November 16, 2011
High-bandwidth, low-latency Mellanox Infiniband solutions enable servers with rCUDA to remotely access GPU servers with negligible overhead costs
SEATTLE, Nov. 16 -- SC11 -- Mellanox Technologies, Ltd., a leading supplier of end-to-end connectivity solutions for datacenter servers and storage systems, the Universitat Politecnica de Valencia and the Universitat Jaume I, today announced the joint demonstration of rCUDA (remote CUDA) over InfiniBand during SC11 at Mellanox booth #522. rCUDA enables applications to access multiple GPUs, even if the GPUs are not physically attached to a computing node by a PCIe bus. When deployed with Mellanox InfiniBand solutions, all the nodes in the cluster efficiently share all the GPUs, resulting in significantly reduced GPU idle time, application execution time, energy consumption and number of GPUs required to run the application. This leads to networking flexibility, higher performance and reduced OPEX spending.
When compared to local CUDA within a single node, rCUDA with Mellanox InfiniBand has a runtime overhead below 3 percent, rendering negligible performance cost for most rCUDA-accelerated applications to access remote GPUs.
"We are delighted to be working with the Universitat Politecnica de Valencia and the Universitat Jaume I to bring the performance benefits of Mellanox InfiniBand interconnect solutions to their rCUDA middleware," said Gilad Shainer, Sr. Director, HPC & Technical Marketing. "Mellanox InfiniBand technology, coupled with rCUDA, delivers the highest performance for GPU virtualization technology while reducing operational costs and infrastructure complexity."
To learn more about rCUDA or request a free OpenSource version of the package, visit the Universitat Politecnica de Valencia (http://www.gap.upv.es/rCUDA) or Universitat Jaume I ( http://www.hpca.uji.es/rCUDA).
Mellanox Technologies is a leading supplier of end-to-end InfiniBand and Ethernet connectivity solutions and services for servers and storage. Mellanox products optimize datacenter performance and deliver industry-leading bandwidth, scalability, power conservation and cost-effectiveness while converging multiple legacy network technologies into one future-proof architecture. The company offers innovative solutions that address a wide range of markets including HPC, enterprise, mega warehouse datacenters, cloud computing, Internet and Web 2.0. Founded in 1999, Mellanox Technologies is headquartered in Sunnyvale, Calif., and Yokneam, Israel. For more information, visit Mellanox at www.mellanox.com.
Source: Mellanox Technologies, Ltd.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?