FlashSoft viewpoint: The importance of the Fusion IO + IO Turbine acquisition

The Fusion-io IPO earlier this year captured a lot of attention, and their recent financial results show that there's a very healthy demand for flash in the enterprise. Fusion-io's plan to acquire IO Turbine further validates the importance of server-tier caching, which we have experienced first-hand in numerous successful implementations of FlashSoft's software.

So, why is SSD-based caching in a server such a big deal? Because right now the typical server is running at just a fraction of its potential because it is bottlenecked by IO latency.

For decades, CPU and system memory have grown dramatically more powerful. But the speed at which a server can read and write data hasn't kept pace.

This is a serious problem for all kinds of enterprise applications and databases, and it's an equally tough constraint on virtualization environments. There have been work-arounds to relieve IO latency, often using flash, or even RAM in storage systems or appliances. These solutions relieve some of the pain, but don't cure the fundamental IO problem.

However there is a cure: a high performance, read-write cache based on non-volatile memory that resides right in the server. This can improve IO performance to the point that it's no longer the primary constraint on server performance and workload. Servers run at their full potential, increasing performance and saving a great deal of money.

With server-tier caching, the IO bottleneck simply no longer exists. Consider the value of this kind of solution: if today's painkillers for the IO bottleneck sell for millions of dollars, how much is a cure worth?

Even more significantly, flash-based caching represents a fundamental architectural change that disrupts the idea of "servers" and "storage" as two distinct, complementary technology categories. This "tier minus one" cache moves what had been thought of as a "storage" capability inside the server. The impact of this change could be enormous within our industry.

Given the importance of this technology, we should expect to hear announcements of new products and initiatives. But in the years we've spent developing our technology, we've seen that it's neither easy nor straightforward to bring a production-quality read-write caching capability to market.

When we introduced FlashSoft a couple of months ago, we announced a software product that (a) was immediately available for purchase and (b) was available for prospective customers to evaluate for themselves. We've been successful with that approach, and plan to follow it in our future product releases.

When we say we have a product available, we mean it's available for you to try and - if you choose - to purchase and put into production use, right now.

Fusion-io has a great technology foundation, and we congratulate them on their news. Joint FlashSoft and Fusion-io customers already use our software to turn Fusion-io hardware into a high-performance read-write cache with great success. We value our relationship with Fusion-io, and we expect to work more with Fusion-io and their customers in the future.