No detailed info, and still vapour, you can’t buy it :(
You can certainly put in an order and pay your money now (and I have), for delivery in March or April if you got in early with a promocode given to a number of news sites (for 10 each), or otherwise demonstrate to them that it’s to their advantage to get you one of the first ones.
SeeedStudio is not in the habit of advertising vapour. They’ve got a long track record of delivering the products and I’ve been buying Arduino clones and the like from them for a decade.
HN discussion has more details like price, GPU, etc
Also, made in China…
What’s your point? Lots of things are made in China, and not all of them are terrible. The company behind this has made some decent products in the past…
(I agree with others though that this post doesn’t belong here on lobste.rs…)
This is getting to be a very interesting product if you want to experiment with Linux on RISC-V.
It should handily outperform a Pi 3 or 3+, at only a little more than Pi price. Sure, the Pi is advertised for $35 but it’s quite a lot more than that by the time you add an SD card, power supply, keyboard, mouse, HDMI cable, monitor.
The BeagleV doesn’t come with any of those accessories either, so how is that point relevant?
This is a $100 board, the Pi is a $35 board. The BeagleV does some interesting coprocessors that are cool if you’re doing AI, but I’m skeptical that the DSP processor is going to be effective as a GPU replacement.
It’s relevant because it’s not $100 vs $35 (i.e. three times more), but $100 + N vs $35 + N where N might be anything from $20 to $200 or more depending on whether you want to use one of these as a proper workstation and what you already have lying around that you can re-purpose.
Personally I have a whole bunch of different ARM and AVR and RISC-V SBCs sitting around headless connected to USB hubs or ethernet/wifi and only need a power supply and SD card and USB/ethernet cable for them, but that’s not the common case.
Please let’s not clutter up the front page with product advertising and marketing pages.
RISC-V has the potential to break the stranglehold of a proprietary ISA on hobbyist-class SBCs. This is a development that’s of great interest to me, especially as I was thinking of buying a brace of Pis to rackmount in a home server.
I’m glad you were happy!
Please consider all of the things that lobsters might find neat to buy. Now, consider all of the people who get paid large sums of money to advertise to those lobsters. Now, consider how slowly the front page moves, and how one might exploit this community to market things. Now, consider how tiresome and messy that would get. It’s happened before.
Advertising is the mindkiller, and projects rattling cups (especially when other sites like the orange site are more than happy to let them ply their wares) without giving back here are malicious actors.
Didn’t realize this would be off-topic/spam. But topics such as https://lobste.rs/s/p8e4wd/raspberry_pi_400_70_desktop_pc has been discussed here before, so I’m confused.
I had that one flagged as off-topic, and four merges suggests something off with it.
I personally found this very interesting. RISC-V board I could potentially get my hands on is a big deal that feels very on topic.
Virtual CES seems heavier that way than in-person CES was. Maybe that is pure recency bias on my part.
I’m finding CPUs generally uninteresting when it comes to this kind of product. Sure, they save the manufacturers a few pennies in ARM licensing fees, and it’s nice having more competition in the CPU space, but a CPU is just a small part of a computer.
For a hardware engineer, I can see a reason to like RISC-V. For the programmer, Armv8 vs mips vs risc-v doesn’t matter much: They’re riffs on the same theme. The instruction sets are usually documented well, and it’s not too hard to interface with the processors.
The parts that matter are peripherals. I’d be excited about this if it came with openly documented peripherals. GPUs, network cards, wifi, and everything else. The fact that this has the Nvidia deep learning accelerator makes me very skeptical that this is going to run any relatively off the shelf software, and is probably going to be stuffed full of undocumented or proprietary bits.
NVDLA is all open source and documented.