Disclosure: Some links on this page are monetized by the Skimlinks, Amazon, Rakuten Advertising, and eBay, affiliate programs. All prices are subject to change, and this article only reflects the prices available at time of publication.
The ROMA Laptop is the first notebook computer powered by a RISC-V processor. First announced this summer, it’s now available for purchase from Alibaba for $1,499 and up. Sort of.
At this point the manufacturer only plans to ship around 100 units by the end of the year and up to 1,000 more in the first quarter of 2023. And you’ll need to contact the supplier for exact pricing and specification details before placing an order. But at least we know know more about that chip powering the laptop.
The laptop is the result of a partnership between China’s DeepComputing and Xcalibyte, and it’s powered by a new Alibaba T-Head TH1520 processor. It features four 2.5 GHz Xuantie C910 64-bit RISC-V processor cores, a neural processing unit with up to 4 TOPS performance, and an Imagination GPU.
The laptop also has a 14.1 inch, 1920 x 1080 pixel display, 16GB of RAM, 256GB of storage, a backlit keyboard, HDMI, USB-C and USB-A ports, an SD card reader, Ethernet port, WiFi 5 and Bluetooth support, and a 1080p webcam.
At this point the ROMA laptop is probably a notebook that only a developer could love. But it’s still a bit of a landmark device that could eventually pave the way for wider adoption of RISC-V processors in mainstream computers, offering an open alternative to the x86 and ARM processors that currently dominate the space.
via CNX Software, RISC-V blog (1)(2), and Tom’s Hardware
The SOC cores appear comparable to A73.
https://www.anandtech.com/show/15991/hot-chips-2020-live-blog-alibaba-xuantie910-riscv-cpu-300pm-pt
Given the amount of muck being chucked here, it’s fair to suggest that this product announcement has unsettled the comfy establishment. Well… buckle up, Dorothy, there’s much more to come : RISC-V is set to become the primary processor in a lot of consumer devices.
I’m actually not particularly interested in this open-source-OS-running-on-a-published-schematic-of-an-open-specification-ISA laptop. It’s limited edition production run and corresponding high price makes it the purview of the developer. And I’m happy about that : let them have the kit they need to continue their work on rebasing the software needed for the RISC-V platform.
What I am interested in is the use of T-Head’s TH1520 SoC : something of the performance muscle of a top of the range smartphone from 2017, and more than enough grunt for everyday computing. I hope Sipeed slap this little beauty onto a low-profile board that’s suitable for inclusion into Shift phone, or a (much more attractive) Framework laptop, and that they also issue a dock for using it as a SCB.
Actually I’d like to see it upset the establishment more and don’t think it’s doing enough for that. To do that it needs to work on mainline kernels, and have a UEFI, among other things. And for all I know maybe this laptop does, but we can’t expect that out of RISC-V as a whole at this time.
I agree.
China doesn’t want to bolster a good device and community, to make a great open-source product. They will if they have to. But they’re more inclined to take shortcuts every step of the way, which yields something completely different.
What China would like to do, is to bring RISC-V to close-source so no-one can see into the device, install any code they want, and be out of the reach of Foreign Influence.
From what can be seen, RISC-V lacks the leadership to bring their technology into the mainstream without it getting corrupted.
They should have Cortex-A73 level of processors designed and stamped out already, and it should already have been shipped in a product with fully open source LinuxOS (Debian, Fedora, Arch) included. The next step would be a vanilla AndroidOS (AOSP), then perhaps a Google-Play Version. Then it’s to get out a functioning UEFI/BIOS to have Windows10/ARM booting up. They can’t just design these theoretically and hope everyone else does the work, they need to build the base platform/framework first to drive their innovation to success. See how 64-bit ARM and later ARMv9 pivoted here.
“mainline kernels, and have a UEFI, among other things.”
they already have it with loongArch
I’m both enthusiastic and concerned about the future of RISC-V, and this product lands more in my concerns.
I’m enthusiastic to see a new open source computer architecture that might avoid some of the downsides that existing architectures bring from private ownership, and closed source designs.
I’m concerned about the influence that China is going to have on the landscape of products that we’re going to end up with when the dust settles.
What kind of products can we, as consumers, expect? Are we going to see 99% of RISC-V products coming out of China with extremely closed-source bootloaders, and security concerns for the western world? Or are we going to see products that want to belong to a larger open-source ecosystem of cross-compatible operating systems and software?
Is the RISC-V Foundation going to be defining the standards of a larger ecosystem of products? Can we expect the RISC-V Foundation to encourage an ecosystem of RISC-V processors and computers that will all share similar BIOS and Bootloader functionality? Or is every manufacturer responsible for their own implementation and compatibility? I suspect the latter, but it’s hard to tell. From my perspective, it seems the goal is just to define the standards of the architecture, and nothing else.
My understanding is that the openness of the RISC-V architecture is related to the relationship between the RISC-V design and the companies that manufacture the chips themselves. Everything else beyond that is up to the implementors?
An OPEN ISA doesn’t imply an open computer or open drivers or open bootloader. In fact it doesn’t imply open CPU or SOC. A RISC-V CPU or SOC can have hidden parts and components as much as creator wants or needs.
The only point is that coming from China it could be out of western agencies control. And China doesn’t have direct power over western citizens.
I was fairly certain that was the case.
China’s reasons for wanting to use RISC-V don’t seem to be aligned with my enthusiasm for RISC-V, and I don’t think this laptop represents anything other than being a product that was made in China with a few less western technologies than an x86 or ARM laptop.
China is going to take RISC-V nowhere interesting. They don’t have the semiconductor foundries needed to make competitive products. RISC-V is only interesting to China because it frees them from western-controlled technologies. But that should hardly matter to a Chinese business, unless this technology is being incentivized by the CCP, which I’m sure is the case.
I need laptop with transmeta but in new process 5 or 7nm
Open alternative? This thing isn’t open. No real implementation of RISC-V is open. Or at least, that’s not how it’s going to be. The information needed to produce a RISC-V chip is licensed permissively. Anyone can take it, stick whatever they want into it, remove whatever they want, and keep that proprietary. Of course, removing stuff also means that software compiled for generic RISC-V machines won’t run on your device so you probably don’t want to do that, but you can add whatever proprietary botnet extensions you want and you’re not obliged to distribute your extensions. And once your proprietary implementation of RISC-V begins to dominate the RISC-V market it won’t matter any more than software compiled on other machines doesn’t work on yours, and it’s at that point that you can charge a lot for compilers that use your extensions to make programs that run faster.
dont forget about wifi/ethernet blobs and GPU and bios/uefi ;P
It’s kind of funny how we already know that RISC-V is going to be a horrible mess.
We’ll be lucky if someone brings out a Raspberry Pi type machine that’s only partly locked down.
I fully expect dozens of Chinese SOCs, all with one-off outdated kernels and no updates, ever.