Avoiding Amdahl's Law RISC-V Architecture Exploration for AI & ML Compute

on YouTube ↗10m

" summary

Imperas builds configurable virtual platforms that simulate RISC-V systems and run full software stacks a year before silicon. The demo runs almost a billion instructions, models a RISC-V vector engine, and evaluates a 150+ core ML array whose test suite takes about two hours at 500M ops. The simulator supports heterogeneous RISC-V/ARM platforms, provides multi-core debugging and performance traces, and runs roughly an order of magnitude slower than native x86 while still enabling architecture choices.

AI-written summary · about

" chapters (3)

1. 0:00Introduction
2. 1:16The Challenge
3. 5:26Machine Learning

" related talks

• 10mLightning Talk: Software Development for ML and RISC-V Vector Acceler...Simon Davidmann, Lee Moore · RISC-V Forum: Vector & Machine Learning 2021
• 19mAccelerating Edge AI with the Help of RISC-VFlorian Zaruba · RISC-V Summit North America 2023
• 13mKeynote: Taking RISC-V Intelligence EverywhereCharlie Su · RISC-V Summit North America 2023
• 13mUnleashing ML Processing Power Through RISC-V Vectors: Applications, Implementation and...B. Barker · RISC-V Summit North America 2025
• 30mRISC-V: Enabling Open Physical AILuca Benini · RISC-V Summit Europe 2026
• 22mAdvancing AI Computing With Optimized Matrix Multiplication Techniques With RISC-V CPUChun-Nan Ke, Heng-Kuan Lee · RISC-V Summit North America 2023
• 45mThe Benefits of Building New AI Accelerators with RISC-VMartin Maas, Cliff Young · RISC-V Summit North America 2024
• 11mSimulate, trace, and evaluate a RISC-V system leveraging very long vectorsPablo Vizcaino · RISC-V Summit Europe 2024