BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

MemSQL Brings Facebook-Style Big Data Analysis To The Masses

Following
This article is more than 10 years old.

Big companies are so worried about risk these days that it is a miracle any of them ever buy anything from a startup. That's probably the reason that people used to say you can never get fired for buying from IBM. But a little San Francisco-based database startup is giving big competitors like IBM and Oracle a run for their money when it comes to selling databases to big organizations.

Why would a big company take a chance with a small startup? In the case of real-time big data analysis software provider, MemSQL, the answer is that it delivers better performance at a lower price compared to Oracle. Oracle charges customers "about $100,000 per terabyte (TB) when you break down its $3 million hardware cost" said Eric Frenkiel, MemSQL CEO, in a January 22 interview. Random Access Memory (RAM) on a commodity server goes for $4,000 per TB.

German Sailing Grand Prix Kiel 2006. Team: BMW Oracle Racing. (Photo credit: Wikipedia)

That 96% lower cost per unit of data stored is an excellent example of what I call a Quantum Value Leap (QVL). And evidently that QVL is enough to persuade otherwise reticent companies to buy MemSQL's product. So far, MemSQL's customers include "Comcast, Zynga, Shutterstock and Ziff Davis," notes Frenkiel.

MemSQL -- it was founded in 2011, launched its product in April 2013 and now employs about 40 people -- offers two advantages to such companies -- it's faster and cheaper. That's because it operates in a "distributed" manner -- meaning that it takes a big problem, breaks it into many smaller ones, and sends each of the smaller problems to an inexpensive server to be crunched. Then the answers to all the small problems are pulled together so companies can analyze the results.

Distributed technology means that the problem gets solved faster and the ability to crunch data on cheap servers means that companies can save lots of money on hardware. Frenkiel pointed out that Oracle introduced a $3 million computer for the purpose of doing Big Data analysis last fall with 30 TB of memory -- generating margins in the 50% to 60% range for Oracle.

The commodity servers that run MemSQL are much less costly. And it is hard to see how Oracle would be able to justify giving up those attractive hardware profit margins just to give customers a better deal.

Meanwhile, MemSQL is targeting a market for Big Data technologies that "is forecasted to reach $32.4 billion by 2017 or about six times the growth rate of the overall information and communication technology market," according to Frenkiel.

As I learned when I interviewed Frenkiel in September 2012, he was an engineer at Facebook along with his co-founder, Nikita Shamgunov. They wanted to improve on flaws in "legacy Big Data solutions" and they got their company accepted into Y Combinator.

That helped them develop their product and they have since raised a total of $45 million in outside capital. They announced a $35 million Series B funding from Accel Partners, Khosla Ventures and others after they raised $10 million in their first quarter 2013 Series A round.

Accel is focusing on many investments in Big Data and sees MemSQL as an important part of that portfolio. Kevin Efrusy, general partner at Accel Partners, explained. “Through Accel’s Big Data Fund, we see a very broad database landscape. We’ve backed breakout open-source leaders Couchbase on the transactional side and Cloudera, with its Enterprise Data Hub, on the analytic side, but we also believe there is a very distinct place for MemSQL in real-time analytics."

Effrusy sees a big advantage to companies from using MemSQL's technology. "No executive wants to wait for insights anymore, and MemSQL’s in-memory, SQL technology means they don’t have to. MemSQL has a strong team and a great track record for delivery, so we are thrilled to be their partner,” said Effrusy.

For his part, Frenkiel attributes MemSQL's growth to a better price/performance ratio than competing products. “Demand for our in-memory database technology has been tremendous. Customers are drawn to our solution because it delivers extremely fast performance on commodity hardware and has the ability to store and query multiple data formats in a single database. The result is an extremely fast Big Data solution with versatility and ease of use that lets customers see immediate ROI,” argued Frenkiel.

MemSQL will use the $35 million to expand product development, support its growing customer base, and expand its sales force. Noted Frenkiel, "Our goal is to make big data available to smaller companies by offering a more cost-effective solution. We want to deliver  to the masses Facebook- and Google-style big data analysis."