In a move to beef up its flash memory tech, IBM on Thursday announced plans to acquire Texas Memory Systems. The Houston-based privately held company develops high-performance flash memory solutions that Big Blue is betting will fit nicely into its Smarter Computing division. Financial terms of the deal were not disclosed.
Although privately held, Texas Memory Systems is no startup. The company was founded in 1978 and sells high-performance solid state storage solutions. Unlike hard disk drives that rely on spinning disks and robotic arms, solid state systems are high-speed data storage solutions based on Flash or RAM memory that promise significantly faster throughput and data access, while consuming less power.
Reducing Server Sprawl
Commenting on the acquisition, IBM's Brian Truskowski, general manager of Systems Storage and Networking explained that the TMS strategy aligns well with IBM's own Smarter Computing approach to information technology by helping clients realize increased performance and efficiencies at lower costs.
"Solid state technology, in particular, is a critical component of our new Smarter Storage approach to the design and deployment of storage infrastructures," said Truskowski, "and part of a holistic approach that exploits flash in conjunction with disk and tape technologies to solve complex problems."
Specifically, Texas Memory Systems offers its solid state solutions as the RamSan family of shared rackmount systems and Peripheral Component Interconnect Express (PCIe) cards. The products are designed to help companies improve performance and also reduce server sprawl, power consumption, cooling, and floor space requirements, all of which in turn can help clients save money and boost performance.
With the Texas Memory Systems acquisition, IBM is preparing for a growing opportunity. IDC estimates the amount of solid state storage solutions being shipped into the enterprise will grow significantly, reaching nearly 3 exabytes by 2016.
Flash Disrupting Traditional Servers
Alex Kurtz, a stock analyst at Sterne Agee, told us IBM's Texas Memory Systems acquisition suggest that the Flash storage market is finally reaching a threshold of disruption to the traditional server and storage markets and architectures.
"Other key Flash private companies include the likes of Violin Memory, Nimbus, Pure Storage and hybrid Flash/SATA vendor Nimble, to name a few," Kurtz said. "We don't see a material impact on Fusion-io business at this point, as IBM has been less than 10 percent of revenue for the last few years, with demand partially driven by Fusion-io resources into IBM accounts. On the fringes, a slight headwind to revenue over the next few years, potentially offset by the NetApp initiative that is not in our numbers."
Kurtz noted that EMC has aggressively diversified its Flash storage strategy with VFCache, along with Project Thunder and with the XtremIO acquisition. As it approaches $700 million in revenue, he believes EMC may have to begin to consider acquiring Fusion-io if its efforts with VFCache doesn't materialize in the first half of 2013 to avoid a repeat scenario of NetApp a decade ago in the NAS market.
"NetApp's upcoming partnership with Fusion-io may also act as an accelerant to EMC considering Fusion-io if that joint product development generates traction as well. The upcoming refresh of VFCache at VMworld, contacts tell us, will be a critical trend to follow in the coming months as it may reveal a tell on where EMC moves next on their PCIe strategy with Micron. Remember EMC sold Quantum DXi before acquiring Data Domain," Kurtz said.
"Ultimately, the window for Fusion-io as an acquisition target would likely be short as it evolves into a much larger purchase -- $3 billion-plus at this point we estimate -- and begins to build out its own networked storage strategy as an independent company."