Phase-Change-Memory Devices for Non-Von Neumann Computing
Another fascinating new non-von Neumann approach is that of computational memory, where the physics of nanoscale memory devices as well as the organization of PCM devices in cross-bar arrays are exploited to perform certain computational tasks within the memory unit. I will present large-scale experimental demonstrations using about one million PCM devices organized to perform high-level computational primitives, such as image compression and reconstruction, linear solvers and temporal correlation detection. The results show that this co-existence of computation and storage at the nanometer scale could be the enabler for new ultra-dense, low-power, and massively parallel computing systems.
My talk will focus first on spiking neural networks (SNNs), the third-generation neural networks, which are widely believed to be computationally more powerful because of the added temporal dimension. Phase change memory (PCM) devices could emulate neuronal and synaptic dynamics. Such phase-change neurons also exhibit intrinsic stochasticity, and can thus be used for the representation of high-frequency signals via population coding. In general, such phase-change neurons and synapses can be used to realize SNNs and associated learning rules in a highly efficient manner.
The tremendous increase in data that is being generated could significantly improve our understanding of today's incredibly complex economies and societies. It ushers in a new era of computing, namely, the cognitive era, in which data is considered a new natural resource. However, cognitive computers based on the classical von-Neumann computing architecture pose considerable challenges in terms of area and power consumption. This has triggered research efforts to unravel and understand the highly efficient computational paradigm of the human brain, with the aim of creating brain-inspired artificial cognitive systems. Most recently, post-silicon nanoelectronic devices with memristive properties are also finding applications beyond the realm of memory. It is becoming increasingly clear that for application areas such as cognitive computing, we need to transition to computing architectures in which memory and logic coexist in some form. Brain-inspired neuromorphic computing and the fascinating new area of in-memory computing or computational memory are two key non-von Neumann approaches being researched.