Violin Systems was one of the first storage vendors to bring an all-flash storage system to market. While the company had early success in the market, a combination of fierce competition and poor management decisions lead to the company filing for Chapter 11 protection, Violin is back now, as Violin Systems, with a clean balance sheet and a core group of loyal customers. Does it make sense for organizations to include the company in its all-flash decision making process?
A Flash History Lesson
Violin’s roots were actually in DRAM-based arrays, so designing a hardware system that would fully exploit DRAM was the top priority. As flash became more enterprise-ready, Violin quickly transitioned to all-flash arrays. Its key differentiator was the internal design of the system which delivered very high performance at incredibly consistent low latencies.
At the heart of the differences from its competitors was Violin choice to use memory modules instead of solid-state drives (SSDs). The decision allowed them to squeeze more capacity in their systems and to spread the write load across more modules, increasing durability while at the same time delivering on that low latency. It also had a sophisticated flash interconnect that further reduced latency.
The software included with the early Violin hardware, while functional was “basic.” But in the early days of flash arrays, “basic” was necessary so that the software didn’t interfere with flash performance.
The problem was the next wave of all-flash vendors focused on software design instead of hardware design. This next wave all-flash vendors leveraged commodity servers for their storage systems and SSD for the flash media. Thanks to continued advancements from Intel, the processing capabilities of the CPU within those systems could drive more sophisticated software. The performance the next wave of all-flash systems delivered, while not up to Violin’s abilities, was “good enough” for many enterprises.
With their performance concerns being met by “good enough” enterprises started to demand more feature rich software that had capabilities they had come accustomed to in hard disk-based systems like snapshots and replication and the ability to scale-out instead of up.
Enterprises also wanted flash specific features like data deduplication and compression. These two elements drove the price per GB of all-flash down to the point that it could rival the price point of a performance-oriented hard disk array. While Violin added a robust software offering to its systems, they suddenly went from ahead of the curve to behind the eight ball.
The transition to commodity-based all-flash arrays with a heavy emphasis on software combined with several management miscues left Violin in a precarious financial state. Essentially, Violin was a company with great technology but a bad balance sheet. Eventually, the company had to file for bankruptcy.
Out of the Ashes
Quantum Strategic Partners (QSP) bought the assets of Violin Systems. The investment advisor to QSP is Soros, a private investment management firm that serves as the principal investor to a number of private investment funds that have a multi-billion dollars aggregate asset value.
As a result of the purchase by QSP, Violin Systems’ is debt free with a fully funded P&L. Unlike many storage companies, they now have a short path to profitability. They can afford to deliver their products either in a pay as you grow model or as a subscription sales model.
Violin Systems Today
As part of their reemergence, Violin did not have to reinvent their products. It is indeed a re-start-up, with a great financial backer, 120+ customers worldwide and growing revenue. They had great technology and has along the way caught up and in some case surpassed their competition in software capabilities. At the same time, the all-flash market is hitting “wave 3” where “good enough” performance is not good enough for an increasing number of data centers.
Most impressive is that Violin is not hiding from its past stumble. It kept the same name and Ebrahim Abbasi, the CEO, is very transparent and forthright in explaining what went wrong, why the situation is different now and why he believes Violin’s products are ideal for data centers looking to extract the full potential from memory-based storage. He also stated that he has the backing of his new investors to move into the future with a NVMf platform driving the latency even lower, and a new SW platform that includes a complete suite of data services plus SDS, Scale-out, QoS, predictive intelligence and file/block/object on the same platform.
Violin is moving forward with lessons learned from the past, a clear strategy, a focused plan, a committed employee base (almost no attrition since emergence), and a loyal install base.
They have also re-focused their solutions into three categories. The 7700 and 7750 systems are focused on enterprise production systems and deliver zero downtime and zero data loss. The 7300 and 7450 are high capacity platforms delivering 3X to 20X deduplication efficiencies, and the 7600 and 7650 platforms are focused on high performance delivering two million IOPS with very consistent and very low latencies, as low as 100 microsecond latency for 1M IOPS.
It would be easy to write Violin off and not give it a second chance. And a few pundits in the industry are doing just that. But considering they’ve never actually bought or installed a Violin array, the reader should consider the source of the complaint. Instead, IT planners should look at the Violin installed base, which through the most tumultuous of times stuck with the company.
Then IT planners should look at the state of the company right now; well funded, no debt and a realistic path to profitability. Few other all-flash startups are in a position to make all of those claims.
Finally, IT planners should look at the state of the actual solution. The hardware easily remains one of the best in the class and is ready for the third wave of storage performance. The software, at a minimum, rivals the competing solutions and services remain a strength of the organization.