Zero ASIC’s chief executive Andreas Olofsson has called for “all chip designers” to try out the open-source SiliconCompiler build system, showcasing his top ten reasons for giving it a go – including a provenance feature which supports the Software Package Data Exchange (SPDX) standard.
“We have been working hard on our open source SiliconCompiler build system for a year now and it’s coming together really nicely,” says Olofsson. “I have been doing chip design and CAD development since 1998 (Perl, TCL, make…) and this is light years ahead of anything in the industry. You owe it to yourself to at least try this out!”
In support of the project, which aims to offer compiler-equivalent functionality to chip design in the same way as for software development, Olofsson listed “ten reasons” to try it out: the fact it’s free and open-source; improved support for complexity; 350 pages of “well-crafted documentation” to help users out; a parallel architecture well-suited to cloud computing; a design which offers potential for “massive optimisation;” its programmability; the inclusion of a powerful object-oriented Python API; its ready availability as a PyPi package; and the opportunity to “learn Python.”
The tenth reason: its position as “the first chip build system with a provenance manifest” via a well-defined schema which includes baked-in support for the Software Package Data Exchange (SPDX) standard.
SPDX, adopted as the ISO/IEC 5962:2021 standard in September last year, is designed to ease provenance, licence, security, and other supply chain tracking, communication, and compliance. AB Open’s Andrew Back served as one of the contributors to SPDX 1.0, while representing then-employer British Telecom (BT) at the Linux Foundation.
“Compiling simple programs into silicon should be like using llvm or gcc: fast, automated, and accessible,” project creators Andreas Olofsson, William Ransohoff, Noah Moroze, and Zachary Yedidia explained of their vision late last year. “SiliconCompiler is an open source compiler framework that automates translation from source code to silicon using a standardised compiler data Schema, a Python object oriented API, and a distributed systems execution model.”
More information is available on Olofsson’s LinkedIn post, while the project source code is published to GitHub under the permissive Apache 2.0 license.