Beth, Are there any security concerns with "computing in the cloud?" It seems like whenever something likes thise comes up, we're assured it's 100 percent secure, and then we hear about someling like the RSA debacle, which was "under the cloud". Now it's "under the gun". Do you think, for example, that the US Defense Department would do high-speed computign in the cloud? I, for one, hope not.
The security issue is obviously front and center with any form of cloud computing and while I think many of the issues are being ironed out, there are definitely holes and no complete guarantees that breaches won't occur. That said, the kind of HPC cloud service Altair is offering is likely aimed at companies that don't have the financial resources or the data center budgets to have their own HPC capabilities on site. As far as the Department of Defense or any sensitive government agency work, my guess is their HPC needs will not head to the cloud and remain tucked away in the university centers and government buildings where they are played out today.
I believe this will be a very successful project if two factors in the impementation can be accommodated. The first will be ease of use. This is a very useful service and if it can be made easy to use then the benefits to the customers will be the very flexible licensing and capabilities that can be brought to bear.
The second factor is going to be cost. The service model is built on the convenience and cost savings of being able to obtain the capabilities in a cost effective manner compared to operating a similar service in house.
It seems a no brainer that if the above two factors are adeqautely addressed then the service offering is going to be a successful business venture. The customers will benefit and so will Altair.
Once the premise described above is accepted the effort should be directed to helping customers utilize the system and take advantage of the flexiblity offered. I would expect to see tutorials and examples of how this service can be utilized.
I've said this before: cloud is inherently HPC-capable, because you're accessing scalable compute resources. So unless you're talking specifically about supporting apps running under HPC operating systems (like Windows HPC Server), then I submit that, yes, of course you have HPC capability in the cloud, assuming you're willing to pay for the cycles. That's not to critcize either Altair or IBM. As they say on the street, I'm just sayin'...
I've got to believe this is more cost effective than buying time on HPC systems and much more cost effective than investing in on-site HPC, at least for most users. Automotive, aerospace, electronics and medical industries (to name a few) should benefit.
The cloud model is particularly effective for small- to mid-size companies. As Beth notes, this may not be such a good fit for companies with the resources to provide computer umph and sophisticated IT support. Which makes me scratch my head about the federal government embracing the cloud model.
Truchard will be presented the award at the 2014 Golden Mousetrap Awards ceremony during the co-located events Pacific Design & Manufacturing, MD&M West, WestPack, PLASTEC West, Electronics West, ATX West, and AeroCon.
In a bid to boost the viability of lithium-based electric car batteries, a team at Lawrence Berkeley National Laboratory has developed a chemistry that could possibly double an EV’s driving range while cutting its battery cost in half.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.