Following up on Jason's note to the gem5-dev list, I wanted to share a brief post I wrote for the SIGARCH blog: https://www.sigarch.org/remembering-nathan-binkert
I think it's safe to say that gem5 would not exist if it were not for Nate. About 15 years ago, when he was my PhD student at Michigan, he wanted to do research on architectures for networking. He needed a full-system simulator with decent network I/O modeling capabilities, but such a thing did not exist. SimOS was pretty much the only non-commercial full-system simulator around, but there were several reasons it wasn't usable. I don't remember them all anymore, but it had some awkward licensing restrictions, was old and unsupported (all the people who worked on it at Stanford had moved on to found VMware), and I don't think it had cross-architecture support (i.e., you needed an Alpha machine to run SimOS/Alpha). Meanwhile in the rest of my group we were using a pretty heavily modified version of SimpleScalar with the Alpha ISA. Most people probably would have just chosen a different thesis topic at that point, but as I said on the SIGARCH blog, Nate was unafraid of any challenge, so he dove in to adding Alpha full-system support to our internal simulator. Nate was also the one who first insisted that we make our new simulator---which by now no longer had any traces of SimpleScalar outside of the EIO trace reader---available as unencumbered open source. And that was the genesis of M5. Nate was also instrumental in the M5/GEMS merger that produced gem5, putting in the lion's share of the early work to integrate Ruby into M5's object structure. Many others have made huge contributions along the way, and gem5 is now much bigger than any one person, but from my perspective Nate is the one who really got it all started. Rest in peace, my friend. Steve
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
