From the Editor | Main Blog Index
January 07, 2010
It's tempting to think that the beginning of a new decade will usher in some sort of major paradigm shift in HPC. But that's unlikely to happen, primarily because the pace of change in HPC, and for IT in general, is already built into the fabric of the business. That's due primarily to Moore's Law, which has been generalized into the broader expectation that next year's computers will be cheaper, smaller and more energy-efficient than last year's. That enables most IT users to be conservative about migrating to new architectures or software.
In fact, the end of Moore's Law would be the true precipitating event for a major paradigm shift in the computer biz. But don't hold your breath. Even the most pessimistic projections give us another five years of shrinking silicon. And even if semiconductor-based transistor technology eventually runs out of steam, it's likely that other nanotechnologies like carbon nanotubes and spintronics will carry on.
That said, innovation is still being driven by specific discontinuities: the memory wall, the stalling of processor clock speed, power consumption, mechanical limitations on hard drive performance, and system complexity, among others. In response we're seeing the rise of multi/manycore CPUs, GPGPU computing, flash memory storage and cloud computing. All of these technologies stand to play a big role in HPC.
Of the technologies mentioned, I think the one most likely to make the biggest impact in HPC over the next couple of years is GPU computing. Thanks mainly to the efforts of NVIDIA and its CUDA architecture development, general purpose GPU computing is the fastest moving HPC technology today and has the steepest adoption trajectory. (Using the same criteria, if I broadened the scope to the entire IT industry, I'd tap cloud computing as the leading technology driver.) When the first "Fermi" Tesla GPU products hit the streets in mid-2010, it will be the first time a vector (or vector-like) processor based on a commodity architecture will be available to the supercomputing community. By all accounts, that qualifies as a game changer.
Other HPC innovation in this decade will be driven by new parallel programming development environments from the likes of Intel, Microsoft, NVIDIA, and others. The result, unfortunately, is that we'll end up with a confusing array of languages and libraries, which in many cases will be tied to specific vendors: Ct (Intel), CUDA (NVIDIA), Microsoft (Parallel LINQ), The MathWorks (Parallel MATLAB), and so on. If history is any guide, this is not likely to get sorted out anytime soon.
As far as the outlook for the HPC business itself, the smart money is predicting a modest growth rate in vendor revenue, perhaps less so for server firms, more so for storage and InfiniBand sellers. Overall, the IT sector looks pretty strong right now revenue-wise, with decent job growth in the long-term forecast. The fact that at least half the US stimulus money is still to be spent should bolster HPC work in government and academia, at least through the end of 2010.
On the other hand, there are still too many server vendors with too little differentiation out there, so I wouldn't be surprised to see some more consolidation this year, even as the market rebounds. And if credit and venture capital funding remains tight, we will also continue to see software firms being gobbled up, as we saw with RapidMind, Cilk Arts, Interactive Supercomputing, Visual Numerics, and most recently, TotalView Technologies.
By the way, despite some of the positive economic news over the last few months, there are plenty of analysts out there forecasting a rather sluggish recovery in 2010 and beyond. Washington Post business columnist Steven Pearlstein writes that the uptick in the recent economic numbers may be temporary:
The surprising strength of the bounce-back testifies to the wisdom of the underlying strengths of the U.S. economy and the success of the policies, but is likely to peter out as the stimulus begins to wear off and the inventory correction is completed. Economist Paul Krugman probably has it about right when he says there is a one-in-three chance that the economy will dip back into recession, with the "optimistic" scenario that the economy will neither shrink nor grow but bounce along the bottom.
Welcome to the new normal.
Posted by Michael Feldman - January 07, 2010 @ 5:30 PM, Pacific Standard Time
![]()
Michael Feldman is the editor of HPCwire.
No Recent Blog Comments
Contributing commentator, Andrew Jones, offers a break in the news cycle with an assessment of what the national "size matters" contest means for the U.S. and other nations...
Read more...
Today at the International Supercomputing Conference in Leipzing, Germany, Jack Dongarra presented on a proposed benchmark that could carry a bit more weight than its older Linpack companion. The high performance conjugate gradient (HPCG) concept takes into account new architectures for new applications, while shedding the floating point....
Read more...
Not content to let the Tianhe-2 announcement ride alone, Intel rolled out a series of announcements around its Knights Corner and Xeon Phi products--all of which are aimed at adding some options and variety for a wider base of potential users across the HPC spectrum. Today at the International Supercomputing Conference, the company's Raj....
Read more...
Jun 19, 2013 |
Supercomputer architectures have evolved considerably over the last 20 years, particularly in the number of processors that are linked together. One aspect of HPC architecture that hasn't changed is the MPI programming model.
Read more...
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Read more...
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Read more...
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Read more...
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
Read more...
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
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?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.