix (which is software). the newer chips don't have the problem via a workaround, if i read correctly.

so fair to the guys that buy phenom now.... isn't it? while competitor doesn't have any probles at all and keeps selling Q6600 like crazy...

edit: It appears to affect the chips worse when overclocking (high frequenzies). now taht explains the delay of 9900 and 9600 black edition....

Last edited by GC_PaNzerFIN (2007-12-04 07:34:30)

3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8

2007-12-04 09:08:13

#13

ReTox: Member; +100|6976|State of RETOXification

Freezer7Pro wrote:
This is one of the reasons I don't like AMD...

All chip makers have had a problem similar to this at one time or another. Intel had its famous floating point bug that did math wrong and let's not even talk about Cyrix. Even the C2D-Extreme has issues, just ask server admins who have rack mounts that don't restart or shutdown properly. Still, it's never good for a chip maker to have this issue.

I am dismayed that they seem to have known about the bug and still released the chips.

2007-12-04 09:13:38

#14

mikkel: Member; +383|7078

Ender2309 wrote:
GC_PaNzerFIN wrote:
mikkel wrote:
This is just cheap journalism. You find very few CPUs that don't have engineering samples with bugs, and it's not too uncommon for processors to have bugs in their first steppings.

AMD clarified when information on this bug was released that it only happens under extremely high loads, and that they have been unable to recreate it at all in a normal user environment. I really wouldn't put too much into it. Get your processors before the fix!
I would. The fix drops performance about/more than 10%.... and that info is from amd....
you don't have to get the fix (which is software). the newer chips don't have the problem via a workaround, if i read correctly.

As far as I can read, the next generation of processors -is- the fix. This is supposedly an engineering error that cannot be fixed with any microcode, so it'll be purely hardware.

2007-12-04 09:20:30

#15

GC_PaNzerFIN: Work and study @ Technical Uni; +528|6891|Finland

The Bios fix for this will probably disable the L3 cache of the processor. that would explain the 10% performance decrease. Yep, it is harware and there is no way to fix it except disabling the part of the chip that is causing the problems. They'll fix it with the new revision to get the lost 10% back. but sure sucks for buyers of current Phenoms... Making processor slower after the release is something that I don't like to see.

3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8

2007-12-05 01:44:26

#16

Crypto_420: Member; +25|7320|Portland

AMD said, the bug only happen under EXTREME LOADS at high voltages. Tomshardware threw every test at it while it was overclocked to 3.0 and never had an issue. The reason why AMD is selling the models with this issue, NO home computer user could ever put enough stress on it to cause the problem.

Plus AMD has already planned a head, the Quad Cores with 1 bad core will be sold as the X3.

BF2S Forums AMD Phenoms have a hardware bug! :(

mikkel wrote:

jsnipy wrote:

GC_PaNzerFIN wrote:

mikkel wrote:

Ender2309 wrote:

Freezer7Pro wrote:

Ender2309 wrote:

GC_PaNzerFIN wrote:

mikkel wrote:

Board footer