14 DAY TRIAL // Advanced Micro Devices may already have the hardware needed to power servers, but it could use some help on the software side, assistance that, apparently, it now got, courtesy of ScaleMP.
Users might not exactly know what ScaleMP deals in, since it isn't actually a company with a stake on the consumer front, unlike Advanced Micro Devices.
What the company does do is research and develop technologies that enable virtualization on high-end computing systems.
Apparently, this expertise is what Sunnyvale, California-based AMD was looking for, as a new press release reveals that the two have entered a collaboration.
The goal is to enable AMD-based servers to use more than four Opteron processors at a time, something that can't really be accomplished, unless one uses custom chipset and motherboards with special chip communication and shared memory.
Basically, ScaleMP will help in creating a virtual symmetric multiprocessing system (SMP) which will combine multiple servers and let more than four physical systems work as a single logical system.
“Supporting AMD servers has been a goal of ours. AMD has a long history of offering powerful, yet cost-effective solutions and we have always seen a natural fit,” said Shai Fultheim, founder and CEO of ScaleMP.
“Through this collaboration, ScaleMP enables AMD customers to create large virtual SMPs that will lower the TCO for organizations – enabling customers to focus their time and budgets on their applications and workloads.”
The vSMP Foundation, as the project is called, will be able to handle up to 128 AMD processor-based servers, for up to 64 TB RAM and 512 CPU sockets, thus thousands of cores.
October 1 is when a limited release will be shipped to customers, with general availability set for November 1 (2011).
"The ability for our customers to aggregate individual AMD Opteron processor-based servers into a virtual high-end symmetric multiprocessor (SMP) computer with vSMP Foundation is a game-changer," said Margaret Lewis, director, product marketing, AMD.
"Our customers can now easily build a cost-effective large memory and CPU system out of up to 128 servers to help meet their high performance computing needs."