Nothing really new about this, the most relevant for most people is that Esperanto will hopefully put a lot of love into the open source BOOM repository.
Boom isn't nearly as relevant as what Esperanto is trying to deliver. In fact, one can say Boom is the reason why companies have been reluctant to switch to RISC-V. Experienced engineers don't want a generator or other tools such as Chisel. They want standard synthesizable code and IP cores that were optimized for leading edge processes using standard CAD tools. This is what Esperanto is promising and is far more important than Boom.
Where have I said that there help for boom is the main thing that they are delivering?
There involvment will help and some of their work will be upsteamed and that will help other people do the same. Also the maintainer will work for them and have access to a lot of knowlage and get some time to maintain the project.
Also there are many interesting projects that do care about using better tools such as Chisel, you also don't need 'leading edge processes' for every application either.
Any improvment on Boom will help projects such as Low Risc.
I keep reading articles like this about plans to productionize RISC-V processors. I’m rooting for RISC-V, but from what I’ve read it’s difficult to get a sense of progress here.
Is there anything on the market that offers price/perf close to Intel’s consumer chips? How powerful are RISC-V out-of-order cores compared to Intel’s recent microarchitectures? How about compared to Apple’s ARM cores?
I think this is a problem with your expectations or reading of the evidence.
Esperanto are the only once who really want to offer extreamly high performance and they are not going primarly after Intel, but rather specialised applications.
There are multible companies offering IP and SiFive will realease a linux capable SoC soon, but thats it. LowRisc will hopeful produce hardware in 2018.
In terms of open source there is BOOM and in terms of IP Esperanto will sell you IP for the ET-Maxion that should be fairly high performance.
9 comments
[ 0.31 ms ] story [ 23.1 ms ] threadhttps://github.com/ucb-bar/riscv-boom
That will help Open-Source projects such as LowRisc.
http://www.lowrisc.org/
There involvment will help and some of their work will be upsteamed and that will help other people do the same. Also the maintainer will work for them and have access to a lot of knowlage and get some time to maintain the project.
Also there are many interesting projects that do care about using better tools such as Chisel, you also don't need 'leading edge processes' for every application either.
Any improvment on Boom will help projects such as Low Risc.
Is there anything on the market that offers price/perf close to Intel’s consumer chips? How powerful are RISC-V out-of-order cores compared to Intel’s recent microarchitectures? How about compared to Apple’s ARM cores?
BOOM is not really the same class as Intel, or higher end Apple cores. BOOM targeted something around an ARM A9 in perf.
Esperanto are the only once who really want to offer extreamly high performance and they are not going primarly after Intel, but rather specialised applications.
There are multible companies offering IP and SiFive will realease a linux capable SoC soon, but thats it. LowRisc will hopeful produce hardware in 2018.
In terms of open source there is BOOM and in terms of IP Esperanto will sell you IP for the ET-Maxion that should be fairly high performance.