High Performance Computing
topic on 4 shows · 4 statements across 4 episodes
Innovators & Investors 
Latent Space 
Catalyst 
the a16z Podcast
4 statements about High Performance Computing, every show
Leading scientific supercomputers cost up to $700M plus $300M in power
“Some of the most expensive HPCs right now doing science and engineering work are costing six to seven hundred million dollars a pop. And they require perhaps two to three hundred million dollars of electricity to run over a five year period.”
Smith: Hardware manufacturers cannot keep up with high-performance computing demands
“We see a lot of those hardware manufacturers of GPU and other
data center hardware just blazing a trail today.
we're really happy that they're building better solutions, but the reality is they can't keep up with high performance computing needs today.”
Conrad: HPC Hardware Reliability Issues Are Unlikely to Disappear
“Like the hardware problems aren't going away until the underlying vendors fix things. But honestly, I don't think that's likely because you're always pushing the limits at HPC. This is the case of trying to build a supercomputer”
Lisa Su: Hardware bets take 3 to 5 years to prove out
“It, we didn't know. Like, we didn't know at the time when we were making that decision was it gonna be the right bet, but we knew that we had to make that bet, and you really don't figure that out until three to five years out, so your question about how do yo…”