John Arthur

I joined IBM in 2010 to work on the DARPA-funded SyNAPSE Project. I am currently a principal research scientist and hardware manager in the Brain-Inspired Computing Group at IBM Almaden Research Center. Previously, I did postdoctoral and graduate work in bioengineering at Stanford University and University of Pennsylvania, respectively.

We published a 2nd chip paper in Science (2023) about NorthPole. NorthPole is a high-performance, high-efficiency neural network inference accelerator. It uses local memory within a parallel, distributed array of 256 cores, linked by NoCs. Fabbed in a 12nm finFET process, the chip has 224 MB of SRAM and uses 22 billion transistors. It represents the most efficient chip to date for several inference tasks such as Resnet-50 classification and Yolov4 detection.

https://www.science.org/doi/10.1126/science.adh1174

We published a paper in IEEE Computer (2019) summarizing the TrueNorth project from inception to presentation of our recent 64-Million-Neuron System.

https://ieeexplore.ieee.org/document/8713821

We published a paper in PNAS (2016) on our algorithm, Eedn, for learning deep neural networks for TrueNorth.

http://www.pnas.org/content/113/41/11441.full

We published a paper in Science (2014) on our chip, TrueNorth, with 1 million neurons and 256 million synapses.

http://www.sciencemag.org/content/345/6197/668.full

www.johnarthur.org