Homepage Rankings and Research Companies Channelcast Marketing Matters CRNtv Events WOTC NetApp Digital Newsroom WatchGuard Digital Newsroom Cisco Partner Summit Digital 2020 HPE Zone The Business Continuity Center Enterprise Tech Provider Masergy Zenith Partner Program Newsroom Hitachi Vantara Digital Newsroom IBM Newsroom Juniper Newsroom Intel Partner Connect 2021 Avaya Newsroom Experiences That Matter The IoT Integrator NetApp Data Fabric Intel Tech Provider Zone

This Is How Amazon Web Services Is Turning Its Cloud Into A Supercomputer

High performance computing workloads have always been resistant to moving off-premises. AWS intends to change that with specialized hardware and cutting-edge architectures that deliver supercomputing performance on demand.

1   2   3   ... 11 Next

 A Supercomputing Cloud

AWS has a reputation for winning hesitant, often skeptical customers to the cloud. But for years now, the IaaS kingpin has faced a tough sell when trying to convince high performance computing practitioners to ditch their data centers.

The scientists and engineers that run those resource-demanding applications have long resisted Amazon's overtures: They are more comfortable deploying (and more often waiting to deploy) their compute intensive jobs on expensive mainframe supercomputers.

But AWS thinks it's finally found the formula for matching the elasticity of cloud computing with the scale-out capabilities required for high performance computing, Peter DeSantis, Amazon's vice president of global infrastructure and customer support, told attendees of a nighttime keynote that kicked off the AWS re:Invent conference last week in Las Vegas.

Over the last year, the goal to host "some of the most amazing and demanding workloads out there" has become reality, DeSantis said in an annual talk that provides a unique glimpse at the technical wonders of AWS infrastructure.

Advances in custom hardware, primarily networking, now enable HPC applications "to run on AWS the same as any other workload," DeSantis said.

As DeSantis dived deep into the innovations underpinning AWS' progress, Derek Magill, HPC solution architect at Flux7, a pioneering AWS partner that just launched an HPC practice, listened closely.

That DeSantis dedicated his highly anticipated re:Invent keynote to HPC in itself is an important signal to channel partners implementing the technology, Magill later told CRN.

"Not only do we believe it, but Peter DeSantis getting up there and spending an hour-and-a-half talking about HPC on AWS, I think it really validated our move to HPC," Magill said.

And DeSantis did a good job explaining how AWS has overcome the technological barriers that have kept HPC workloads on-premises.

"HPC people talk about how we're different and special and our needs can't be met by cloud," Magill said. But DeSantis’ keynote went far in knocking down some of their biggest excuses.

1   2   3   ... 11 Next

sponsored resources