‘RISC-V is sloooow – Marcin Juszkiewicz’

Encountered this here on Lemmy a few days ago, haven't looked into it properly. If you search for the article's title, you should find the post and comments.

To my knowledge, modern CPUs have a lot of hardware acceleration for various common algorithms, specifically regarding media. This is orthogonal to the architecture itself, and I'm not sure that risc-v platforms implemented all that stuff, seeing as it's been developed for x86/x64 over decades.

"2030? So still a long time awa-holy shit that's four years from now"

I feel old

Why does your question sound like you are prompting an LLM? I sincerily hope we are not transitioning to communicating like this, because if I have to read stuff like this once again I think I will have to get offline indefinitly.

To answer your question only Nostradamus can chew whatever you are asking mate, I think that thinking about what computer produced in two years you should buy in 4 years might be a cool video idea for an upcoming youtube channel, but absolutely useless for anything practical today.

Here's a video (YouTube) I saw recently that talks about the experience of using a Milk-V Jupiter in 2025.

I have a Star64 from Pine64 that I have been meaning to play with. I would say 2030 may be early to expect daily driving a RISC-V device, but you can start playing with RISC-V and becoming familiar with the system so the transition becomes easier once the platform becomes more mature.

In the first place, consider why you even want to switch to RISC-V. If it's because of an enthusiasm for open-source and hearing the ISA described as open, know that any performant hardware you'll get likely won't be as open as you expect. The SoC won't be open-source, the CPU cores in it won't be open-source, the firmware and bootloader might be an open-source u-boot fork but there's a good chance it's proprietary. Even the actual implemented ISA won't be open since major core designers add custom instructions that aren't part of the RISC-V spec.

Distros like Ubuntu and Fedora seem slated to treat RISC-V as a main architecture that has close to the same number of packages and the same update schedule as x86/ARM by the end of next year, if not sooner. Just like is also the case for ARM, proprietary software like games can run with a nontrivial performance overhead, and other binary software distributed through other channels outside the distro repos (like docker containers, third-party apt/yum repos, or appimage) is often only distributed for x86 even for things that are open-source and can be compiled for other arches without issue.

The software situation can be either a major annoyance or completely seamless depending on how closely you stick to just the distro repos.

Hardware vendors will probably have stuff comparable enough to recent Intel/AMD for desktop in about a year from now. Likely not better, but within the same realm at least. Within another couple years after that you'll almost definitely see more than one of the established major SoC vendors (like Qualcomm, Nvidia, AMD, or Samsung) release something RISC-V in the desktop, server, or mobile space, which is sure to be competitive with x86 and ARM hardware in that space.

Laptops might not see anything good. An alternate ISA can be viable on servers and mobile (both being Linux-first ecosystems), and desktop can easily inherit from stuff made for server, but laptop has unique hardware needs and the market isn't there for vendors to bother investing too much R&D on laptop chips that can't run Windows nor Mac. RISC-V laptops do exist but they're basically taking chips designed for SBC/edge and throwing them in a laptop shell, with the result naturally being awful at power draw since it was never meant to be a good laptop chip, and the iGPU situation is a mess too. That's unlikely to change in the next few years.

I would love to be able to build a Risc V PC. This is currently impossible is it not? The only Risc V machine I've seen for sale is a Raspberry Pi like thing.

is this the nuclear fusion by 2020 of our generation?

I'm not gonna say "don't do it", but I've dug into this deeply and I've turned away from RISC V.

RISC V is slowly being pidgeon-holed into embedded systems. This is not a bad thing, the embedded market (cars, tvs, industrial controls) is huge and diverse.

RISC V has had a very rough start to introduction into bridge-and-bus systems the way we know from Intel/amd because there have simply been too many iterations of CPU registers and capability flags for integrators to take the platform seriously enough to commit to piling a bunch of effort to design, produce, and lead sales on any RISC V platform. Even arm (especially v9) has settled some of these platform issues and is ahead of RISC V in adoption in the integrated platform space (as opposed to embedded).

Long story long, it is extremely difficult to write device drivers for RISC V because one would have to write half a dozen architecture versions, just for a niche platform that barely sells. Conversely, an embedded controller for, say, a vehicle gets a preliminary build and few revisions, ongoing support isn't part of planning the same way.

I'm still checking on videos on youtube where people use currently available boards/laptops and test and compare them to ARMs. In my opinion planning 2027-2030 hardware purchases and a full commitment is still too early.

There will most- ikely be a lot of movement be it chip-wise and with linux support.

As non_burglar said, RISC-V is so far mostly used for small chips for embedded systems. This could change in the future, of course, but you might not have a good RISC-V laptop by 2027-28. There is also a concern that the chips themselves need not be open-source, but there have been open-source designs such as XiangShan, which is comparable to A76.

Currently, your options are DeepComputing's DC-ROMA, which uses a SpacemiT SoC K1, and Framework's Laptop 13, which has a StarFive JH7110. Neither CPU is fully open.

On the software side, there is better news. Debian has accepted RISC-V as one of its main architectures.

It shouldn't be hard by 2030 I imagine; particularly if you primarily or exclusively use open source software. The RVA23 chips announced I usually see people comment them as having synthetic benchmark scores at about the Apple M1 level. I regularly use a laptop with a Skylake dual core in it and a Raspberry Pi 5 run off a microsd rather than a m.2 NVME hat. With that in mind, if RISC-V designs don't get any better than that in the next 4 years, they'll still be better than hardware that I will still be using. I still use a Raspberry Pi 3. At work every now and then I'll throw a gitlab runner on a 10 year old desktop to have another thing building when things are busy

There are RISC-V developer boards today with PCI-E slots that you can throw in pretty much any AMD graphics card. The big distributions Debian, Fedora, Ubuntu, Red Hat - they all support risc-v. felix86 is equivalent to box64 and FEX for x86 to ARM:

https://felix86.com/felix86-26-04/

Software support is solid already today. It's hardware availability for the announced RVA23 designs that's not mature yet. 4 more years and I imagine in most cases the experience of Linux on RISC-V hardware not being much different than on ARM or x86 hardware

compat wise I think many packages are now available for rv specifically or as .noarch.

perf wise, I think we're still a ways off, we're not seeing rv SoCs at the same level of perf / efficiency as arm, and whilst that's just a matter of time, I'm not sure you'll have many compelling offerings even a couple years from now, though potentially in 2030?

you can check in with experiences using devices like the PineTab V or even the custom RV mainboard for the Framework 13. There are also several SoCs produced by SiFive on SBCs, some are card sized, some are mATX. These are primarily positioned as development devices, but they may give you some idea of what things are like right now

Like most ISA's, RISC-V needs it's decades in the R&D oven to get passed certain hurdles.

Everything from design, standards, pre-fabrication, fabrication and manufacturing has to get to a point that makes RISC-V comparable to ARM or even x86.

I think my metric is having hardware available to the consumer that can run heavy productivity suites, like NLE's and real-time media processing in general. Once you get there, albeit through accelerators and subroutines, that's when RISC-V has arrived.