Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

MTIA v1: Meta’s first-generation AI inference accelerator

ai.facebook.com

Post reply on HN
Print view View original

41–50 of 50 posts

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#41

Post by nikhilsimha »

flangola7 wrote: post #32
htrp wrote: post #23

This looks like a customized ASIC specializing solely in recommendation systems possibly focused on ads ranking >We found that GPUs were not always optimal for running Meta’s specific recommendation workloads at the levels of efficiency required at our scale. Our solution to this challenge was to design a family of recommendation-specific Meta Training and Inference Accelerator (MTIA) ASICs.

What a tragic waste of human effort and potential.

I hope they take comfort in the fact that this is open-source.

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#42

Post by throwuwu »

flangola7 wrote: post #32
htrp wrote: post #23

This looks like a customized ASIC specializing solely in recommendation systems possibly focused on ads ranking >We found that GPUs were not always optimal for running Meta’s specific recommendation workloads at the levels of efficiency required at our scale. Our solution to this challenge was to design a family of recommendation-specific Meta Training and Inference Accelerator (MTIA) ASICs.

What a tragic waste of human effort and potential.

Same thing was said about GPUs when they were just for games

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#44

Post by u017937387 »

innagadadavida wrote: post #18
bhouston wrote: post #8

Comparing MTIA v1 vs Google Cloud TPU v4: MTIA v1's specs: The accelerator is fabricated in TSMC 7nm process and runs at 800 MHz, providing 102.4 TOPS at INT8 precision and 51.2 TFLOPS at FP16 precision. It has a thermal design power (TDP) of 25 W. Up to 128 GB of ram LPDDR5. Googles Cloud TPU v4: 275 teraflops (bf16 or int8), 90/170/192 W. 32 GiB of HBM2 RAM, 1200 GBps. From here: https://cloud.google.com/tpu/docs/s…

Is there something that compares these to more consumer offering like Apple’s ANE?

[flagged]

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#47

Post by flangola7 »

throwuwu wrote: post #42
flangola7 wrote: post #32

Earlier quoted context omitted.

What a tragic waste of human effort and potential.

Same thing was said about GPUs when they were just for games

I don't remember that. Games are art and offer at least some benefit to the world. Optimizing Facebook attention algorithms harms society.

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#48

Post by fortyseven »

throwuwu wrote: post #42
flangola7 wrote: post #32

Earlier quoted context omitted.

What a tragic waste of human effort and potential.

Same thing was said about GPUs when they were just for games

That, but the exact opposite. Games were the primary motivation for GPUs in desktop computers in the 90s.

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#49

Post by throwuwu »

flangola7 wrote: post #47
throwuwu wrote: post #42

Earlier quoted context omitted.

Same thing was said about GPUs when they were just for games

I don't remember that. Games are art and offer at least some benefit to the world. Optimizing Facebook attention algorithms harms society.

You must be Gen Z because the prevailing attitude towards video games of all kinds was mostly negative in the 20th century. They even tried to blame Doom for school shootings.

Re: MTIA v1: Meta’s first-generation AI inference accelerator

#50

Post by flangola7 »

throwuwu wrote: post #49
flangola7 wrote: post #47

Earlier quoted context omitted.

I don't remember that. Games are art and offer at least some benefit to the world. Optimizing Facebook attention algorithms harms society.

You must be Gen Z because the prevailing attitude towards video games of all kinds was mostly negative in the 20th century. They even tried to blame Doom for school shootings.

I was born in 89. I'm talking about GPUs specifically, not just video games.
Post reply on HN