FireSim: Open-Source Easy-to-use FPGA-Accelerated Cycle-Exact Hardware Simulation in the Cloud
Abstract
We present FireSim, an easy-to-use, open-source, FPGA-accelerated cycle-accurate hardware simulation platform that runs on Amazon EC2 F1. FireSim automatically transforms and instruments open-hardware designs (e.g. RISC-V Rocket Chip, BOOM, Hwacha, NVDLA, etc.) with the MIDAS framework into fast (10s-100s MHz), deterministic, FPGA-based simulators that enable productive pre-silicon verification and performance validation. To model I/O, FireSim includes synthesizeable and timing-accurate models for standard interfaces like DRAM, Ethernet, UART, and others. By providing a framework to automate the management of FPGA infrastructure, FireSim also lets software developers get a head-start on building software for a novel hardware design, by letting these developers interact with the pre-silicon hardware design as they would a virtual machine. In effect, both hardware and software developers work from a single source of truth: the RTL for the hardware design. Originally developed to simulate new datacenter architectures, FireSim is capable of scaling to simulating thousands of multi-core compute nodes, with an optional cycle-accurate network simulation tying them together. Leveraging AWS EC2 F1, FireSim removes the high capex traditionally involved in large-scale FPGA-based simulation, democratizing access to realistic pre-silicon hardware modeling of new designs. For designs that contain RISC-V SoCs, FireSim also provides compatible Linux distributions (Buildroot, Fedora) and automates the process of building complex workloads on top of these Linux distributions. By harnessing a standardized host platform and providing a large amount of automation/tooling, FireSim drastically simplifies the process of building and deploying large-scale FPGA-based hardware simulations. In this talk, we cover the open-source FireSim framework, explore how users can use and modify the existing designs available in FireSim, and show how users can integrate and measure their own hardware designs. Presented by David Biancolin and Alon Amid
