c0t0d0s0.org

Lostcircuits published a really interesting article about AMD´s Phenom processors. Really worth a read - AMD's Phenom Processor - Beyond Erratum 298. It explains the Erratum 298 (the TLB bug) and makes some interesting statements about the performance:

Arguably, there is that thing called legacy support and legacy applications are still prevailing in today’s computing environment. On the other hand, there are truly multithreaded applications and as it turned out, the more thread level parallelism is used in any application, the better fared the Phenom processor. There is a bit more to this. Multicore processing is fine and dandy as long as the access to the data is warranted. Otherwise, there is starvation of the cores.

With a highly threaded application, the Pheonom can compete with a higher clocked Core2-based processor. On single thread load, the Phenom has problems to be as performant as the Intel products. This situation sounds familiar to me ...

Some interesting point about the IBM/Solaris announcement at "IBM and Sun, sitting in a tree, S-O-L-A-R-I-S-I-N-G:

Honestly? I think it is one of the few times I’d say “win-win” applies.

Times start to get more interesting again in the epic battle between Intel and AMD The Inquirer reports about performance promises of AMD

These figures were from a few months ago, so things may well have changed since then. But in the Integer test, a Barcelona 2.3GHz yields 21% higher score than Clovertown 2.66 GHz, but Floating Point test leaves a staggering 50% performance deficit for Clovertown, and this is not something 45 nanometre Penryn can solve overnight. Unless, of course the clock deficit for AMD is such that Intel speeds past.

There are some wild speculation in the usual suspicious forums about some SETI@HOME statistics of a mysterious AMD processor.

Nice comment about 80 core Intel at the register:

Terry Shannon will be turning in his grave. All X86 processors are really Alpha chips now. Pity it took AMD and Intel 10 years to catch up. And it seems DEC was quite right when it introduced the Alpha processor and said it would be good until the year 2025.

In my opinion, the Alphacide was the baddest thing that happened to the computer industry. The computing industry would look a whole different now with it.

PS: Now it´s up to Sun with Rock/Niagara/NiagaraII to bring some new concepts into computing.

While reading Scientias good article about AMD Q4 earnings, i thought a little bit about the situation in the x86 market:

AMD touts the quad core K8L, Intel speaks about the 45 nm Core Duo, both tells the world that they are faster than the other. And both is correct for certain timeframes. Sounds familar. Yeah, it´s leapfrogging. When two architectures are rather similar (UltraSPARC and Power, AMD x64 or Intel x64) then the companies starts to play technical leapfrogging. At one point of time you lead the competition, at an different point of time it´s the competition.

ZDnet has some interesting informations about the upcoming quad core opteron. Looking forward to see the Sun Fire X4600 with this processors.

Jonathan writes in his The world just changedblog about the real news:

Which is all to say, just when you thought you'd had the industry all figured out, good things happen - and although the coverage in the mainstream media seems focused on our adding Intel chips to our product line, I'd like to believe the real news is that Intel has agreed to add the Solaris operating system to theirs.

But the media seems to look more on the competion between AMD and Intel than to the real effects of this announcement. With the annoucement, we have partnership with both processor heavyweights. This can be only useful to push Solaris forward in the x86 sphere.

After some thoughts while sitting in an aircraft towards Dresden (yes, the customer not unrelated to announcement yesterday), i came to the conclusion, that the announcement yesterday evening can make it easier to sell AMD systems. Sounds counterintuitive, but there is a interesting effect in selling hardware: I call it the subterfuge effect. When there is a weakness in your portfolio, your argumentation is a subterfuge, when there is a strength your arguments are reasonable.

Nice example: Against a wide knows that TPC-C is not an indicator for system speed, it´s an function over the number of harddisks. When you have no or only an average TPC-C value, nobody believes you. Even when you explain it on bit level, many customers see this as an subterfuge. When you lead the TPC-C, the customer believes you, when you say: "TPC-C is utter bullshit". But obviously you wouldn´t do, as this would obliterate an selling point.

With AMD and Intel it´s the same. When you have only Intel or AMD, every argument for your point is only a subterfuge: "You say this, because you can´t sell us Intel". When you have both it´s a good advise.

So, the deal with Intel can actually help us to sell AMD systems.

Ich nutze mein Blog mal als Erwiderung zu Benchmarking Bullshit:

Zum Spamproblem, sag mir mal ungefaehr, wann du gepostet hast, dann guck ich mal ins Logfile. Akismet ist manchmal etwas rigoros. Ich ueberführ deinen Kommentar dann ins Blog.

Zur Sache: Ja, die Konkurrenz kommt ein Jahr, nachdem wir die erste T2000 rausgebracht haben, so langsam ran .... ist ja auch zu erwarten. Deswegen steht ja auch Niagara II bevor. Was die Benutzung von fully buffered angeht: Sagen wirs mal so, ich gehe davon aus, das wir auf die Leistung gerechnet, weniger bis gleich viel verbrauchen werden. Wenn die Maschine doppelt so viel leistet, und doppelt so viel Strom verbraucht, dann bekomme ich aus einem bestimmten Quantum Energie trotzdem die gleiche Leistung. Ich kann zu den genaueren Parametern noch nicht viel öffentlich sagen, aber ich sehe auf der Basis der mir bekannten Dinge keinen Grund zu sagen, das mit fully buffered die Strategie aufgehoben wird. Nur so viel, Niagara2 ist kein Sprung um 10 oder 20 Prozent ... das wird wesentlich mehr.

