In the roughly decade and a half since the Android mobile operating system appeared on the scene it has been primarily sold on devices with an ARM core at their heart, but along the way it has also appeared for other architectures. If you had a MIPS Android phone you may have been in the minority, but Intel phones enjoyed some popularity, and the up-and-coming new kid in the world of Android is RISC-V. For anyone interested in this last architecture it’s worth looking at the Google Open Source blog, in which they’ve published an overview of the current status of the project.
In short, it’s full steam ahead — as the development environment and emulation is in place for RISC-V Android. It’s certain we’ll start seeing RISC-V phones on the market soon, but perhaps that’s not the part which should interest readers the most. Over the last decade we have seen an explosion of inexpensive ARM single board computers, and though some of them such as the Raspberry Pi owe their heritage to set-top-box SoCs, it’s fair to say that a strong driver for this trend has been the proliferation of powerful mobile chips. A take-up of RISC-V driven by Android would mean a similar explosion of powerful SoCs with those cores, leading we hope to much more accessible and powerful RISC-V computing. Sadly we expect them to still come with proprietary peripherals leading to plenty of closed source blobs, but we can’t have everything.
If you’d like to read more about the whole blob situation and RISC-V, we’ve got you covered.
I’ve still got at least one MIPS Android device around here somewhere. It’ll be good to see another contender again in the low power space.
“A take-up of RISC-V driven by Android would mean a similar explosion of powerful SoCs with those cores”
I’m a huge fan of RISC-V, but I’d argue this is a bit of a stretch. It isn’t a given that fast development of powerful RISC-V cores, but it would be one more incentive towards that. I do think that it’s likely that over time many designs will shift towards RISC-V to ameliorate ARM licensing issues, but I would bet on Android TV boxes and the such before it shifts into low-end smartphones and then maybe high-end ones. I would be very surprised if this last step happened sooner than a decade from now.
RISC-V’s progress has been extremely fast relative to the limitations of hardware design cycles, but there are always naysayers that point out that it’s slower than the most optimistic predictions, and use that to claim that it’s a flash in the pan. Managing expectations gives detractors less ammo.
I would prefer it to be powerpc, decades of optimizations and software already created, let’s see if Raptor’s joint venture project for an ISA 3.1 processor (power 10 equivalent) goes ahead, the possibility of having a version adjusted for mobile would be very interesting.
Alpha for the win!
> we can’t have everything.
yet.
it’s also a matter of investing to make things come true.
because if others do it, they make you pay for it.
The chips from sifive all require closed source blobs to boot, and pretty much every other RISC-V SoC comes from some random chinese company due to their need to break from ARM. I don’t exactly trust their bootloaders.
Now if there was such a thing as an Android release that wasn’t tainted by Google,
this might be good news.