I’m catching up on the news today and see that Nvidia forecasts $1T in GPU sales over the next 18 months.
-
I’m catching up on the news today and see that Nvidia forecasts $1T in GPU sales over the next 18 months. That is crazy. I don’t know how that works at any level.
Also on the nvidia news front, they are making a version of the gpu for orbital datacenters. Now I’ve been around for a long, long time and I’ve heard some pretty dumb ideas in that time, and I feel like orbital datacenters is at least in the top 10.
@jerry they try to force somthing impossible to do because if it was possible, it would be the answer to ignore every single privacy law.
sadly they might be arround yes men, it will not bend what is scientifically possible in space with our current knowledge and technology.
-
@Lily_and_frog @ferrix @jerry I suppose you could try and dig a tunnel into the mantle. That might be worse, but only because we've barely even drilled that far.
-
@jerry The one good thing I can say about orbital datacenters is:
The idea is so incredibly foolish that, no matter how hard they try, they won't get it to work.
Imagine trying to cool a GPU in a vacuum thermos. -
@aredridel @ferrix @jerry and you could make it water cooled, and it would be shielded from radiation, and much easier to install replacement parts
-
I’m catching up on the news today and see that Nvidia forecasts $1T in GPU sales over the next 18 months. That is crazy. I don’t know how that works at any level.
Also on the nvidia news front, they are making a version of the gpu for orbital datacenters. Now I’ve been around for a long, long time and I’ve heard some pretty dumb ideas in that time, and I feel like orbital datacenters is at least in the top 10.
@jerry yeah *NOW* they start believing in solar power.
Of course Dementia Don will probably start talking about building a coal fired power plant.
-
@jerry I am an aerospace engineer and literally have a PhD in this stuff.
I can confirm that space is essentially the worst place imaginable you could locate a datacenter.
Nvidia is 100% cashing in on this trend/bubble just because people are willing to entertain the idea, and by the time it all crashes, Nvidia doesn't care because they already sold their units.@Artemis201 @jerry I thought you liked like a rocket scientist..... must have been the picture oh, and your profile!
-
I’m catching up on the news today and see that Nvidia forecasts $1T in GPU sales over the next 18 months. That is crazy. I don’t know how that works at any level.
Also on the nvidia news front, they are making a version of the gpu for orbital datacenters. Now I’ve been around for a long, long time and I’ve heard some pretty dumb ideas in that time, and I feel like orbital datacenters is at least in the top 10.
I am now terrified to ask what the top 10 list you do have is.
Nothing about the casino economy has any contact with reality. And when we take into consideration that a handful of billionaires own about 80% of all stocks, it’s just an inside game of their manipulating stock prices.
-
@popio @paco @jerry Ethernet itself just uses electricity, so effectively the EM field moves at the speed of light. The real problems would be the voltage drop over such a long wire (so I guess it would have to use a non-standard super high voltage to get that far) and, much more importantly, somehow preventing the interference that would almost be 100% guaranteed to be far outside the range of anything ethernet was ever built for.
Oh, and, you know, storms and such would be... a problem...
Latency would actually be pretty incredible, but, everything going all explodey from time to time might be troublesome.
-
@jerry I am an aerospace engineer and literally have a PhD in this stuff.
I can confirm that space is essentially the worst place imaginable you could locate a datacenter.
Nvidia is 100% cashing in on this trend/bubble just because people are willing to entertain the idea, and by the time it all crashes, Nvidia doesn't care because they already sold their units.Maybe it is just a concept of a GPU. Elon Musk farts out the stupidest idea in the world and the funny money casino spins another round. Consider how much money the tech Bros have and how much ketamine they ingest on a daily basis.
-
@Lily_and_frog @ferrix @jerry I suppose you could try and dig a tunnel into the mantle. That might be worse, but only because we've barely even drilled that far.
@Artemis201 @Lily_and_frog @ferrix @jerry
Hyperloop data! Cue boring company!
-
@aredridel @ferrix @jerry and you could make it water cooled, and it would be shielded from radiation, and much easier to install replacement parts
@Artemis201 @aredridel @jerry ok I'm ready to invest in LavaFlare let's go
-
I’m catching up on the news today and see that Nvidia forecasts $1T in GPU sales over the next 18 months. That is crazy. I don’t know how that works at any level.
Also on the nvidia news front, they are making a version of the gpu for orbital datacenters. Now I’ve been around for a long, long time and I’ve heard some pretty dumb ideas in that time, and I feel like orbital datacenters is at least in the top 10.
@jerry no server will ever make enough money when sent to space. it just really does not make sense.
Datacenters in orbit? Top tier shit idea #1
-
So you’re saying that we clone Elon Musk’s ego?
-
@chieroglyph @jerry you wouldnt be wrong. the size of transistors today is small enough that subatomic particles cause shorts and bitflips, and itll be interesting to see how the shit they're gonna try and solve for that
Encase in lead obvsly
-
@popio @paco @jerry Ethernet itself just uses electricity, so effectively the EM field moves at the speed of light. The real problems would be the voltage drop over such a long wire (so I guess it would have to use a non-standard super high voltage to get that far) and, much more importantly, somehow preventing the interference that would almost be 100% guaranteed to be far outside the range of anything ethernet was ever built for.
Oh, and, you know, storms and such would be... a problem...
Latency would actually be pretty incredible, but, everything going all explodey from time to time might be troublesome.
@nazokiyoubinbou @popio @paco @jerry Actually, the propagation velocity in a coaxial cable is a function of the dielectric constant and for typical materials like Teflon and other fluorinated plastics the propagation velocity is around 0.66 * C.
So over sufficiently long ranges, ignoring resistive losses etc, you will actually see significantly less latency over the air than in a coaxial cable.
The propagation velocity of optical/RF signals in vacuum is C by definition, in air it's very slightly slower but not by enough to matter. In glass it's reduced by the refractive index of the fiber, I don't have typical numbers off the top of my head there.
