Meta Unveils Next-Gen AI Chip

Meta unveils MTIA chip to lessen reliance on Nvidia and save energy.

[{"selector":"#anim-762b0f2f-1259-4d4f-a28f-dfca229aaaec","keyframes":{"transform":["translate3d(-147.49999%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-1c869330-3a0d-4213-afeb-e5718dca5fa6","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-c2ca2b61-67e4-4547-89c4-e02d8c245fa0","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

MTIA tackles rising AI computing needs across Meta's platforms.

[{"selector":"#anim-7bf61489-59f1-4aaa-9b18-7490c8dc4bca","keyframes":{"transform":["translate3d(-136.50001%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-6211a53d-14e0-460d-956b-7a0df8b8d68d","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-57ed8ebd-a144-45ea-b45f-17b2b3d65cbe","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Meta invests in custom hardware and software for efficient infrastructure use.

[{"selector":"#anim-788180ec-a097-482b-8c4e-54bc27321cb6","keyframes":{"transform":["translate3d(-143.00001%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-02320a05-dbb1-4f74-9ad8-7999f44bbc65","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a7e5b702-6a89-4713-8da4-dac987a608a4","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Zuckerberg plans to acquire 350,000 Nvidia H100 chips to reduce dependency.

[{"selector":"#anim-a462fefb-92fc-45ef-a17c-6eefeb54a25a","keyframes":{"transform":["translate3d(-137.49999%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-198d4775-41d4-4d37-b97b-cfaf2da15b24","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-47466f8a-d198-42d1-be5a-b053bf22ad9d","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

MTIA, from Taiwan Semiconductor, offers triple the performance, supporting generative AI.

[{"selector":"#anim-6e5b9414-e5da-4751-acdf-3c7e9145f09d","keyframes":{"transform":["translate3d(-136.50001%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-ce7a8f26-0567-49fd-a135-ed9f8adf65b7","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-48fd936b-bfbd-42b4-8527-6c39006e0157","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Meta Unveils Next-Gen AI Chip

Meta unveils MTIA chip to lessen reliance on Nvidia and save energy.

MTIA tackles rising AI computing needs across Meta's platforms.

Meta invests in custom hardware and software for efficient infrastructure use.

Zuckerberg plans to acquire 350,000 Nvidia H100 chips to reduce dependency.

MTIA, from Taiwan Semiconductor, offers triple the performance, supporting generative AI.

Read more stories