
Virtualization and Cloud executives share their predictions for 2014. Read them in this VMblog.com series exclusive.
Contributed article by Salil Deshpande, Managing Director with Bain Capital Ventures
Predictions from a Software Infrastructure VC
I'm
one of two Managing Directors in Bain Capital Ventures' new Palo Alto office,
which has made more than a dozen sizable investments (not counting seeds) in
less than two years. With approximately $4B under management and offices in Palo
Alto, Boston and New York, we're the venture arm within Bain Capital, with $70B
under management. We are currently investing out of a $660M stage-agnostic
venture fund raised in 2012 in projects ranging from $500K seeds to $50M
growth-stage investments. I focus on investing in software infrastructure
startups. The guys at InfoWorld cornered me in a dark alley, put a gun to my
head, and asked me to make predictions, so here they are.
Predictions:
- Ancillary problems in the
cloud
When you move your apps to the cloud, you move your enterprise's ancillary problems to the cloud as well. Infrastructure software such has MuleSoft (which integrates silos of applications in the cloud and on premise) and Aria (which handles billing for cloud applications) will become big important businesses. I’ve invested in both MuleSoft (at my prior firm) and Aria (at Bain Capital Ventures).
- Memory is the new Disk
A number of cloud applications are transactional, but not classically transactional - they do not need, and in fact are slowed down by, classical transactional semantics. These apps need to manage their data in memory, often doing new-style transactions in memory, never writing them to disk. We've invested in Redis (an in-memory NoSQL database) and Hazelcast (in-memory data grids for Java).
- Reactive Applications
There's a new way to write cloud applications takes the advantage of multiple cores and nodes. That new way is starting to be called "reactive applications." Reactive applications, also known as event-driven or event-oriented applications, are applications written in such a way that they don't ever stay blocked on anything else. Anytime an application blocks waiting for something to happen, such as a database call or other long operation, it makes the application less easy to run in parallel and thus less scalable. And with Moore's Law tapping out and additional computing power coming from more nodes and more cores rather than cores getting faster, the non blocking way of writing more applications is ever more important. Node.js popularized this style of writing applications. But with Node.js, you end up writing your server apps in Javascript. Enterprises that want to write server apps in Java, and in the future, Scala, will use frameworks called Akka and Play from a company called Typesafe which lets them retain the benefits of Java, and partially transition to Scala over the next decade.
##
About the Author
Salil
focuses on software infrastructure, open source, middleware and enterprise
software.
Prior
to BCV Salil spent seven years as General Partner at Bay Partners, where he
invested approximately $85M into 20 companies including DynaTrace (application
performance management; acquired by Compuware), Buddy Media (platform for brands
to engage with customers on social media; acquired by Salesforce.com), SpringSource (open source;
acquired by VMware), Engine Yard (open source, Platform as a Service), MuleSoft
(open source, cloud and enterprise integration), Lending Club (peer to peer
lending), G2One (open source, Groovy, Grails; acquired by SpringSource),
Sonatype (open source, software component management), Jambool (virtual
currencies for social graphs; acquired by Google), Covestor (marketplace for
emerging hedge fund managers), FantasyBook (fantasy sports apps on Facebook;
acquired by Citizen Sports, now part of Yahoo), Triggit (retargeting on
Facebook), Dropcam (webcams with cloud DVR service), and Vaxart (oral vaccine
platform). Salil also inherited Tealeaf (customer experience management;
acquired by IBM) and served on its board for three years prior to its
acquisition by IBM. Salil's numerous angel investments include ZeroTurnaround
(faster Java redeployments), and Datastax (the Cassandra NoSQL database).