![[Day Permalink]](http://radio.weblogs.com/0112083/images/permaLink.gif){kind=small}
Wednesday, November 10, 2004
![[Item Permalink]](http://radio.weblogs.com/0112083/images/ur_bullet.gif){kind=small} -- Comment()
Trading CPU Time Like Corn? "Sun Microsystems President Jonathan Schwartz is convinced users will purchase time on his company's machines by buying it on eBay."
|
|
System X fastest academic supercomputer -- Comment()
Earth Simulator, G5 Cluster Drop In 'Top 500' List: "The November Top 500 supercomputer list has been published at SC2004. Topping the charts is IBM and the US Department of Energy's 'BlueGene/L DD2' beta system, at 70.72 TFlops, followed by NASA's 'Columbia' at 51.87.TFlops. For the first time in several publications of this list, Japan's Earth Simulator is no longer in the number one slot, falling to third. Virginia Tech's 'System X' Xserve G5 cluster, while 20% faster than the original cluster that debuted at number 3 last November, has fallen to number 7 due to the new entries, but remains the fastest supercomputer at an academic institution." Last year I wrote about my impressions of the Apple G5 Dual Supercluster at Virginia Tech. At that time the machine was an amazing job, and it it still. Now they are using special Xserve systems with 2.3 GHz processors, and thus the aggregate speed is faster. The communication network is the same, so the system has the same limitation concerning bandwidth and latency. Is is amazing how cost-efficiently Virginia Tech achieved 12.25 teraflop/s with System X. Of course, PowerPC 970fx is an impressive processor, which explains a lot.
|
|
Solve hi-res printing problems with AirPort Extreme Base Station -- Comment()
I have been having intermittent printing problems with a Canon i455 connected to the USB port of an AirPort Extreme Base Station at home. When I print wirelessly through the wlan connection, the print job hangs at some point, and I have to reboot the printer and the AEBS. This problem started last summer, before that wireless printing worked fine. The problem only occurs when printing files with large graphics such as photos. Small print jobs (Word documents or web pages with small graphics) print just fine. Yesterday I first tried four times to print 1.3-1.4 MB photos from iPhoto, and each time I only got a partial printout. Then I noticed a cpu-consuming process called PrintJobMgr in the process view. Just for fun I tried printing again, and when a new PringJobMgr process started, I went to Terminal and increased the priority of this process with the command sudo renice -20 pid (Here pid is the id of the PringJobMgr process.) Now the printing went fine without problems. I printed three more photos, both succeeded fine (you have to give the renice command several times, because each print starts a new PrintJobMgr process). I'm using a 600 MHz iBook with an AirPort card, running Mac OS X 10.3.5. Perhaps the problems are caused by having a too slow machine, which is not able to feed the printer fast enough through the AirPort connection. Printing photos with the i455 directly connected to the iBook works fine. This problem is a nuisance. On the other hand, Canon i455 is not certified to work with AEBS, so who to blame. On the other hand, the problem should be easy to fix by making sure the printer does not time out when there are delays in data transfer. I posted this hint at the Apple discussion forum, where many users seem to have similar problems.
|
