241–250 of 377 posts
Getting some NUC-like machines makes a lot more sense to me. You’ll get 2.5Gb/s Ethernet at the least and way more FlOPS as well.
Earlier quoted context omitted.
I did some calculations on this. Procuring a Mac Studio with the latest Mx Ultra processor and maxing out the memory seems to be the most cost effective way to break into 100b+ parameter model space.
Now that we know that Apple has added tensor units to the GPU cores the M5 series of chips will be using, I might be asking myself if I couldn't wait a bit.
Earlier quoted context omitted.
There are an absolutely stunning number of ways to lose a whole bunch of money very quickly if you're not careful renting compute. $3,000 is well under many "oopsie billsies" from cloud providers. And that's outside of the whole "I own it" side of the conversation, where things like latency, control, flexibility, & privacy are all compelling reasons to be willing to spend slightly more. I still run quite a number of…
At the local/hobby scale, it’s very much a “do whatever” area. But I can rent a 4090 for a little under a dollar an hour, and I can rent a b200 for $6, it’s very hard to claim I’ll use 10k+ hours of gpu time on a b2000 I buy for myself.
Like, if you buy that card it can still be processing things for you a decade from now.
Or you can get 3 months of rental time.
---
And yes, there is definitely a point where renting makes more sense because the capital outlay becomes prohibitive, and you're not reasonably capable of consuming the full output of the hardware.
But the cloud is a huge cash cow for a reason... You're paying exorbitant prices to rent compared to the cost of ownership.
brand fanboyism, and vendor locking, the bane of our society the common denominator is always capital gain capitalism is the reason why we haven't been able to go back to the moon and build bases there
and yet spaceX is for-profit and is pushing the boundaries again, after NASA (a not-for-profit government-created monopoly) stagnated. blanket-blaming capitalism without good reasoning is becoming the new red-flag of "can't think critically"
private space companies, despite decades of hype and funding, have stagnated by comparison
the fact that SpaceX depends heavily on government contracts just to function is yet another proof: their "innovation" isn't self sustaining, it's underwritten by taxpayer money
are you denying that NASA landed on the Moon?
Elon psyop doesn't work on me, i know who is behind it all, they need a charismatic sales man for the masses, just like Ford, Disney, Reagan and all, masking structural power with a digestible story for the masses
> blanket-blaming capitalism without good reasoning is becoming the new red-flag of "can't think critically"
it's quite the opposite, people unable to take criticism of capitalism, talk about "critical thinking", how is China doing?
Earlier quoted context omitted.
Bad for your power bill though.
What does a few rpis cost on a monthly basis?
Handy: https://700c.dk/?powercalc
My Pi CM4 NAS with a PCIe switch, SATA and USB3 controllers, 6 SATA SSDs, 2 VMs, 2 LXC containers, and a Nextcloud snap pretty much sits at 17 watts most of the time, hitting 20 when a lot is being asked of it, and 26-27W at absolute max with all I/O and CPU cores pegged. €3.85/mo if I pay ESB, but I like to think that it runs fully off the solar and batteries :)
Earlier quoted context omitted.
I think the only exception is specifically for studying network/communciation-topologies. I've seen a couple clusters (ca. 10-50 Pi's) in universities for both research and teaching.
There are so many network emulators you can use, such as Mininet or GNS3.
Guys, don’t take the claim so literally. He’s a successful tech poster. He makes good money showing off his purchases the good money complaining about how expensive they were. But certainly don’t imitate his choices, his economics aren’t your economics!
There is a reason all the big supercomputers have started using GPUs in the last decade. They are much more efficient. If you want 32bit parallel performance just buy some consumer GPUs and hook them up. If you need 64bit buy some prosumer GPUs like the RTX 6000 Pro and you are done. Nobody is really building CPU clusters these days.
Well, El Capitan uses AMD CPUs (which have integrated GPU capabilities) and it is right on top of the rankings lately. Frontier is right behind it with the same arrangement. Having honest to god dedicated GPUs on their own data bus with their own memory isn't necessarily the fastest way to roll.
For comparison there are 9,988,224 GPU compute units in El Capitan and only 1,051,392 CPU cores. Roughly one CPU core to push data to 10 GPU CUs.