AMD Processors
Decrease font size
Increase font size
Topic Title: What's is the problem with Phenom??
Topic Summary:
Created On: 12/17/2007 05:09 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 12/17/2007 05:09 PM
User is offline View Users Profile Print this message

Author Icon
caracol
Junior Member

Posts: 4
Joined: 12/17/2007

Hello all, first my name is carlos i'm from Portugal and i'm writing here in the forum because in the web we can read lots of things about the phenom processors but i think the best place to ask some questions its here. ok i think ppl allready saw tests about the phenoms processors. my question is what is happning with this processors, they have architecture to give more perfomance that we can see in tests.
I hope ppl can explain me and other ppl that have the same question about it.
Thanks all

Edited: 12/17/2007 at 05:12 PM by caracol
 12/18/2007 11:34 AM
User is offline View Users Profile Print this message

Author Icon
Decembermouse
Mad Scientist

Posts: 2149
Joined: 03/31/2004

I heard that it has something to do with the L3 cache, but that's all I know. Yup, here it is, from Wikipedia:

"Before Phenom's release, it was discovered that a bug in the L3 cache in AMD's quad-core processors can cause a system lockup in rare circumstances. Phenom is affected by this bug up to and including stepping B2. There exist software workarounds to prevent this problem from occurring, although the workarounds typically employed incur a considerable performance penalty. This penalty was not accounted for in early pre-release previews of Phenom, so the performance of early Phenoms delivered to customers is expected to be somewhat below the benchmarks in these previews. Updated CPUs with the bug fixed will be named Phenom 9550 and 9650, and are supposed to arrive "mid to late Q1 2008." "

-------------------------
ASRock M3A785GMH/128M
Phenom II X4 925c2 @ 2.80GHz (95W)
OCZ Platinum PC3-1333 2x2GB 7-7-7-21
WD5000AADS
Raidmax RX-500AF
Check it out.
 12/18/2007 03:32 PM
User is offline View Users Profile Print this message

Author Icon
Overmind
Assimilator

Posts: 8052
Joined: 01/22/2004

The problem is more of marketing than a real visible problem. There is an eror on one of the CPUs registries, but the probability of it occuring is very low and requires very specific circumstances.

Let's not forget intel q6600 even B3 revision had about 30 known issues that were later solved with G0 stepping. And I did not hear anyone complain about it.

AMD could of certainly played this one a 'blame the OS thing' and I have the outmost respect for them because they didn't, even if the problem is software bypassable.

-------------------------

World's best Red Alert 2: Yuri's Revenge mod and Star Trek: Starfleet Command 3 mod: Overmind.ro
 12/19/2007 06:29 AM
User is offline View Users Profile Print this message

Author Icon
Rhythmeister
Junior Member

Posts: 4
Joined: 12/19/2007

"...Considerable performance penalty..." nothing, that's sum Intel rep' writing lies about AMD again. I read from the Inquirer (almost impartial!) that it'll be about 5-10% performance hit IF that. You know they're gonna be good so start saving!

-------------------------
FX-62 @3.0GHz
TF7025-M2
Club 3D HD3870
2GB HyperX (3, 3, 3, 10 1T at 2.35V!)
Silverstone ST56F
Lanbox lite
...'nuff said ;)
 12/19/2007 08:14 AM
User is offline View Users Profile Print this message

Author Icon
caracol
Junior Member

Posts: 4
Joined: 12/17/2007

well i dont care if they have less performance now i think amd will fix this bugs and kick some intel processors *****. i'm allready waiting for my 9700 and my hd3870
ty all
 12/19/2007 10:04 AM
User is offline View Users Profile Print this message

Author Icon
Haasonline
Junior Member

Posts: 3
Joined: 11/30/2007

The problem with the Phenom is a TLB bug which has his effect on the L3 cache. Due this bug the L3 can not run at full speed (still 1GHz, future 2,6GHz). There is a BIOS TLB work-around availible that will decrease the preformance by 5% up to 10% over the entire core. The B3 stepping will have a other fix that will have a preformance decrease of 1% as the rumors say.

Overall the core speed of the Phenom can be from stock up to 3,2 GHz stable without the BIOS fix. AMD has told the coming Black Edition will be the B3 stepping and will be able to show the real power of the K10 (K8L) core. So there is nothing to worry about. Also rumors tells us that AMD is already producing the B3 stepping from the beginning of December.

Back to the bug: Well the effect of the bug itself is low, sometimes you can get a BSOD if the core is running at 100% for 14 hours. For a normal user this bug will not be visible, but for important users like servers this is a problem so they apply the fix. If you ask me i think the entire talk about this bug is overhyped by all.

-------------------------
Athlon X2 3800+ Toledo 2x512kb @ 2200Mhz | Asus A8N-E | 2x 512Mb DC DDR400 @ 220MHz | Asus X1600Pro 256Mb | Raptor 74Gb 10.000 toeren | Stultorum infinitus numerus est
Statistics
112018 users are registered to the AMD Processors forum.
There are currently 0 users logged in.

FuseTalk Hosting Executive Plan v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.



Contact AMD Terms and Conditions ©2007 Advanced Micro Devices, Inc. Privacy Trademark information