“You have to optimize the whole stack all the way from the power… down to the GPUs… to really reduce your cost per token”

Monetary Matters with Jack Farley · Token Bill Dwarfs CPU Bill | Ben Pouladian on Meta's Muse, Anthropic S1, and Why All Roads Still Lead to NVIDIA

Play from 6:57
6:57

Audio streams from the publisher’s own feed. Verbatim does not host or copy it.

What was described

Advocated optimizing the entire stack - from power delivery to GPU selection and networking - to lower cost per token in data centers

Said nearby on the same subject

“We need to get much more efficient at delivering the power networks… moving to 800 volt DC data center… eliminate so many conversion steps”
PG
Pat Gelsinger
The a16z Show ·
How they work Share this moment
“You’d be surprised how underutilized GPUs are… if you also just aggregate things globally… squeeze out of the available hardware to lower these costs”
♫
Said on Unsupervised Learning with Jacob Effron
Why it worked Share this moment
“They were able to eliminate something like 98% of the total data center cost with these like chicken coop type data centers.”
♫
Said on The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch
Why it worked 98% cost reduction Share this moment
“Some companies have actually saved money by applying AI, but it's being replaced by all the spend with the tokens and data centers…”
JQ
Jim Quick PwC
Full Episodes | Insurtech Leadership Podcast ·
Hard number Share this moment

Moments like this, every Monday

The strongest new moments on your topics, in the speakers' own words. Free for three topics.

No card. No password. One-click unsubscribe.

Track a competitor?

Pro watches up to 25 names or phrases and emails you the morning after they are said.

See Pro, $29 a month

Verbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.