While we waste four cores, scientists use a million at a time

A group of Stanford researchers recently ran a complex fluid dynamics workload across more than a million cores on the Sequoia supercomputer. It’s an impressive feat and might foretell a future where parallel programming becomes commonplace even on our smartphones.

PaaS providers, ThroughPuter wants to parallelize your cloud

A startup called ThroughPuter has a message for the world’s cloud computing providers: Come talk to us, we can make your cloud fly for serious applications. The company has built a hardware platform designed from ground up to run parallel-processing applications in a multitenant environment.

Why Intel put $24M behind cloud, big data

Intel Capital announced $24 million in new investments, and cloud computing and big data companies were the big beneficiaries. They underscore Intel’s understanding that it has to prop up software partners to keep Intel dominant as computing evolves.

Today in Cloud

I’ve been reading a lot about parallel programming this morning, particularly as it relates to big data and analytics. As expected, most of it is Hadoop-related, but Microsoft’s forthcoming Dryad model is also getting attention now that it’s on track for commercial availability. Whatever the foundation, though, we’re bound to see nearly every software vendor worth its salt start pushing advanced BI products very soon. Perhaps the ultimate lesson learned from web pioneers won’t be how to build infrastructures that mirror theirs, but, rather, how to manage and utilize data as effectively as them.

What We’re Reading About the Cloud: August 19

Hadoop, the big data analytics software is so hot right now. Heck anything big data is so hot right now. Today’s links offer insights to Hadoop alternatives, how to use Hadoop and an endorsement of Microsoft’s platform as a service strategy.