Fusion-io, a leading provider of system, application & database acceleration, today announced a total of 15 petabytes of enterprise flash shipped, a milestone carved over the past 12 months. This is worth 199 years of unceasing play of HD video or 15, 360 times the entire print contents of the Library of Congress. This is an indication of a swing to ioMemory technology on a global scale, so as to end EMC’s lead on Enterprise Flash.

“Traditional server-storage architectures are being challenged in light of the cost-optimized performance of a server-attached approach to NAND flash deployment in the enterprise,” said Benjamin Woo, Program Vice President of Worldwide Storage Systems Research for IDC.

“With 15 petabytes of Fusion’s server-deployed ioMemory technology shipped in just the last 12 months, it is increasingly clear that server-side flash has moved past the bleeding edge and is making its way into the mainstream of enterprise computing.”

The company boasts the efficiency step up brought about by the ioMemory technology, which we caught a glimpse of at VMWorld last fall.  Not only does it reduce operational cost, but server scale-out and software licensing costs as well. This financial reduction is attributed to each server’s increase in workload. It is best suited in today’s peril of tight budget resulting to over 2,000 users upgrading to Fusion’s ioMemory technology, including half of the Fortune 50.

“With a hundred times the capacity density of DRAM and less expensive on a per gigabyte basis, NAND Flash stands to solve the fundamental data supply problem that has led to 20 percent or lower average processor utilization, even in the face of virtualization,” said David Flynn, CEO of Fusion-io.

“By integrating flash directly into servers using memory controller methodologies and virtual memory management techniques, ioMemory and the Fusion-io Virtual Storage Layer (VSL) make it possible for even large data sets to be housed effectively in memory. As applications can then utilize the full processing potential of the server, workloads can often be increased by 500 percent or more.”