1. Benchmark fuer die PowerEdge 8-CoreMaschine findet sich hier
2. Der Strombedarf der bei T2000 angegeben ist, korrespondiert mit einer 32 GB Ausstattung. Die Vergleichssysteme sind sowohl bei der 8 Core Fusi als auch bei der 8 Core Dell Maschine jeweils 32 GB, wie man den entsprechenden Diclosures entnehmen kann.
3. Die Fusi-Maschine liegt in der Tat bei ueber 18000 SpecWebs. Die Dell Maschine liegt in der Tat auch ueber den SpecWeb angaben. Nur irgendwas stimmt hier nicht. Paul Murphy hat am 13 Dezember dazu treffend angemerkt:

The configuration issues aren't the only things that raise questions here. For example, the Sun machine recorded no validation errors versus 346 for the Dell (and 513 for the Fujitsu). More interestingly, the Sun machine's results are consistent across the three iterations allowed for each of the three benchmark components. On the banking test, for example, the Sun machine produced composite scores of 32157.5, 33290.2, and 33203.8 for a high - low difference of only 133 - 0.03%
In contrast both Xeon results, that for the Dell 2950 and that for the comparable Fujitsu, show enormous variation. Thus Dell's three banking runs produced scores of 40,333.6, 23,989.6, and 23,710.9 for a variation of 16,623 - 70% of the last two; Similarly, the eight core Fujitsu records scores of 68,659.4, 23,375.7, and 22,908.3 for a variation of 45,701 - very nearly twice the average (23,495) of all four low scores!

Mir ist es vollständig klar, das es irgendwann maschinen gibt, die im Benchmark schneller als eine T2000 sind, die Frage ist aber für welchen Einsatz von Energie und Geld. Insbesondere für den Geldeinsatz ist der oben angesprochene Kommentar von Paul eine sehr interessante Lektüre.

4. Was den Unterschied zwischen der Dell-Maschine und der HP-Maschine angeht (380G5), die trotz nominell ähnlicher Konfiguration erhebliche Unterschiede geben: Klar ist das möglich. Interessant ist, Banking und ecommerce sind identisch, nur bei der Supportanwendung faellt die HP-Maschine deutlich zurück. Da Support ein sehr I/O beziehungsweise netzwerklastiger Teil ist und keine SSL-Transaktionen beinhaltet, steht zu vermuten das der wegfallende nivelierende Faktor Prozessorleistung fuer die SSL-Verschluesselung den weg auf systemische oder konfigurationstechnische Probleme öffnet.


Worauf ich aber eigentlich hinaus wollte mit dem Hinweis auf den Benchmark ist das Skalierungsproblem der Intelplattform. "Buy four cores, get performance of one" ist nicht wirklich berauschend. Benchmarks sind in der Tat so gut wie immer Bullshit, aber sie öffnen interessante Einblicke, wenn man sie miteinander in Verbindung setzt.

Nun liege ich in meinem Bett, kann nicht schlafen, und überlege von einer Seite zur anderen. Zermarter mir das Hirn. Es hat sich in den vergangenen Tagen viel getan, und ich weiss noch nicht, ob ich damit gluecklich bin und gluecklich werden kann. Ich fühle mich ziemlich eingeengt von den Korsettstangen, die mir da angeboten werden. Gemessen an der Sensorik in der Bauchgegend habe ich so garkein gutes Gefühl.

Jon Stokes of ars.technica wrote a nice summary of the stuff already known about the next gen core dubbed K8L:A quick look at AMD's quad-core Barcelona.

I mentioned it a few days ago, that many benchmark sites choose to run their benchmarks with an incapable operating system. The Inquirer wrote a nice article about different system concepts of QFX and Kentsfield and why the market will fall in favour of AMD´s concept of a Quad Core system. But at the end, it´s a question of workload which concept is better for you: Kentsfield, when you have a CPU bound workload, QFX when you have an everything-else-bound workload.

I knew it. I knew it right from the start. The first samples of AMD 4x4 are available and the benchmarking starts. And it seems to be inevitable: Benchmarking an UMA-architecture against a NUMA-architecture with a non-NUMA-aware operating system. Well, when will this fanboiz-sites get a little more sophisiticated in their benchmark-spindoctoring ... at least the quality sites points to the fact that 4x4 needs Vista or an real operating system to take off ...

PS: For a real 4x4 versus Kentsfield benchmark i suggest: Vista , one core for operating system, occassional mail and browsing while photoshops works on something real (filter set on a 36M Hasselblad raw) , two cores for photoshop itself , one cores for rendering a DVD.

Gavin wrote an excellent summary on the fault management architecture on Opteron Rev. F systems.