A Chinese startup called Milk-V has begun selling computers with RISC-V processors in China, with plans to expand to the global market soon.

The company’s first product is the Milk-V Duo, a tiny single-board PC with a 1 GHz dual-core chip and a $9 starting price. A unnamed quad-core system with a StarFive JH7110 processor and a Raspberry Pi-like design is coming soon. And the company’s most powerful Milk-V Pioneer, which packs 64 RISC-V cores into a micro ATX motherboard.

Milk-V Pioneer

The MILK-V Pioneer features a Sophon SG2042 processor, which features 64 T-Head XuanTie C920 64-bit CPU cores running at 2 GHz.

The board supports up to 128GB of 4-channel DDR4-3200 memory thanks to 4 DIMM slots and has 5 SATA connectors, two PCIe 3.0 x4 M.2 slots and a microSD card reader that can be used for storage.,

Other ports and connectors include:

  • 1 x M.2 E-Key for wireless card
  • 2 x 2.5 GbE Ethernet ports
  • 8 x USB 3.2 Type-A ports (10 Gbps)
  • 24-pin ATX power connector
  • 1 x PCIe x16 slot for graphics

The board measures 244 x 244mm (9.6″ x 9.6″) and should fit in most cases designed for micro-ATX motherboards.

MILK-V also plans to offer a “Pioneer Box” that comes with 64GB of RAM, 1TB of storage, an Intel X520-T2 network card (with two 10 GbE RJ45 ports), an AMD R5 230 graphics card (with HDMI, VGA, and DVI ports), and a 350 watt MSI A350 power supply as well as a case and cooling fan.

Milk-V Pioneer Box

While the company is based on China, there are also plans to sell its products internationally via AliExpress. And Milk-V plans to launch a Crowd Supply crowdfunding campaign for the Milk-V Pioneer. The company has already published some source code and a CPU user manual to a Github repository, and says hardware schematics and 2D motherboard design files will also be shared before the crowdfunding campaign goes live.

According to Milk-V, the board should work with operating systems including Fedora, Debian, Ubuntu, Arch, and Deepin.

via TechPowerUp, Bret.dk, @MilkV_Official, and Sophon

Support Liliputing

Liliputing's primary sources of revenue are advertising and affiliate links (if you click the "Shop" button at the top of the page and buy something on Amazon, for example, we'll get a small commission).

But there are several ways you can support the site directly even if you're using an ad blocker* and hate online shopping.

Contribute to our Patreon campaign


Contribute via PayPal

* If you are using an ad blocker like uBlock Origin and seeing a pop-up message at the bottom of the screen, we have a guide that may help you disable it.

Subscribe to Liliputing via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 16,199 other subscribers

Join the Conversation


Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  1. I just don’t understand what RISC V is competing with or an alternative to. Is it x86? ARM? They want to be in phones? Servers?

    1. x86 and Arm are both controlled by companies who limit what you can do with them either from a technical, legal or economic perspective.

      RISC-V is free to use so anyone can produce a RISC-V design to serve any kind of market, that’s the appeal.

    2. The point of RISC-V is to be an open spec alternative anywhere a powerful modern CPU architecture could be used. The maintainers of the spec at RISC-V International do not specify where that would be. It is up to anyone to decide to put their own money to do something useful with those specifications, and try to make a profit. So if someone has an idea and skills to produce a RISC-V processor for high end servers, then that is on them to implement. If another party thinks they have figured a lower power RISC-V chip for phones, they take that risk to pursue their ideas. And in either cases, no one has to pay a license to implement the architecture nor be told what they can or can not do, for example such as the ongoing fiasco between Arm and Qualcomm. RISC-V International does not care, as long as the hardware implements documented RISC-V instructions correctly and passes certifications.

      RISC-V is for hardware very similarly to the Linux kernel where anyone can freely download the source code and make use of it any way they choose. There is no direction for Linux either, other than ingest code from anyone to make it work on the hardware of their choice because of their individual needs and want to share their accomplishments. As long as code is good and innovative and proves “useful” to everyone using Linux, then it will most likely be added to kernel code.

      However there is one distinct difference between RISC-V and Linux that is misunderstood and assumed by many when relating the two. Linux GPL license requires anyone that modifies the source code for their product, must make those changes publicly available as well for possible inclusion back into the core code. RISC-V does not enforce such a rule, and any modifications can be added and kept hidden if desired. So for Linux, all parties benefits from others work and new ideas. But for RISC-V, the only real benefactor is the producer of the hardware because they got access to the hard work put into the base hardware designs as a starting point for nothing. So RISC-V is not truly FOSS as Linux like some keep repeating.

    3. As Reagan said below, RISC-V is an open-platform with no-one in charge, AND, no necessity to share architectural improvements to the platform. So it doesn’t grow and benefit everyone like GPL-2 based Linux Distro does. And even if it could, I don’t think it would work, since there isn’t a leadership in place. Too many cooks spoil the broth as they say.

      So while in theory I like Open Source ideology and products, but in practice it doesn’t work that well and doesn’t produce the best product.

      With that said, going full-ham on proprietary isn’t it either. It leads to monopoly, stagnation, and worse for consumer. Case in point; Intel Core i7-7700k in 2017 was abysmal. But it’s still something, more than what can be said about RISC-V in its perpetual infancy.

      So the ideal scenario is a balance of the two. Perhaps a framework or standard which the hardware is based upon, controlled by an entity to lead the improvements, but also freedom within that framework for companies to make improvements, changes, and collaborate. This really sounds like ARM.

      Just see the rapid innovation ARM went from 2009 to 2013. Then in 2014 they established the big.LITTLE processing and offered a lineup of Cortex-A7 (tiny), Cortex-A9 (small), Cortex-A17 (medium) cores. Whilst a few year later in 2016 they had the Cortex-A35 (tiny), Cortex-A53 (small), Cortex-A73 (medium) cores. Now we’ve got the Cortex-Nothing (tiny), Cortex-A510 (small), Cortex-A710 (medium), Cortex-X2 (large). And to go even further beyond, we have Apple’s designs with the S5w (tiny), A14e (smallish), A14p (largish), M1p (LARGE).

      The standard ARM cores are better than their Intel core alternatives, whilst their custom cores (Apple) easily outpace them by 2-3 generations. Not that you can make a proper comparison due to the use of different software base Windows/OS X vs macOS/Android.

  2. I’m still waiting to see if framework or other laptop manufacturers will make a RISC-V laptop. I know of course that desktops can feature more powerful components and are extremely upgradeable, but after I bought my first laptop an eon ago, I just can’t go back to desktops again.

    Despite my other excitement here for things like the Asus Rog Ally and that DOS handheld. It’s cool to oooh and ahhh cool gadgets like that, but my real excitement lies with RISC-V. I’d really love to get my hands on a decent RISC-V laptop.

    If I really did get one and it had full linux support, I would e-cycle what I currently have and probably never look back.

  3. If the price for the bare Pioneer Board is within $500 I will put in an order to try my hand on contributing to the Linux community. That would be too good to pass up for a 64-core desktop. This might finally be the product to kick start RISC-V development into high gear to finally get Linux fully optimized for it and launch into a stable state as a viable platform.

  4. There are calls beneath their twitter announcement of the RPi 3B+ form factor JH7110-based SBC for them to join forces with antmicro to produce a RISC-V compute module product like the arvsom.