We recently hosted a webinar together with 451 Research Analyst Jason Stamper from 451 Research Group to discuss Insights to Action using In-Memory Computing. It was so successful that we decided to share some of the highlights with you in this blog series. This is the second post based on Jason’s insights.
In my division in 451 Research I focus on data platforms and analytics, and how they have started to simplify some of the complexity that you would see in a typical organization when it comes to data platforms and analytics. 451 Group surveys many thousands of end users each month and we believe that it’s very important to be pragmatic about technology.
The Economics of Data Change
There are technologies that come and go very rapidly. So it’s important to look for which ones are actually solving a genuine business case and not to set out on investments in new technologies that don’t necessarily have a valid reason to do so.
Hadoop is actually a very good example of where a lot of companies that we speak to, have found themselves going down something of a garden path, where they put an awful amount lot of data into Hadoop thinking that’s it’s low cost. But when they come to recover that data and analyze that data, they find that there’s a huge burden on the IT department, and very low amount of self-service on end users being able to analyze that data.
New technologies are infiltrating their way into IT departments to try and solve some of these challenges and to try and regain some of that trust and be more responsive to the needs of the business.
If you look at the different technologies that are being applied in the data platforms and the streaming data analytics space, there’s actually quite a few:
Our operational databases, old faithful, have been around since the ’60s and even earlier, and they’re still absolutely critical parts of most organizations handling mission-critical applications. There’s a lot of investment that’s gone into them in terms of skills and the technologies involved.
But there are quite of a lot of other technologies that are now being used more and more. Analytics databases reporting and analytics data integration and management, performance management, stream processing is increasingly key. We believe that all of these technologies have their role to play and it’s very hard to say that you should just choose one over another because of course every use case’s skills and requirements are slightly different.
Other than role though, you can see from our research that this market is growing strongly still.
I don’t know how many times people have predicted the demise of the relational database, certainly they did around the rise of object databases and that didn’t happen. They’re still obviously here today. But over those nine market segments we just covered, there’s at least 284 vendors. New ones are being added all the time and you can see good strong growth of about 14% until 2020.
Of course that’s not just because of the data volumes that are changing, but the complexity in handling that data and ensuring that it gets to the right business people at the right time and in the right format.
The Data Grid Impact on IoT
Research conducted by 451 Group suggests that distributed data grids and caches, such as those offered by GigaSpaces, are one of the fastest growing segments of those in the data platforms and analytics market that we track.
You can see they were predicting 32% CAGR growth between last year when we started the research, and 2020. That’s based on our analysis of vendor revenues. In private companies we make estimates of revenues. It’s also based on surveys about buying patterns and decisions, and we put all of that into models, and we model that outcome up with our predictions.
Hadoop is even higher. Then as you move down the stack you’ll see that the operational databases we don’t expect to have nearly as good growth obviously very, very large chunk though in the operational database segment that isn’t going away.
As you may know, once you have chosen a particular operational database as your main platform for mission-critical applications, it can be difficult to move to a different one. So that, although it’s not growing particularly fast, it’s relatively stable for the companies in that space. Stream processing is an interesting sector. We’re seeing that particularly take up as companies are streaming data in there from sensors and smart devices in the era of the Internet of Things.
The distributed data grid/cache is also interesting because it’s really being used not just to speed up application performance times, but as a data platform in its own right. Increasingly these aren’t just caches. They’re much more than that. They’re able to handle numerous different data types. They also offer the likes of high availability and redundancy and fault tolerance.
So rather than just being seen as a replacement to or an addition to some of the older databases that perhaps aren’t keeping up with the latest demands, they’re increasingly being seen as a data platform in their own right, which we think is an interesting trend.
The IoT Hype
As for the Internet of Things (IoT), about 9% are already using it. It hasn’t grown particularly strongly from 2014, which gives you some sense that a lot of the hype around IoT needs a little bit of a reality check.
Having said that, I do think it is changing the world already. If you look at those planning to use IoT in the six months, it’s around 3%. Another 7% expecting to start evaluating and testing it. So many of those since we did the survey would already be using those technologies.
So where are we with IoT today? Is it real? It absolutely is real. It’s a real thing. At 451 Research Group, we talk to the customers all the time who say that they’re rolling out different IoT projects. Whether it’s industrial IoT, connected cars, they’re helping to build smarter cities, smart energy systems, metering in the home. Connected devices of all sorts are being increasingly used.
Today though, it probably isn’t really a true Internet of Things, it’s more of different networks of things. You’ll find for example that your Fitbit or your Nike device doesn’t necessarily talk to all of your other devices even though they’re smart. They might be on certain networks. They can’t necessarily connect to other devices.
We see that as partly a protocol thing, partly because a lot of companies have their own walled gardens as well. But we do see that changing. The protocols and the standards are evolving. I think companies are seeing the value in being part of an open ecosystem in many cases rather than keeping walled gardens around their own particular networks, and we’re seeing more and more collaboration and sharing of data amongst different systems and different providers of technology.
For more IoT and In-Memory Computing insights watch our full Webinar with 451 Group or read the full IoT blog series here: