IoT data must be prioritized
But as the Internet of Things evolves and wearables begin to flourish, what are the pressures for memory and Flash storage?
It’s already clear these new devices will have unique requirements for memory and data storage vis-à-vis power, performance, and form factor, but at the same time many IoT devices and wearables will simply be harvesting and transmitting data to the cloud and datacenters to be processed.
When it comes to IoT, memory is pretty broad, since it can be found in devices such as sensors, computers, or smartphones, says Cliff Leimbach, analyst at IHS. What’s happening is there are more yet smaller ways to connect to the Internet to share data. What will dictate memory requirements is how much processing devices will have to do. “If something just has to transmit stuff back, it doesn’t need a lot of memory.”
Leimbach says the DRAM market is continuing to grow steadily at 30% per year. While the tablet explosion came out of nowhere in the past five years, only 6% of all DRAM bits right now are going to tablets, for example. He said for something to make a sizeable impact, it has to have a lot of DRAM. “For these new devices to make a dent, they would have to have a lot of thinking behind them.”
The use of NAND Flash would be affected similarly by IoT devices in that the amount of data needing to be stored would be a factor. If it were a sensor gathering and holding data temporarily to be transmitted elsewhere, the storage requirements would be minimal.
Where big data is having a greater impact on DRAM, whether it’s generated from IoT or other sources, is on the server side, where more processing power is need to crunch the data, and faster storage media such as SSDs might be needed. But where the big data is coming from is almost irrelevant, as Mark Peters, analyst with Enterprise Strategy Group, notes. IoT doesn’t preclude the economics of data storage. “The choices remain the same.”
As storage arrays continue to incorporate more SSDs to serve up information faster, enterprises still have to decide whether the cost of using faster Flash is worthwhile, he says. “It really depends on what you are going to do with all of this data IoT generates. You need look at the volume of data collected, and the speed that data needs to be served. It depends on what [the] application is going to be.”
Peters cited a couple of scenarios where faster storage may or may not be needed to support the processing of data being collected remotely by IoT devices. For example, data from a toll booth equipped with sensors to monitor emissions from passing traffic doesn’t need to be processed quickly, but if the objective is to catch a stolen car within 300 yards, then moving that data faster becomes critical.
The same goes for data coming out of smart appliances in the home. “Everyone loves the idea IoT means all devices in the home are intelligent.” Having a toaster that connects to its manufacturer so it, in turn, can let you know it’s got a month left before its filaments are worn is convenient, but it doesn’t require fast processing. But if that toaster has caught fire while you’re away from home, obviously that information should be dealt with as quickly as possible.
Ultimately, there’s no direct line from IoT to flash storage, says Peters. “It’s not a function of IoT or increase of data volume. It’s a function of what you are going to do with the data.”
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
