Lightning Talk: Software Development for ML and RISC-V Vector Acceler...
summary
Imperas provides fast virtual platforms that boot Linux in ~15 seconds and simulate heterogeneous Arm + RISC-V arrays to develop software long before silicon; one customer runs software a year before RTL on a system with over 150 cores and >1 billion transistors. eSol configures an Arm Cortex-A57 plus 16 RISC-V vector cores to accelerate AlexNet, using configurable vector lengths, toolchain support (GCC/LLVM intrinsics), and a test generator with ~5,000 vector tests.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 0:40Simulation
- 1:02Models
- 1:50Example Platform
- 2:05Machine Learning
- 2:30Multiprocessor
- 3:52AI Hardware Example
- 4:26Trinity
- 4:51Demonstration
- 8:13Configurable
- 8:29Vector Configuration
- 8:55Tests
- 9:12Summary
- 9:29Leaders
- 9:42Outro
related talks
10mAvoiding Amdahl's Law RISC-V Architecture Exploration for AI & ML Compute
13mKeynote: Taking RISC-V Intelligence Everywhere
13mUnleashing ML Processing Power Through RISC-V Vectors: Applications, Implementation and...
09mLightning Talk: RISC-V and AI: How an Open-Standards ISA Underpins Portability f...
19mAccelerating Edge AI with the Help of RISC-V
40mKeynote Panel: The Future of High Perfor...
17mRISC-V is Ready for Powering the Era of Intelligent General Computing
45mThe Benefits of Building New AI Accelerators with RISC-V