RSSAMD quad-core 'erratum' creates problems for early adopters
By Scott M. Fulton, III | Published December 6, 2007, 3:05 PM
A "design and process tuning step" is being blamed for the first quad-core AMD processors at 2.4 GHz being shipped with a BIOS fix, containing a workaround for what now appears to be a serious erratum.
An AMD spokesperson told BetaNews this afternoon that the first wave of its quad-core Opteron server CPUs and Phenom desktop CPUs were shipped on the November 19 launch date with a known erratum -- a documented bug. Customers received CPUs along with a BIOS fix that includes a workaround.
"Respective to Quad-Core AMD Opteron," AMD's Phil Hughes told BetaNews, "we are only shipping processors earmarked for specific end-user installations where customers have had the opportunity to validate the stability and robustness of the solution where it leverages the BIOS fix or some other potential software workarounds. Quad Core AMD Opteron processors shipping for general availability in the Q108 timeframe will not have this erratum."
Hughes acknowledged shipment delays for the latest wave of Barcelona architecture CPUs, but reiterated earlier comments made by AMD executives that the company continues to plan to ship "hundreds of thousands of quad-core processors" during the fourth quarter of the year...a quarter that is now three weeks from being over.
The problem, Hughes confirmed, was first brought to light on Monday by The Tech Report, and concerns a critical element of the CPU called the transaction lookaside buffer (TLB). Here's what it does: In modern computer environments, each application thinks it has the entire memory space all to itself. Virtual memory enables this application to "think" it has 4 GB of unencumbered address space, while the CPU is busy translating the addresses the application presents into real addresses in physical memory.
To do that, the operating system presents the CPU with a virtual memory handler. But modern CPUs bypass this handler, using a faster local cache where it can buffer translated addresses in advance. This is the TLB, which for AMD quad-cores is located in the L3 cache. The TLB is designed for failover, so in the event of a "TLB miss" (when the table doesn't translate correctly), the OS virtual memory handler is called in for backup.
Apparently, a certain level of stress which AMD's own engineers did not test for, causes more TLB misses than usual. And though AMD has yet to say so explicitly, it would appear its BIOS fix could be turning off the TLB altogether...which could explain what many hardware enthusiast sites are reporting to be a 10% performance hit compared to unpatched BIOS.
Despite the apparent setback, Hughes said his company is sticking by its projections of shipping "hundreds of thousands" of quad-core processors this quarter.
"Quad Core AMD Opteron processor is the most advanced x86 processor ever introduced to the market," he stated, "and as such there are design and process tuning steps that have taken longer than expected."
I bought a Quad FX board which they now aren't going to support >< Anyone know if the quad code Opteron CPUs will work in place (same socket). I originaly bought it "AMD" saying it would be upgradeable and now they are droping the Quad FX brand.
Score: 0
|Remember when the Opterons were first released, 1 million of them were faulty and there was a massive recall. That was a DAH moment for AMD.
It’s good that this time AMD caught the error upfront. Or should we say IBM? Most of the Opteron technology concept is an IBM technology manufactured under license by AMD.
I don’t know what else to say about AMD and I’m no longer holding my breath for them. That’s because the problems they have start from the top management and no clean up was done in that area.
Score: 0
|I do hope AMD does not ship their newest product with this huge design flaw.
Of course if they already started on production I woudn't be shocked if they did.
Score: 0
|Red Hat is already distributing a quirk (workaround) patch for Linux, which supposedly has a negligable performance hit. It wouldn't surprise me if Microsoft wrote a patch for the Windows kernel with the same purpose shortly.
