yummycandy
Commodore
- Registriert
- März 2005
- Beiträge
- 4.373
http://www.anandtech.com/show/10230...-openpower-hpc-server-with-power8-cpus-nvlink
Interessant ist vor allem der Platzbedarf und der Energieverbrauch.
Das finde ich richtig krank, weil 2004 noch gar nicht solange her ist.
mfg ;-)
Dazu passend noch folgender Artikel:
http://www.heise.de/newsticker/meld...nstlicher-Intelligenz-ausstatten-3163547.html
The new HPC platform developed by IBM, NVIDIA and Wistron (which is one of the major contract makers of servers) is based on several IBM POWER8 processors and several NVIDIA Tesla P100 accelerators. At present, the three companies do not reveal a lot of details about their new HPC platform, but it is safe to assume that it has two IBM POWER8 CPUs and four or eight NVIDIA Tesla P100 accelerators. Assuming that every GP100 chip has four 20 GB/s NVLink interconnections, four GPUs is the maximum number of GPUs per CPU, which makes sense from bandwidth point of view. Keeping in mind that it is possible to install eight Tesla P100 into a 2P server (see the example of the NVIDIA DGX-1), the new HPC platform by IBM could do just that. Still, even with four Tesla P100 accelerators the systems would offer a rather formidable performance.
Interessant ist vor allem der Platzbedarf und der Energieverbrauch.
IBM’s POWER8 CPUs have 12 cores, each of which can handle eight hardware threads at the same time thanks to 16 execution pipelines. The 12-core POWER8 CPU can run at fairly high clock-rates of 3 – 3.5 GHz and integrate a total of 6 MB of L2 cache (512 KB per core) as well as 96 MB of L3 cache. Each POWER8 processor supports up to 1 TB of DDR3 or DDR4 memory with up to 230 GB/s sustained bandwidth (by comparison, Intel’s Xeon E5 v4 chips “only” support up to 76.8 GB/s of bandwidth with DDR4-2400). Since the POWER8 was designed both for high-end servers and supercomputers in mind, it also integrates a massive amount of PCIe controllers as well as multiple NVIDIA’s NVLinks to connect to special-purpose accelerators as well as the forthcoming Tesla compute processors.
A HPC node with eight such accelerators will have a peak 32-bit compute performance of 84.8 TFLOPS, whereas its 64-bit compute capability will be 42.4 TFLOPS. Just for comparison: NEC’s Earth Simulator supercomputer, which was the world’s most powerful system from June 2002 to June 2004, had a performance of 35.86 TFLOPS running the Linpack benchmark. The Earth Simulator consumed 20 kW of POWER, it consisted of 640 nodes with eight vector processors and 16 GB of memory at each node, for a total of 5120 CPUs and 10 TB of RAM. Thanks to Tesla P100, it is now possible to get performance of the Earth Simulator from just one 2U box.
Das finde ich richtig krank, weil 2004 noch gar nicht solange her ist.
mfg ;-)
Ergänzung ()
Dazu passend noch folgender Artikel:
http://www.heise.de/newsticker/meld...nstlicher-Intelligenz-ausstatten-3163547.html
SAPs Business Suite 4 HANA (S/4Hana) soll künftig mit künstlicher Intelligenz von IBM arbeiten. Auf Basis von sogenanntem Cognitive Computing will man komplexe Fragen beantworten, etwa zu strategischen Unternehmensentscheidungen. Diese Technik zeichnet aus, dass sie auf Grundlage unscharfer Informationen, Annahmen, Spekulationen und Wahrscheinlichkeiten nach Antworten sucht. Watson soll grundsätzlich die Kluft zwischen Menschen und Maschinen verkleinern. Dazu gehört, dass Ratsuchende ihre Fragen in natürlicher Sprache stellen dürfen. Das Ganze besitzt gehörige gesellschaftliche Sprengkraft, denn es ließen sich damit potenziell Verwaltungs- und Supportprozesse automatisieren, was zu Arbeitsplatzverlusten führen könnte.