Fusion-io Puts a Little Flash Into Cisco Blades to Boost Productivity

The company says ioMemory in Cisco Unified Computing System Servers will amplify in-server performance and cut latency.

June 13, 2012

3 Min Read
NetworkComputing logo in a gray background | NetworkComputing

Fusion-io has struck up an OEM relationship with Cisco that will see its ioMemory flash platform architected into the Cisco Unified Computing System (Cisco UCS) Server, which the companies say will reduce latency and ultimately improve productivity for enterprises.

Fusion-io's flash memory platform will be integrated into all UCS B-series blade servers to amplify in-server performance, something customers have been looking for, said Gary Orenstein, VP of product marketing. "For too long, customers have been forced to deliver performance by aggregating disk drives," he says. "It's become prohibitive."

Adding more drives means drawing more power and more physical space for equipment, he says. "There's not a company in the world that says it wants bigger data centers, except a real estate company."

Cisco UCS unites compute, network, storage access and virtualization into a cohesive system designed to reduce total cost of ownership and increase business agility. "Cisco architected UCS for highly intensive workloads from the start," says Orenstein, and with ultra-low-latency ioMemory architected into UCS blade servers, storage performance can be "decoupled" from capacity through the integration of a powerful new memory tier uniquely designed to accelerate applications.

This is faster and more efficient, says Orenstein, and means enterprises need no longer rely on power-hungry SAN arrays. Fusion-io ioMemory provides an alternative to either adding more disks or increasing DRAM. "People are starting to see it as a new tier," he says. "Disks are too slow, and DRAM is too expensive. And DRAM is not persistent."

While reducing power consumption and space requirements is valuable, he says, the big benefits to enterprises are the lower latency and resulting productivity that come with quicker response times for applications and databases.

Big data isn't just about volume, he says--data is coming in faster and needs to be processed faster. "The price per transaction is very compelling for flash memory," he explains.

By improving latency and processing queries faster, enterprises are able to put more virtual machines on a single blade server, as well as applications they might not have considered before.

David Floyer, resident chief technology officer with the Wikibon community, says the productivity that comes from improved latency is the key benefit of Fusion-io ioMemory technology. "The technology is all about writing faster," he says. "If you look at applications today, the biggest constraint on complexity is I/O."

Floyers says Fusion-io's SDK and APIs are what make the flash memory platform so powerful.

In April, Fusion-ios released its first SDK to provide developers with native access to the ioMemory flash platform. It includes APIs within user-space libraries, as well as reference application examples made available as open source. By integrating applications directly with this new persistent memory tier, developers are able to optimize enterprise, Web and big data applications through direct programmatic access to the ioMemory computing layer for the first time.

Ultimately, Floyer says, the ioMemory platform can increase the number of I/O calls and database calls. "You can increase the scope of the application and improve productivity for the user," he says.

The initial Fusion ioMemory products for Cisco UCS Blade Servers will be based on the Fusion ioMemory2 platform and are expected to ship in the second half of 2012.

Read more about:

2012
SUBSCRIBE TO OUR NEWSLETTER
Stay informed! Sign up to get expert advice and insight delivered direct to your inbox

You May Also Like


More Insights