Two presentations were given at the February meeting. Both will
also be given at the CHEP’09 conference in Prague in March.
Sverre Jarp
pondered if the
Atom (N330) processor is ready for High Energy Physics.
In accordance with their tests, it might have potential although
its pure performance is far behind the Xeon processor. However,
two issues are still worrisome: the power consumption (of the
Atom chipset) and the overall memory limitation of 2GB. The Atom
N330 is the biggest member in the current family (with 2 cores
and 4 threads), but it must be kept in mind that both Atom and
Xeon families will evolve. Sverre concluded saying that the Atom
N330 processor would be interesting but that the above-mentioned
issues need urgently to be tackled before it can reach High
Energy Physics.
Andrzej Nowak
presented
an update on perfmon and the struggle to get into the Linux
Kernel.
After a review of perfmon and its capabilities as well as the
tragicomic saga of the efforts to get perfmon in the Linux
kernel, he ended by saying that perfmon support at CERN will at
any rate continue and that perfmon will be closer integrated
with the developers in the physics experiments.
Both presentations are available on the openlab Web site.
Finally,
Sverre Jarp
reminded the audience that the next Minor Review Meeting will
take place on March, 17th.
Charlotte-Anaïs Cornaton, Sverre Jarp
|