GPUs aren’t always your best bet, Twitter ML tests suggest

GPUs are a powerful tool for machine-learning workloads, though they’re not necessarily the right tool for every AI job, according to Michael Bronstein, Twitter’s head of graph learning research. His team recently showed Graphcore’s AI hardware offered an “order of magnitude speedup when comparing a single IPU processor to an …

[Pascal Monett]

"Twitter’s head of graph learning research"

So Twitter has a graph learning research division.

Is it for finding a way to do content moderation ? Because if not, I don't really see the point.

[anonymous boring coward]

" are commonly used in social networks and recommender systems"

How thrilling...

[Gordon 10]

So basically a custom processor is better than a GPU - well duh! This nothing but a vendor sales pitch and hence worthless.

[DrG]

Really

Dedicated tool vendor says dedicated tool is more efficient than general tool at specific task.

...that's a double-dose of obvious with a side of bias...

[pimppetgaeghsr]

I wonder how much Graphcore is paying people for these articles. Maybe a nice dinner out after delivery of some units for investigation?

Seems some of their best employees want out now.