A new era of computing is on the horizon.
Storage class memory – solid state, non-volatile, fast, and cheap – is coming. It will replace both (memory) RAM and storage (HDD/SSD) because it does the job of both. It will enable all sorts of new data-centric computing tasks, like intensive pattern recognition in immense data sets (images, genetics, physics, astronomy, etc.) and do it quickly and efficiently. These are things that our brains are actually quite good at.
Today’s computers are not designed like our brain, and are not well suited to perform these types of computations. Interestingly, two announcements this week may open the door to unlocking the commercial potential of in-memory computing. One is a challenge from an industry legend, and the other is a new processor type from an industry giant.
At this week’s Flash Memory Summit, Steve Wozniak laid out the challenge in his remarks to the conference, as reported by eWeek:
Woz Challenges Storage Developers
Wozniak put out a direct challenge to software developers during his 30-minute stage appearance: He’d like to see somebody come up with an operating system and programming language that treats storage a lot differently.
“Now that we have NAND flash, and it can be implemented in ways that are nearly as fast as DRAM (dynamic random access memory), do we need to write our software with both variables in memory and records in storage? Wouldn’t it be neat if the operating system had no storage commands, and the programming language had no storage commands?” Wozniak said.
“I’m waiting to see someone do one example of that and show it to the world.” (1)
The other announcement is the SyNAPSE chip from IBM, which was demonstrated not far from the Flash Memory Summit at IBM’s Almaden Research Center. As reported by MIT’s Technology Review:
IBM Chip Processes Data Similar to the Way Your Brain Does
The efficiency of conventional computers is limited because they store data and program instructions in a block of memory that’s separate from the processor that carries out instructions. As the processor works through its instructions in a linear sequence, it has to constantly shuttle information back and forth from the memory store—a bottleneck that slows things down and wastes energy.
IBM’s new chip doesn’t have separate memory and processing blocks, because its neurons and synapses intertwine the two functions. And it doesn’t work on data in a linear sequence of operations; individual neurons simply fire when the spikes they receive from other neurons cause them to. (2)
It looks like IBM has provided one of the fundamental building blocks necessary for the next generation of computing, and maybe one of the audience members will take up Woz’ challenge and provide the other.
This new era of computing will create extraordinary new value for computing customers, and enormous opportunity for solution providers. Perhaps, dare I say, these new technologies may be highly disruptive!
[A version of this post was originally published on LinkedIn.]