4. ONNX enables models to be trained in one framework and transferred to another for inference.
CPUGPU
ML HW
DSPFPGA
High level API &
Framework Frontends
Hardware Vendor
Libraries & Devices
Any tools exporting ONNX models can benefit ONNX-compatible runtimes and libraries designed
to maximize performance on some of the best hardware in the industry.
ONNX.ai
github.com/onnx
37. DNN Processing Units
効率性柔軟性
Soft DPU
(FPGA)
Contro
l Unit
(CU)
Registers
Arithmeti
c Logic
Unit
(ALU)
CPUs GPUs
ASICsHard
DPU
Cerebras
Google TPU
Graphcore
Groq
Intel Nervana
Movidius
Wave Computing
Etc.
BrainWave
Baidu SDA
Deephi Tech
ESE
Teradeep
Etc.
39. F F F
L0
L1
F F F
L0
Pretrained DNN Model
in CNTK, etc.
Scalable DNN Hardware
Microservice
BrainWave
Soft DPU
Instr Decoder
& Control
Neural FU
Network switches
FPGAs
40.
41. Azure ML integration
End-to-end deployment and model lifecycle support
Hardware Accelerated
Model Gallery
Brainwave
Compiler & Runtime
“Brainslice” Soft
Neural Processing Unit
42. Model
Management
Service
Azure ML orchestratorPython and TensorFlow
Featurize images and train classifier
Classifier
(TF/LGBM)
Preprocessing
(TensorFlow, C++
API)
Control Plane
Service
Brain Wave Runtime
FPGA
CPU
43. http://aka.ms/aml-real-time-ai
Models are easy to create and deploy into Azure cloud
Write once, deploy anywhere – to intelligent cloud or edge
Manage and update your models using Azure IoT Edge
53. Cloud: Azure 高機能 Edge 軽量 Edge
概要
An Azure host that
spans from CPU to GPU
and FPGA VMs
A server with slots to insert CPUs, GPUs, and FPGAs or a x64 or
ARM system that needs to be plugged in to work
A Sensor with a SoC
(ARM CPU, DSPs)
and memory that can
operate on batteries
CPU
CPU,GPU or Arria 10
FPGA
Arria 10
FPGA
NVIDIA GPU x64 CPU ARM CPU
HW accelerated
DSP,CPU,GPU
モデルパッケー
ジ
Native to Windows
and container elsewhere
Windows
Native
- Linux
container
- Windows ML
- Linux
container
- Windows ML
- Linux
container
- (Ideally) container
- Android Native
- iOS Native
- RT OS