Feeds:
Posts
Comments

Posts Tagged ‘High Performance Computing’

Well, I was at the long awaited public Oracle/Sun strategy briefing yesterday. A rather long affair that certainly would have been enough time to cover all aspects of where the combined company is heading. They did a pretty good job of it. There were a lot of statements that basically said “we are investing in Sun’s product X and Oracle’s product Y continues to be the strategic direction”, i.e. a lot of “and” and not a lot of “but”. Especially during the software strategy talk. But despite this inclusive theme there were however some glaring oversights and a missed opportunity to provide clarity and state what they will NOT do.

As I told some people I met during lunch, it’ll be interesting to sit down later and ponder over what was NOT said or what was glossed over during the presentations and compare that with the statements that WERE made.

Being an HPC guy, my ears perked up when I heard the Lustre parallel file system mentioned as an example of an important open source project during Charles Phillips opening address. But as it turned out, that was the extent of telling us about the path forward with regards to HPC for Oracle/Sun. It was also the extent of Lustre directions. With nothing explicitly said about HPC, I and others are left to speculate and read between the lines.

What WAS said was that Sun’s x64 systems would be focused on integrated clusters for the enterprise. They emphasized “integrated” and “enterprise”. I guess you can interpret that in several ways, but to me that sounds pretty much like the Exadata system that was launched in the fall and very different from selling general purpose servers (that btw also can be used for HPC). Was that a bone thrown towards Dell?

Oracle’s On Demand centers use Dell servers and NetApp storage as far as I know. I can imagine these will be switched to Sun servers and storage going forward. NetApp got sort of a black eye when Larry Ellison positioned Sun’s ZFS storage appliance as a next generation NextApp, just better, faster and cheaper. There was no further reference to Dell however. The gloves never came off. A lot of Oracle software run on Dell hardware…

HP wasn’t mentioned much either, IBM was used for almost all competitive comparisons. I guess I’d put what happens with the Dell and HP relationships in the “glossed over” category.

Good to hear that they are hiring though. That message wasn’t glossed over.

Update: HPCwire made similar observation with regards to the future of HPC at Oracle/Sun

Read Full Post »

Going back to the HPC education issue I mentioned in my previous post, that actually touches on the other theme that was clear today at the sessions at the 23rd HPCC conference.

Seems like almost no one argues against that the future for continued increased performance in HPC is moving towards a future with more multithreading, multicore, many-core, GPUs and other accelerators, often in a heterogenous mix of thousands (or millions eventually) of each.

This is not a panacea however, there certainly are problems to be solved in those areas as well, the infamous memory wall and energy consumption would be two of them.

The biggest challenge in my mind though is on the software side. Our middleware, tools and applications are just not keeping up. We don’t have the software technology today that makes it easier to automatically take advantage of the inherent parallelism in the hardware infrastructure. We’re today edging into the Petascale era and providing essentially assembly level programming tools. That won’t work for the next level, closer to Exascale.

We need to invest in software that on one hand hides the underlying complexity and makes it easy to scale and on the other hand makes it possible to state the problem to be solved that is close to the natural representation of it. Much like Fortress allows mathematical notations to be used to easier represent equations. We need to bridge the gap between the domain knowledge that can describe the problem and the low level “magic touch” that is needed to get code to scale.

It’s not that this is new news. Many people have pointed this oput, but we don;t seem to make progress towards a solution. It’s not that we as a “collective ostrich” is hiding our head in the sand and hoping it will go away. It won’t.

The problem is that there’s no business case for a single software vendor to take on this huge challenge. This is an area that definitely requires government funding and industry wide attention.

One speaker suggested that HPC needs to be elevated to the same importance as a nationwide energy strategy. It’s that important. I tend to agree with him. We need to do whatever it takes to start to make progress in this area.

I also intend to continue twittering tomorrow under my Bearcrossings twitter id.

Read Full Post »

I just finished up the first day of the 23rd HPCC conference in Newport, RI. It’s a well organized conference with many good presentations. Well worth the time attending. Much good news and progress, discussion of futures but also highlighting some problems and challenges we’re facing.

There were a few recurring themes throughout the day when it comes to problems and challenges. One was that we have a problem with creating the pull for a new generation of students to get into HPC. Several speakers mentioned that university level courses for introduction to HPC or parallel programming are attracting small number and are not growing or actually being canceled. This seems to be especially true in the US, while interest seems to be greater elsewhere.

We as an industry have also failed to make the connection between HPC (which many still instinctively think of as only the classical, high end HPC) and commercial use of HPC technology. One speaker mentioned that the Financial Services industry spent $30B on computing in 2007-8 in search for a competitive edge. There’s sure to be interesting (and well paid) jobs in that area that require a solid HPC foundation. Other areas would be to work in bio-sciences and be part of finding the cure for some of our most dangerous illnesses. Or in the Energy field to solve our energy problems. Lots of big challenges where you can feel proud of being part of a team addressing them, not only monetary rewards.

If we don’t fix this, then we’ll eventually run into an age gap problem, an abyss is almost literally opening up in front of us.

In my next post I’ll write about the second theme among today’s discussions about challenges.

I also intend to continue twittering tomorrow under my Bearcrossings twitter id.

Read Full Post »

I’m researching new uses of technologies and methods originating from High Performance Computing (HPC), some call it Technical Computing and others Supercomputing. I’m focusing on use cases in more unconventional areas such as in business or marketing processes.

There are many examples of technologies that have made this transition beyond HPC and reached a wider user base. A few of them are grid computing, parallel computing, use of accelerators and many others. The Internet itself and web browsing are other famous examples. On the horizon we have Cloud Computing that also can be said to have its roots in HPC.

What I’m looking for are new trends and new use cases where “HPC technologies” help solve largely non-technical problems. For example, are substantial compute resources being used to simulate the performance or predict future behavior of business models or marketing programs before they are launched? If so, how common is this usage? With the rise of Cloud computing, are new technologies now within reach of almost anyone? for small businesses or for home use?

I’m exploring new usages and new trends and don’t want to limit it too much by asking too specific questions at this time. Please leave a comment on this entry if you’re aware of anything that you think may fall in this rather open area, or if you have suggestions for where to look.

Read Full Post »

Follow

Get every new post delivered to your Inbox.