Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Sunday November 01 2020, @03:04AM   Printer-friendly
from the gimme-five! dept.

SiFive has announced a mini-ITX motherboard with its SiFive Freedom U740:

At the heart of the SiFive board is a SiFive FU740 processor coupled with 8 GB DDR4 memory and 32 MB SPI Flash. It comes with a 4x USB 3.2 ports and a 16x PCIe expansion slot. The mini-ITX standard form factor makes it easy to build a RISC-V PC.

[...] SiFive Unmatched board will be available by Q4'20 for USD 665, and you can already register your interest. You will get a mini-ITX board, 32 GB MicroSD, and 3-meter CAT5e ethernet cable. SiFive did not speak on the commercial aspect of the product but are very confident about future development. Android and Chrome support is something we can see in the future. The product looks promising and we are excited to see future development in the RISC-V PC ecosystem.

Performance will probably be comparable to a Raspberry Pi 3. Alternatively:

PolarBerry is a Compact, Linux-capable RISC-V FPGA SBC and module (Crowdfunding)

Powered by Microchip PolarFire RISC-V SoC FPGA, PolarBerry is both a single board computer with Gigabit Ethernet and 40-pin GPIO header, as well as a system-on-module thanks to three Samtec board-to-board connectors.

[...] PolarBerry is not available just yet, but LinuxGizmos reports the SBC/SoM will be soon launched on Crowd Supply for $995 and shipments are expected to start in January 2021. Besides the aforementioned crowdfunding page, additional details may be found on the product page.

See also: SiFive Is Launching The Most Compelling RISC-V Development Board Yet

Previously: SiFive to Debut a RISC-V PC for Developers in October


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 0) by Anonymous Coward on Sunday November 01 2020, @03:44AM (7 children)

    by Anonymous Coward on Sunday November 01 2020, @03:44AM (#1071541)

    I guess these are the prices you have to pay to have a computer with no hardware backdoor for the alphabet bois to watch you browse pornhub.

    • (Score: 3, Informative) by takyon on Sunday November 01 2020, @05:43AM (5 children)

      by takyon (881) <reversethis-{gro ... s} {ta} {noykat}> on Sunday November 01 2020, @05:43AM (#1071552) Journal

      The prices are down to RISC-V boards being an immature, niche product not aimed at PC users in any way. $665 is also not as expensive as it seems, and less than the previous offering [cnx-software.com].

      The performance is the real bummer. They ought to increase the core count to 16+ to boost multi-core performance per dollar, even if those cores are still very slow compared to x86/ARM. Also, it's a mini-ITX motherboard but there are no DIMM slots, just 8 GB of soldered RAM.

      --
      [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
      • (Score: 3, Insightful) by driverless on Sunday November 01 2020, @10:00AM (4 children)

        by driverless (4770) on Sunday November 01 2020, @10:00AM (#1071570)

        It's not just performance, it's also cost as the OP pointed out. For an architecture whose sole selling point over ARM is that you don't have to pay the ARM tax (vs. endless downsides to do with the fact that ARM is everywhere and supported by and with everything), having to pay $665 for something that might cost $150 on ARM isn't doing a good job of demonstrating its cost advantages.

        • (Score: 3, Insightful) by takyon on Sunday November 01 2020, @10:12AM (1 child)

          by takyon (881) <reversethis-{gro ... s} {ta} {noykat}> on Sunday November 01 2020, @10:12AM (#1071571) Journal

          It's not for Joe Pi User. It's for people getting paid to work with RISC-V to have a native development platform.

          The extremely paranoid are going to have to pay a paranoia tax, for now.

          --
          [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
          • (Score: 2) by driverless on Sunday November 01 2020, @10:29AM

            by driverless (4770) on Sunday November 01 2020, @10:29AM (#1071572)

            It's not for Joe Pi User. It's for people getting paid to work with RISC-V to have a native development platform.

            Sure, but unless there's lots and lots of Joe Users getting RISC-V machines in the near future there's not much incentive to pay people to work with a RISC-V development platform now. I've asked for a RISC-V dev. system a few times but always got the same reply "why should we pay you to work on a platform with no market share?". And I expect lots of others are in the same boat.

            In the meantime if anyone wants to send me a RISC-V dev. system to play with, PM me, they can't tell me what I can do in my spare time...

        • (Score: 2) by EETech1 on Monday November 02 2020, @06:29AM (1 child)

          by EETech1 (957) on Monday November 02 2020, @06:29AM (#1071831)

          I used to spend thousands on development tools to support my projects that used $0.20 Atmel Tinys.
          It's a drop in the bucket when you're developing a high volume custom product.

          Development boards are priceless, they allow you to get your prototype rolling before you have the actual hardware you're developing.

          If you don't feel the need to invest in the up front costs that can save millions on the final product, then by all means, use an ARM..

