August 12, 2010
Although Oracle hasn't made any announcements about its commitment to the high performance computing market, it has become pretty clear to me that the company is no longer pursuing the HPC business it inherited from Sun Microsystems. Prior to the merger in January 2010, Sun had an array of HPC offerings (servers, storage and middleware) and was surging back into the market with its Constellation-class blade server product. But from all appearances, the HPC product set and talent it acquired from Sun will be absorbed Borg-like into the Oracle collective.
Even that, perhaps, is a bit too generous. Oracle is apparently shedding HPC staff, in a very un-Borg-like manner. A recent article in The Register reported that Oracle had let go much of the HPC sales team, with the remainder tasked to sell the company's Exadata data warehousing appliances. I, myself, have spoken with two credible sources that told me HPC engineering talent is also being axed. Although this has been rumored to have been going on for some time, the recent RIF last week was said to cut particularly deep.
In truth, Oracle never really had an HPC business to lose. Although the Sun IP and product set were acquired in the merger deal, Sun itself hadn't closed any new deals several months prior to the Oracle buyout. All of the HPC systems deployed in late 2009 were the result of contracts that had been signed much earlier. With the pipeline now flushed, it's highly unlikely we'll see any more HPC system deployments from Oracle.
Sun's flagship HPC platform, the 6048 Blade Constellation system, is still listed on Oracle's website, but it looks like it only comes with blades using last year's x86 silicon. For example, the only Intel blade available for the 6048 chassis is the X6275, which comes with Nehalem EP (Xeon 5500 series) CPUs. As of today, there is no Westmere EP (Intel Xeon 5600) hardware available for this enclosure. The X6270 M2 server, which does have the Westmere chips, is only being offered for Oracle's enterprise-class 6000 system. The sole AMD Opteron blade available with the 6048 is the quad-socket, six-core X6440 module. No Magny-Cours (6100 series 8- and 12-core Opterons) server blades are even listed on Oracle's site.
Essentially this means that all of Sun's Constellation customers don't have an upgrade path for the machines they bought within the last three years. This includes some big name supercomputers like TACC's "Ranger" and Sandia National Labs' "Red Sky" in the US system, Jülich's "JuRoPA2" in Germany, and KISTI's "Tachyon" in Korea. A number of smaller Constellation systems are in the same boat.
It looks like Lustre-based HPC storage is also on the way out, if not already gone. Although the storage hardware products -- StorageTek arrays, J4000 series expansion arrays and the Sun Fire X4540 storage server -- are still listed, it appears that the Lustre file system is no longer being offered on these platforms. Since many of the Constellation supercomputers sold over the past couple years came bundled with Lustre storage, those customers would be hard pressed to expand their existing disk capacity without switching file systems.
Even the term HPC is being stripped out of existing products in some cases. For example, Sun HPC ClusterTools has been renamed to Oracle Message Passing Toolkit. Despite the name change, the toolkit is still downloadable, but I would say its future support for HPC is uncertain.
If I still haven't convinced you that Oracle is cutting HPC from its lineup, consider that the company has no exhibit at the Supercomputing Conference (SC10) in November, and as far as I can tell, is offering no presentations. Given that this is the largest HPC exhibition of the year, this should be a clear signal that Oracle is going to be leaving the teraflopping and petaflopping to others.
Posted by Michael Feldman - August 12, 2010 @ 5:48 PM, Pacific Daylight Time
Michael Feldman is the editor of HPCwire.
No Recent Blog Comments
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?