Chip designer MIPS Technologies announced last year that it was pivoting to RISC-V architecture, and earlier this year the company introduced its first designs based on the open instruction set architecture.
Now MIPS says its first processor core is available for licensing. The company says its MIPS eVocore P8700 multiprocessor IP core is a high-performance solution with out-of-order processing and support for multi-core, multi-threaded, multi-cluster solutions.
MIPS says its IP core has already been licensed “for applications including automotive driver assistance systems” and autonomous driving, but that the processor core could also be used for data centers, storage, and high-performance computing.
While I wouldn’t expect these chips to show up in mainstream laptops or desktops anytime soon, it looks like RISC-V is starting to grow up. While companies like Alibaba and SiFive had been the most visible player in the RISC-V space in recent years, in the past few days we’ve seen several companies announce high-performance RISC-V designs for data centers. In addition to the MIPS announcement, a California-based company called Ventana has introduced what it says is the most powerful RISC-V processor to date.
For its part, MIPS says its P8700 processor supports up to 64 clusters, 512 cores, and 1024 threads. The company promises higher single-threaded performance than unnamed “other RISC-V CPU offerings.”
The company also introduced the eVocore I8500 processor system this year, which MIPS says is designed to offer “best in class performance efficiency.” This design can be used with up to 64 clusters, 512 cores, and 2048 threads. But it’s unclear if the I8500 designs are available for licensing yet.
thanks Mark!
2022 will go down as a milestone year for RISC-V.
There have been RISC-V micro-controllers in tech since at least 2016 (and probably before that).
Co-processors doing dedicated tasks in SoCs since at least 2019 (according to Qualcomm this week).
And now we have RISC-V HPC. Not just in design, like Ventana’s, or a few chips for evaluation like Esperanto’s E-1, but powerful chips, available for licensing now, for use in datacentres, auto / lidar, AI, ML, edge processing… in fact, in any area where you might have hitherto considered an x86 or an ARM chip.
I thought Prof. Dr. David Patterson’s recent comment was particularly profound:
“My colleagues at UC Berkeley and I predict that by the end of this decade, the dominant ISA for future product development will be the open RISC-V architecture.”
It’s hard to argue that close-source products are usually more financially secure, which causes them to become better options.
Think of iOS and macOS. This happens with hardware as well, like the A14 and M1 Max.
And then even proprietary devices which are semi-open do pretty good as well. Think of AndroidOS and Windows. Again, this also applies to the hardware like the Dimensity 1100 and Ryzen 7.
It’s the open-source projects that have little market penetration, mindshare, or lack an expected degree of quality. Think of PureOS and Debian OS.
…so you can imagine how it must be for the hardware side of things.
So I have little hope for The Year of Open-Source. We won’t see Linux dominating our phones, or see RISC-V in our devices. At least not openly. They could be used in a transformed manner, which doesn’t share the same open-source spirit.
Why licensing this when Xuantie C910 is open hardware?
Risc-V is not starting to grow up, is has been continuously growing up for several years. It just started from the very beginning.
It’s a figure of speech.