OpenSolaris on the Top500
In yet another aftershock from the fall 2008 release of the Top500 list, Sun’s HPC blog points to the first OpenSolaris-based system to make the list. As we all know, Linux dominates the list from clusters to crays. The system graces the machine floor of the Japan Aerospace Exploration Agency and bears a Fujitsu badge. Powered by quad core SPARC64 VII’s, the new machine debuted …
Microsoft opens HPC competency center in Russia
A quickie, since this news is getting a little long in the tooth. Found at HPCwire and sent in by email:
T-Platforms and Microsoft have announced the opening of a joint Competency Centre for HPC (High Performance Computing). The Centre’s primary goal is to provide users with a unique service package by two leading companies of the Russian HPC industry as part of the joint program for academic and commercial organizations.
Users will be able to …
InsideTrack: Rocky Mountain Supercomputing Center
Our faithful reader, Rich Hickey, delivered this great report on Butte’s Rocky Mountain Supercomputing Center, courtesy of his wife Roberta. She was kind enough to venture out and attend the city council meeting whereat the center plans were discussed.
Here is the info I got from tonight…the IBM is being delivered December 27th and it is a water cooled power 7 or whatever the heck you said. One step down from a Blue Gene. The RMSC
…
World record crypto hash throughput
My friend Ilya at software maker Cilk Arts sent me an email with a pointer to a recent example application making use of Cilk’s multicore libraries, this time for the MD6 cryptographic hash
We implemented MD6 for multicore processors using the CILK extension to the C programming language [...] The CILK technology makes multicore programming quite straightforward [...] Our implementation of MD6 in CILK used the layer-by-layer approach (so it assumes that the input message
…
Tokyo Tech Upgrades with Tesla
The Tokyo Institute of Technology announced last week that they have upgraded their latest cluster supercomputer, Tsubame, with NVIDIA Tesla GPU accelerators. Specifically, the group added 170 NVIDIA S1070 1U nodes to the cluster. Wowzers.
It took about a week to add the (Tesla S1070) units in mid-October,” said Satoshi Matsuoka, a professor at the Global Scientific Information and Computing Center of the institute. “I knew we could do it.”
If one considers the 4.1Tflops [32bit, single …
Acumem ThreadSpotter multithreaded application tool released
During SC last week Acumem announced the launch of their ThreadSpotter analysis tool
Acumem ThreadSpotter is the only software analysis tool for the complete set of multi-core software issues. It helps programmers identify and eliminate multi-core bottlenecks by looking at:
Multithreaded interaction
Economic bandwidth usage
Resourceful data handlingJust as Acumem SlowSpotter, ThreadSpotter identifies an application’s slowspots, i.e., parts of the code that can be improved to run faster. For each slowspot, the location in the source code is
…
SGI picks new tape library vendor
During SC08 last week storage company Spectra Logic announced that Silicon Graphics has hitched their star (exclusively) to Spectra Logic’s wagon when it comes to tape libraries
Spectra Logic, the data storage innovator, and Silicon Graphics, Inc. (SGI) (NASDAQ: SGIC) announced today that Silicon Graphics has selected Spectra Logic as its tape library vendor. Due to a mutual focus on high-performance computing (HPC) and shared expertise in managing enterprise data sets, Silicon Graphics will exclusively
…
MATLAB GPU engine announces next beta
AccelerEyes announced yesterday a new beta release of their software, Jacket. John Melonakos, CEO of AccelerEyes, sent me an email about the announcement and the product
Jacket is a GPU engine for MATLAB. Jacket enables standard MATLAB code to run on the GPU, connecting the user-friendliness of MATLAB directly to the speed and visual computing capability of the GPU. Jacket is not another GPU API, nor is it simply a collection of GPU MEX functions. Rather,
…
Banner ad funny: Jaguar running at 164 PF
Which, if true, would force me to stop kvetching about the appalling dearth of peak flops delivered to real applications.
An alert reader (who was bored last week stuck at home while everyone else in the HPC world was in Austin), spotted this in a banner ad at HPCwire and sent it to me. Note the missing decimal point in Jaguar’s performance.
NVIDIA Tesla announcements
There were a lot of them. The NVIDIA team was burning the PR oil long into the night. I don’t want to ignore them, but they don’t each warrant their own post either, since they are most useful as an ensemble indication of market direction. So here is a list of the press release subjects plus links to give you an idea of where you can find Tesla gea
…
