Customers and partners have been asking us what the recent Microsoftorganization changes mean for HPC. What does it mean for HPC? The answer—fullsteam ahead.
We call our workload “Big Compute.” Big Compute applications typically requirelarge amounts of compute power for hours or days at a time. Some of our customersdescribe what they are doing as HPC, but others call it risk analysis,rendering, transcoding, or digital design and manufacturing.
We’re still working on the HPC Pack for Windows Server clusters andenabling new Big Compute scenarios in the cloud with Windows Azure. You’llcontinue to see new features and new releases from us on a regular basis. Theteam is really excited about the new capabilities we’ll be bringing to ourcustomers.
Microsoft’s vision for Big Compute in Windows Azure remains democratizingcapabilities like performance and scale. We have demonstrated world classperformance using Windows Server in the cloud with HPC application benchmarksand Top 500 systems. We’ll continue to make clusters easy to manage and extendthem to Windows Azure. Going forward, we want to make running clusterapplications in the cloud possible for users that don’t have or can’t getaccess to clusters.
The economics of the cloud is fundamentally changing cluster computingby making compute power available when you need it, paying only for what youuse. Our customers in research and industry get it. Users with workstations canhave access to clusters for projects without having to invest in infrastructurethat may sit idle. Our enterprise customers are able to keep their on-premisesservers busy, while running peak load in Azure. And now developers cancost-effectively test applications and models at scale. We are part of theEnterprise and Cloud Division at Microsoft for a reason.
Performance and scale remains at our heart. Because we are part of theWindows Azure Group, we are driving capabilities like low-latency RDMAnetworking in the fabric. We make sure that customers can reliably provisionthousands of cores to run their compute jobs. Windows Azure is evolving toprovide a range of capabilities for Big Compute and Big Data. High memoryinstances help more applications run in the cloud, and we are tuning our RDMAand MPI stacks with partners now.
The research community advances the leading edge of HPC. Our team andMicrosoft Research continue to work closely with partners in academia andlabs—we value the relationships and feedback. Our mission and focus is ondemocratizing these capabilities and making HPC style computing broadlyavailable through services from Microsoft and from our partners.
We’ll continue to use this blog to update you on Big Compute and HPC atMicrosoft.
Group Program Manager, Big Compute
Hi! Great to hear that HPC/Big Compute is still getting attention at MS. One big request on the that front is GPU and Xeon Phi based nodes, along with software support (such as C++ AMP and OpenCL drivers) are also launched.
I am an ex-member of the Interactive Supercomputing team at the NERD center in Cambridge, MA. The team I was in produced MS "Cloud Numerics". I was wondering if you had any updates on MS's work on technical computing, or so-called "big data". It's been a month already.....