• 0 Posts
  • 87 Comments
Joined 2 years ago
cake
Cake day: April 13th, 2024

help-circle





  • I take issue with this forced distinction they are making

    Micron, like Samsung and SK Hynix, already supplies memory chips directly to third-party brands such as G.Skill and ADATA. Even without Crucial-branded kits, Micron DRAM continues to reach consumers through other manufacturers, meaning overall supply remains largely unchanged.

    Nobody ever officially suggested the Crucial supply was likely to shift to the other manufacturers for consumers. On the contrary people expect this to be a step towards a general redistribution of manufacturing capacity towards HBM for parallel compute products.

    By comparison, Samsung exiting SATA SSDs removes an entire class of finished consumer products from one of the world’s largest NAND suppliers. Tom argues that this is why the Samsung move is “worse” for consumers: it directly affects how many drives are available, not just who sells them.

    If you wanted you could make the same argument as for Micron. Who says the Samsung NAND couldn’t be bought by other OEMs to make consumer SSDs. It’s just as possible as the Micron supply shifting to other OEMs who make consumer RAM sticks.

    To me neither are likely. The manufacturing capacity both companies are pulling from the consumer market in both cases is going to go to the higher profit margin parallel compute server market. Neither is worse than the other, they are both equally bad news for us consumers.




  • Yeah I don’t know why anyone entertains the idea.

    Lifting things to LEO still costs around 2000 USD per kg, even with modern cheaper prices thanks to reusable rockets. For a datacenter presumably you’d have to go higher where you have less drag, because you can’t keep doing burns for repositioning. So that sounds like it would already make everything so much more cost prohibitive. And the vibrations of a start are probably also not trivial, if your components are all hardened instead of off the shelf that will cost you more too. I see no world where that’s more economical than buying some cheap land in flyover USA and have truckers drive things there.

    Regarding maintenance there are some approaches where you build more redundancy ahead of time and then let broken things rest in place. At least that was the spiel an Azure evangelist gave us once when I was an intern at a webdev shop (in 2012). But still, once enough breaks down (I think it was a third of components) they would usually then exchange an entire container. So yeah still not great for space.

    The energy I don’t know about really, but at least it doesn’t sound impossible that it could be decent for solar, as long as you can deal with more and more holes in your solar sails over time. At least you wont have to deal with diurnal cycles I guess. But the heating is really the killer issue imho. You’d have to radiate off heat in a massive scale. Heat management for the ISS is fairly complex already. I don’t see how they would efficiently do this on a 5 GW scale. And once again a component level issue: all your cooling from the rack out has to be set up for it. No more fans local to systems, everything is heatpipes that need to connect to the entire spacecraft somehow.







  • I think it’s all performative bullshit, not good policy.

    Some decision maker has to appear innovative to his superiours, so he decides to have some number of locations assigned to a trial group and some bullshit installed. Even if it fails, just as long as he finds the right moment to start appearing critical of the experiment he can still pull off his play. After all moving fast and failing fast are also virtue in modern corporate bullshit lingo.



  • If you look at the desktop, there is AMD and there is Apple silicon

    You can get workstations with Ampere Altra CPUs that use an ARM ISA. It’s not significant in the market, more of a server CPU put in a desktop for developers, but it provides a starting point, from which you could cut down the core count and try to boost the clocks.

    There is also the Qualcomm Snapdragon X Plus with some laptops on the market from mainstream brands already (Asus Zenbook A14, Lenovo ThinkPad T14s Gen 6, Dell Inspiron 5441). That conversely could probably scale up to a desktop design fairly quickly.

    You’re right that we’re not there, but I don’t think we’re that far off either. If Intel keeled over there would be a race to fill the gap and they wouldn’t leave the market to AMD alone.