          • (Score: 2) by driverless on Monday November 02 2020, @09:06AM

            by driverless (4770) on Monday November 02 2020, @09:06AM (#1071846)

            Development boards are priceless, they allow you to get your prototype rolling before you have the actual hardware you're developing.

            Problem with the SiFive devices is that it's unclear when the actual hardware will be available, in what volume, or how much it'll cost when/if it appears. Mouser? A few toy dev boards. Digikey? Same. Element14? Nothing. RS? Nothing. I can't even begin to sell this to management when it's like this. Sure, a dev board is priceless when it's for something that will be everywhere in a few months' time and you get a head start on the competition, but the only thing I can say with reasonable certainty about SiFive silicon is that it won't be everywhere at any point I can foresee, and with that I can't sell it to decision-makers as something worth investing in.

    • (Score: 2) by epitaxial on Monday November 02 2020, @12:22AM

      by epitaxial (3165) on Monday November 02 2020, @12:22AM (#1071722)

      You could always buy a Talos POWER9 based system. Don't get sticker shock though. https://www.raptorcs.com/TALOSII/ [raptorcs.com]

  • (Score: 0) by Anonymous Coward on Sunday November 01 2020, @05:37AM

    by Anonymous Coward on Sunday November 01 2020, @05:37AM (#1071550)

    The photo says 16x, then adds Gen 3 x8. So big slot, but only half the lanes it could have. At least not really old generations, and will be able to drive video cards, just not at full bus "speed".

  • (Score: 3, Interesting) by bzipitidoo on Sunday November 01 2020, @02:21PM (2 children)

    by bzipitidoo (4388) on Sunday November 01 2020, @02:21PM (#1071592) Journal

    If it's possible to really put an entire System on a Chip, why do they keep putting these chips on boards? To better spread out and dissipate heat? Is a printed circuit board still the best way to connect chips to peripherals? Seems an area worth a hard look.

    • (Score: 3, Informative) by takyon on Sunday November 01 2020, @02:37PM

      by takyon (881) <reversethis-{gro ... s} {ta} {noykat}> on Sunday November 01 2020, @02:37PM (#1071594) Journal

      A bigger board can have more connectors, and fit into standard mini-ITX cases. It could also have room for RAM slots, but they didn't do that here.

      --
      [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
    • (Score: 2) by Grishnakh on Sunday November 01 2020, @05:45PM

      by Grishnakh (2831) on Sunday November 01 2020, @05:45PM (#1071621)

      If it's possible to really put an entire System on a Chip, why do they keep putting these chips on boards? To better spread out and dissipate heat? Is a printed circuit board still the best way to connect chips to peripherals? Seems an area worth a hard look.

      It's not about heat; PCBs are made of fiberglass, which is not a very good heat conductor, though layers of copper can be added to conduct heat intentionally. PCBs are used because no one's come up with a better way of connecting the chip to all the things it needs to connect to: external memory (even "SoCs" still generally have external memory, so the "system on a chip" moniker isn't really accurate), connectors, peripherals like cameras, etc. Back in the old days, before PCBs were invented, electronic components were connected together with point-to-point wiring. So there was some kind of metal chassis holding major components, and lots of hand-soldered wires connecting everything together. It required an enormous amount of labor to make an electronic assembly, and mistakes were common and quality was poor because of human error. PCBs revolutionized all this, because this moved most of the "wiring" to a single, mass-produced board, and all humans had to do was insert the components into the appropriate holes in the board and solder them. It also reduced the need for a metal chassis to hold the components, as most of them just sat on the board, except for very heavy ones (transformers) or high-heat ones (large power transistors). Later, more automated ways were devised for "stuffing" the components onto the boards, and later, "surface mount" electronics were invented where the components just sat on top of the board, adhered only by solder. This is where we are currently: mass-produced boards are loaded into an automated assembly line where "pick and place" machines automatically stick the components onto the appropriate location on the board (aided by "solder paste", a stick paste with solder in it, which the components stick to), and then after all the components are placed, the whole thing goes into a "reflow oven" which melts all the solder paste and all the components are now fully soldered to their final locations. There's very little labor that goes into a modern electronic device now, which is why TVs don't cost nearly as much as cars, unlike back in the early days of TV.

      Anyway, back to your question: what I described here is the state of the art, and just keeps getting more and more miniaturized. No one's come up with some alternate way of connecting electronics together other than fiberglass-based PCBs which works as well in terms of cost, ease (and automatability) of assembly, and reliability. There are some more specialized kinds of circuit boards out there that get used in special applications (like military or spacecraft), and some of this technology may filter down to consumer-level stuff: different PCB materials, implanting capacitors and resistors into the board itself, etc., but I don't think PCBs are going anywhere for a very long time.

(1)