Score: 0
|come on people easy on the AMD slamming, lets not forget that it's AMD that has caused INTEL to smart up, and over the last seven years INTEL has,
granted AMD has slipped up here without question, but the CPU isn't in mainstream distrubutuion yet, and it comes a time when a product has to be tested in the field,
if one wants to compare INTEL to AMD, then it clearly is no competition yet, INTEL has made many many mistakes, and promises that were never delivered, i think we can all remember the P4 propaganda, how many revisions did this chip go thought in the end, 6 was it,
and lets also not forget that at the time, INTEL claimed that the P4 will eventually be able to breach the 10GHz barrier, it wasn't until AMD came along and said that would never be possible with today's technology, and AMD was right,
granted AMD will have to fix this problem, but i'm sure they will, and hopefully, they won't give up on the L3 cache, as this looks to be promising technology,
SONY being mentioned is a misconception at best, if anything INTEL would be SONY, lies, creates technology that's suppose to be better then it actually is, and doesn't know the meaning of efficiency (Although thanks to AMD Intel are starting to learn about efficiency now, and also learning to reduce the CPU thermal heat)
competition is always a good thing for the consumer,
as a consumer we need them both, if one falls, then the other will become complacent and would not need to push the boundaries of the technology, but with both companies trying to create a better CPU, it makes them both more innovative, and a lot lest complacent,
after AMD launched there Athlon XP CPU, Intel increased there R&D budget by a whopping 62%, with out AMD, this would of never happened,
but of course the same applies the other way round to, we need both of these companies to exist, or we will be the ones that will loose.
Score: 0
|Early Adopters get f*cked once and again and again, and don't seem to get enough of it. And i bet they are mostly the same reincident people of always.
Score: 0
|This is just sad. AMD is quickly becoming the Sony of the CPU industry. Delays, mistakes, and embarrassments.
Score: 0
|Lesse, Sony: Market cap is 10 times AMD's.
Sony: Owns major entertainment media companies well beyond tech.
Sony: profitable. Stock up 40% in past year. AMD: bleeding cash, stock down 56%...
Yeah I can see the similarities.
Score: 0
|That's a bit harsh. The comparsion was about delays, it was a simile. The didn't say AMD are Sony.
Score: 0
|so advanced, it's too advanced for its own pants.
Score: 0
|Let's see. How can we blame Intel for this one?
Score: 0
|Pressure ... Applied in liberal amounts to every market your in.
Score: 0
|Ummm...so, basically, for the early adopters there's no way to reclaim that performance. Nice going as always, AMD (Another Marketing Disaster).
Score: 0
|"we are only shipping processors earmarked for specific end-user installations where customers have had the opportunity to validate the stability and robustness of the solution where it leverages the BIOS fix or some other potential software workarounds."
Money ... I bet they were cheap, but they could have renamed them SX a la Intel.
Score: 0
|not a bad idea. mark the duds as slower chips. wasnt that the plan for the 3 core chips? not sure
Score: 0
|Im sure it played a part. Wasn't the 486SX a failed Mathcopro DX as well? Im sure the 487SX was actually a 486DX that disabled the SX
Score: 0
|The 486SX was not a FAILED 486DX, it was a deliberately disabled version.
Score: 0
|"486SX was not a FAILED 486DX, it was a deliberately disabled version..."
Hee hee hee, did you actually just say that? I used to "repair" motherboards for a fairly successful computer manufacturer....I can assure you that you are correct! Kinda..
They most certainly were "deliberately" disabled...after they FAILED as DX chips! Everyone did it at the time, Cyrix, AMD, and
Intel...but Intel came up with it (kinda) first.
There is some argument that the transition from 386 with an external mathco to a 486 with an integrated mathco didn't go as well as planned.
Let's not forget that Intel is not without it's "shipping with errors" history...but with the marketing they have they can survive...Love AMD (for competitions sake if nothing else), but they need some more marketing $$ focus or they are going to be in a LOT more trouble than they are now!
Just my opinion...
Score: 0
|3 core chip? Cool. Is there such a thing?
Score: 0
|Yes, a triple-core Phenom.
http://arstechnica.com/n...he-tri-core-phenom.html
It was covered here as well.
http://www.betanews.com/...ply&reply_to=972784
Score: 0
|