You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: content/english/neuromorphic-computing/hardware/xylo-synsense/index.md
+10-10Lines changed: 10 additions & 10 deletions
Original file line number
Diff line number
Diff line change
@@ -17,10 +17,10 @@ product:
17
17
announced_date: 2022
18
18
applications: Smart sensing
19
19
chip_type: Digital
20
-
neurons: null
20
+
neurons: 1000
21
21
synapses: 278000
22
-
weight_bits: null
23
-
activation_bits: null
22
+
weight_bits: 8
23
+
activation_bits: 8
24
24
on_chip_learning: false
25
25
power: ~5 mW
26
26
release_year: 2022
@@ -32,7 +32,7 @@ product:
32
32
retired: false
33
33
product_name: Xylo
34
34
summary: Xylo is a 28nm 1000 neuron digital spiking neural network inference chip
35
-
optimized for ultralow power edge deployment of trained SNNs, with flexible architecture
35
+
optimized for ultra-low power edge deployment of trained SNNs, with a flexible architecture
36
36
to map various network topologies.
37
37
title: Xylo - SynSense
38
38
---
@@ -41,19 +41,19 @@ Xylo is a digital spiking neural network (SNN) inference processor developed by
41
41
42
42
Xylo is a series of ultra-low-power devices for sensory inference, featuring a digital SNN core adaptable to various sensory inputs like audio and bio-signals. Its SNN core uses an integer-logic CuBa-LIF neuron model with customizable parameters for each synapse and neuron, supporting a wide range of network architectures. The Xylo Audio 2 model (SYNS61201) specifically includes 8-bit synaptic weights, 16-bit synaptic and membrane states, two synaptic states per neuron, 16 input channels, 1000 hidden neurons, 8 output neurons with 8 output channels, a maximum fan-in of 63, and a total of 64,000 synaptic weights.
43
43
For more detailed technical information, see https://rockpool.ai/devices/xylo-overview.html.
44
-
The Rockpool toolchain contains quantizaton methods designed for Xylo, as well as bit-accurate simulations of Xylo devices.
44
+
The Rockpool toolchain contains quantization methods designed for Xylo and bit-accurate simulations of Xylo devices.
45
45
46
46
## Overview
47
47
48
48
Xylo is an application-specific integrated circuit (ASIC) chip optimized specifically for SNN inference. Key features include:
49
49
50
50
- All-digital design using integer arithmetic for efficient simulation of LIF neuron dynamics
51
-
- Supports up to 1000 LIF neurons with configurable synaptic and membrane time constants, thresholds and biases for each neuron
51
+
- Supports up to 1000 LIF neurons with configurable synaptic and membrane time constants, thresholds, and biases for each neuron
52
52
- 16 input channels and 8 output channels using asynchronous spiking events
53
53
- Flexible network architecture including support for recurrent connectivity to map deep networks
54
-
- Ultralow power consumption, with 219 μW idle power and 93 μW dynamic inference power measured on audio classification application
54
+
- Ultra-low power consumption, with 219 μW idle power and 93 μW dynamic inference power measured on audio classification application
55
55
56
-
The chip is fabricated in a 28nm CMOS process and occupies 6.5 mm2 die area. It can operate at clock frequencies up to 250 MHz.
56
+
The chip is fabricated in a 28nm CMOS process and occupies a 6.5 mm2 die area. It can operate at clock frequencies up to 250 MHz.
57
57
58
58
## Architecture
59
59
@@ -63,11 +63,11 @@ The input and output layers use asynchronous events to communicate spikes, avoid
63
63
64
64
## Software Tools
65
65
66
-
Xylo leverages the Rockpool ecosystem for mapping and deploying SNNs. The Rockpool library and Python API abstract the SNN programming to high levels, enabling machine learning engineers to easily train networks using standard methods like backpropagation. A compiler handles mapping optimized networks onto the Xylo substrate.
66
+
Xylo leverages the Rockpool ecosystem for mapping and deploying SNNs. The Rockpool library and Python API abstract the SNN programming to high levels, enabling machine learning engineers to train networks easily using standard methods like backpropagation. A compiler handles mapping optimized networks onto the Xylo substrate.
67
67
68
68
## Applications
69
69
70
-
The flexibility to implement generic deep network topologies makes Xylo suitable for a variety of edge deployments in domains such as audio, time series, and control. Example applications demonstrated include lowpower keyword spotting, biosignal classification, and robotic control. Ultralow idle and dynamic power consumption enables continuous background processing in powerconstrained environments.
70
+
The flexibility to implement generic deep network topologies makes Xylo suitable for a variety of edge deployments across domains such as audio, time series, and control. Example applications demonstrated include low-power keyword spotting, biosignal classification, and robotic control. Ultra-low idle and dynamic power consumption enables continuous background processing in power-constrained environments.
speaker_bio: "Tobias conducts interdisciplinary research at the intersection of intelligent robotics, computer vision, and computational cognition. My main goal is to develop high-performing, bio-inspired computer vision algorithms that simultaneously examine animals/humans and robots' perceptional capabilities. He is a Lecturer (Assistant Professor) in Queensland University of Technology's Centre for Robotics. He joined the Centre as an Associate Investigator and Research Fellow in January 2020. Previously, he was a postdoctoral researcher in the Personal Robotics Lab at Imperial College London. He received a PhD from Imperial College in January 2019. His thesis was awarded the UK Best Thesis in Robotics Award 2018 and the Eryl Cadwaladr Davies Award for the best thesis in Imperial's EEE Department in 2017-2018. He previously received an M.Sc. degree (distinction) in Artificial Intelligence from The University of Edinburgh in 2014 and a B.Sc. degree in Computer Engineering from Ilmenau University of Technology, Germany, in 2013. His works have attracted two best poster awards, one best paper award, and he was the senior author of the winning submission to the Facebook Mapillary Place Recognition Challenge 2020."
14
+
speaker_bio: "Tobias conducts interdisciplinary research at the intersection of intelligent robotics, computer vision, and computational cognition. My main goal is to develop high-performing, bio-inspired computer vision algorithms that simultaneously examine animals/humans and robots' perceptional capabilities. He is a Senior Lecturer (US: Associate Professor) and Chief Investigator in Queensland University of Technology's Centre for Robotics. He is also a recipient of the prestigious Discovery Early Career Researcher Award (DECRA) by the Australian Research Council. He joined the Centre as an Associate Investigator and Research Fellow in January 2020. Previously, he was a postdoctoral researcher in the Personal Robotics Lab at Imperial College London. He received a PhD from Imperial College in January 2019. His thesis was awarded the UK Best Thesis in Robotics Award 2018 and the Eryl Cadwaladr Davies Award for the best thesis in Imperial's EEE Department in 2017-2018. He previously received an M.Sc. degree (distinction) in Artificial Intelligence from The University of Edinburgh in 2014 and a B.Sc. degree in Computer Engineering from Ilmenau University of Technology, Germany, in 2013. His works have attracted two best poster awards, one best paper award, and he was the senior author of the winning submission to the Facebook Mapillary Place Recognition Challenge 2020."
0 commit comments