GrAI Matter, Paris research gives rise to AI processor for the edge
GrAI One, the first instantiation of the GrAI Matter‘s NeuronFlow technology, is implemented in the 28nm HPC manufacturing process from TSMC.
The NeuronFlow architecture combines ideas of dataflow processing with neuromorphic computing. GrAI Matter claims that GrAI One has low latency making it suitable for applications in autonomous navigation, human-machine interaction and smart healthcare markets.
The die is 20 square millimeters in area and implements a mesh of 196 neuron cores with local neuron/synapse memories. The neuronal cores process either 8bit or 16bit data and with each core representing 1,024 neurons the overall chip can scale to represent up to 200,000 neurons. The chip is essentially a fabric of neuron cores connected by packet-switched network-on-chip.
GrAI One block diagram: Weights are stored in local SRAM and can be shared. Data and weights are passed to neurons that have state in local SRAM and perform basic neural and ALU functions. Source: GrAI Matter Labs.
The chip provides a GPIO interface to offload latency-critical AI workloads of a host processor. At 100 percent neuron core utilization, GrAI One may consume as little as 35mW at 25 MHz clock frequency.
Next: Presented to Linley Gwennap
GrAI Matter Labs has presented its processor at the Linley Fall Processor conference and compared its chip to Loihi from Intel and IBM’s TrueNorth chip. At keyword spotting GrAI One can spot a word in less than 3 microseconds and consumes 0.1mJ per inference compared with Loihi’s 0.8mJ per inference. At hand gesture recognition GrAI one takes less than 1 microsecond and consumes 25mW compared to TrueNorth’s 44mW.
The GrAI One is supported by the GrAIFlow software development kit. This allows both conventional program execution and machine learning computation via industry-standard frameworks and languages such as like TensorFlow, Python and C++. The kit includes compiler, simulator, debugger, graphical editor and compute and network.
GrAI Matter Lab’s technology is based on research carried out at the Vision Institute (UPMC) in Paris on the human brain over the past 20 years.
GrAI Matter Labs was founded in 2016 and already has 46 employees. The company closed a Series A round of equity financing worth $15 million in 2018, led by iBionext, 360 Capital Partners and 3T Finance.
Related links and articles:
News articles:
Intel scales up self-learning neuromorphic computing
Intel pays $2 billion for AI chip firm
AI processor market to exceed $20 billion in 2019
Groq enters production with A0 tensor processor
Gyrfalcon builds out AI chip range, raises efficiency
China backs Israeli deep learning processor startup
Microsoft’s Azure cloud now on Graphcore IPUs
NovuMind benchmarks tensor processor
AI processor startup emerges from Imagination
China startup releases AI processors then raises $100 million
Microsoft, Alexa, Bosch join Intel by investing in Syntiant
Kneron launches improved line of NPUs
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
