May 28, 2013
Let’s get this out of the way now...
We do not generally report on rumors.
But we are making one exception right now.
With so many sources, all of them credible and from across a rather wide swath of the HPC spectrum, we couldn’t ignore the unconfirmed word of what might be a rather stunning victory for China in the race to produce the fastest supercomputer on the planet.
Again, while unconfirmed, from what we have been led to believe from our sets of sources in both industry and academic circles, China may have something on the order of a 50 petaflop system based on the MIC architecture waiting in the wings for June’s Top 500 unveiling. Sources have claimed that Top 500 brass have already been deployed to validate the results and have emerged with the verification.
All of the above have the been consistent elements of the story across sources—the fact that it’s MIC-based, hits the high note at around 50 petaflops, has been validated… but we still haven’t heard any consistent details (at least with enough consistency to report here) about some of the meatier details, including the interconnect, system vendor, or even the exact site of the super.
"The rumor of a new Chinese #1 system of approximately 40 or more petaflops, if true, is another major feather in the cap for the Intel MIC architecture," said Addison Snell, CEO of Intersect360 Research. "But there are additional potential ramifications that are very interesting in what they could mean to the industry."
As Snell explained, "For starters, despite all theories that we cannot reach exascale until 2022 or later, we haven't yet seen a slowdown in peak TOP500 performance. If someone is fielding 40 or 45 petaflops in 2013, it seems to me that someone will make it to an exaflop by the end of the decade. The fact that it is a Chinese system could spur more discussion, or even funding, in the U.S. and elsewhere."
He continued, noting that it will be interesting to see "the full configuration details, especially with regards to the interconnect and to any technology produced in China. If there is or there isn't a substantial amount of Chinese technology, it's interesting to analyze either way."
According to Snell, "One final piece of analysis will be to see how much of the hardware and software, beyond the processors, comes from Intel. Intel has assembled in-house server and interconnect technologies and a robust middleware and development stack. It's conceivable that Intel could be the prime vendor on this or future supercomputing bids."
At this point, we'll stress once again that all of this unconfirmed, but at the very least it does get the conversation going early about what a Chinese system victory means on a number of fronts.
"Of course, in the end, we all want to see what applications will be run -- what new scientific discoveries will be enabled. That's the case with this supercomputer, and every supercomputer past and future. Supercomputing is about pushing the boundaries of discovery; this new Chinese supercomputer, if validated, represents the new upper limit of scalability."
What do you know? Email email@example.com and we'll update as we go....
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?