Heat dissipation isn't as big a concern as it seems: the weight of the solar panels is significantly larger than that of the heatsinks you'd need per typical modern gpu
The cooling is not an issue for their current designs. Look at their AI1 Satellite specs. They clearly have the cooling figured out. The thing is that it's not a datacenter, it's a single rack. A single rack that weighs multiple tons.
You can figure out the weight of the thing based on the total power output, and "power to weight ratio" from SpaceX's own diagrams. Then look up how much it costs to launch per ton, and even look up what they are projecting it will cost with Starship. Even if they get costs down, it's still astronomical. I just can't figure out who would pay that much money to put a rack into space. There's no way the power savings are worth it. Unless you have some niche where you need your workload in space, I can't see the value at all.
It's not just the weight some outside groups have done the calculations and to dissipate the heat at the operating temperatures in the white paper without a heat pump to increase the radiator temp and thus efficiency requires a massive pump that would have to move nearly 69,000 kg of coolant per second continuously without fail.
I once did a blind test on myself. A FLAC audio file and a 128Kbit Ogg vorbis file of the same track that I could switch between as I pleased but without knowing which was playing. Yeah, I cannot tell the difference.
I am absolutely sure others can, but not me. I also think credit goes to far better encoders today than what we had 25 years ago.
Hearing also degrades over time. In my 20s I was a lot more fussy about audio formats and hi-fi gear.
Approaching 50, with less sensitive ears and a bit of tinnitus, I’m happy with the convenience of Bluetooth headphones and whatever format Spotify uses.
Opus - perhaps, but claiming that 128kbps Vorbis is transparent would be rather stretching it (unless it's a mono stream); though how easily it will be detectable depends on the kind of music used to test it. However, if you added, say, Bluetooth A2DP to the mix and made it go through a lossy encoder again the difference should be pretty obvious to anyone with good ears.
I did this with a japanese punk band and when I finally got on a private tracker, I redownloaded them and the quality difference was night and day. I have a huge backlog of mid-grade MP3s from TPB back in the day and I'll occasionally download a second flac version of the albums just to get the dynamic range.
A lot of old stuff was also encoded way worse than one might think from the alleged bitrate. A while ago I went through a ton of old files, and everything that sounded shitty (or, far more commonly, stuff I downloaded but didn’t care about) got yeeted.
I can't tell the difference with most headphones, but with monitors or a good system in a good listening environment there are some details that get lost in compression, but there's essentially 0 difference from losslezd if I rip a CD to opus or mp3 rather than from a stream.
Properly encoded Vorbis, Opus or AAC (even Musepack) certainly reach transparency very quick, but who knows what horrible sources were used on non-official YT music? And what dismal remasters?
I use a dual A7X setup with a Sub10 Mk2 in a proper studio listening environment, and I can easily hear the difference between a low‑quality MP3 and a fully lossless WAV/FLAC. I do have trained ears, but even an untrained listener could probably notice that level of difference.
Yep, OP should certainly retest under gnome. Im not in the weeds myself but I get the vibe that gnome's wayland compositor is somewhat faster and more carefully designed than kwin (at least, Emacs in pgtk mode does substantially better under gnome, whereas on Kwin it uses substantial cpu when scrolling and even lags a bit at high resolutions)
Runpod, digital ocean's gpu cloud, and at least a few others use Linux containers for isolation between tenants (look at Wiz's blog post about the nvidia container toolkit bug; digitalocean just puts everyone in a massive k8s cluster)
the vast majority of the target audience of this blog post would only consider moving to SF because of the tech scene. This isn't a mountain biking or asian food blog.
