March 28, 2011
A recent symposium on high-performance computing touched on the viability and expense behind HPC clouds, noting that sometimes clouds that are crunching on HPC workloads might be far more cost-effective when following a private cloud model.
According to Rutrell Yasin, who reported on the event for Government Computer News, cloud computing for HPC is likely to be far more expensive when used in the public cloud or IaaS sense. Yasin's argument was based on the experiences leaders from Lawrence Berkeley Natinal Lab (LBNL), the National Institutes of Health (NIH) and the National Oceanic and Atmospheric Administration (NOAA) discussed at the symposium.
LBNL has been at the forefront of cloud computing research for HPC workloads and has benchmarked both performance and cost for a number of workload types.
Kathy Yelick from Lawrence Berkeley National Lab noted that while EC2 is a competitive offering in terms of performance, even in terms of the university’s own private cloud, Magellan, the pricing is far from equivalent. Yelick claimed that EC2 would cost roughly 20 cents per CPU hour whereas their own lab runs workloads on its private cloud for approximately two cents per hour.
Almost a year ago we interviewed Yelick about the benchmarking effort and its outcomes and discussed what some of the best use cases for public clouds were likely to be. From what we can gather, in the time between that initial interview and the unnamed symposium Yasin described, there have not been any significant changes in terms of performance or prince, with external clouds being far more expensive.
Certainly the markup on Amazon’s side (or any other infrastructure provider) is more extensive than some might like to think but for high-performance computing, this kind of price comparison—or at least rough benchmarks to determine the equivalency—are especially important. Still, comparing cloud services versus something that requires a massive initial capital outlay falls into the “apples vs. oranges” category.
What was not addressed on the price front, however, is that cost comparisons of private or internal clouds is a different animal than comparing the price of IaaS services. After all, with one you have significant up-front investment to factor into the equation whereas with the more expensive public cloud service, you are paying for the use of existing resources without any capital at the onset.
Full story at GCN
The ever-growing complexity of scientific and engineering problems continues to pose new computational challenges. Thus, we present a novel federation model that enables end-users with the ability to aggregate heterogeneous resource scale problems. The feasibility of this federation model has been proven, in the context of the UberCloud HPC Experiment, by gathering the most comprehensive information to date on the effects of pillars on microfluid channel flow.
Large-scale, worldwide scientific initiatives rely on some cloud-based system to both coordinate efforts and manage computational efforts at peak times that cannot be contained within the combined in-house HPC resources. Last week at Google I/O, Brookhaven National Lab’s Sergey Panitkin discussed the role of the Google Compute Engine in providing computational support to ATLAS, a detector of high-energy particles at the Large Hadron Collider (LHC).
Frank Ding, engineering analysis & technical computing manager at Simpson Strong-Tie, discussed the advantages of utilizing the cloud for occasional scientific computing, identified the obstacles to doing so, and proposed workarounds to some of those obstacles.
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.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.