Top 7 AI Hardware Releases in Past Year

NVIDIA DGX Series NVIDIA's AI supercomputers, including the DGX Station and DGX A100 chip with 640 GB GPU memory.

[{"selector":"#anim-75535c3c-3347-4f57-9d75-698fb7165904","keyframes":{"transform":["translate3d(-115.45741%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-e46dab47-525b-433f-8c3f-aa59868269b3","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-de992537-a439-45b0-8969-04681f75ad06","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Intel Habana Gaudi2: Intel's deep learning processors, outperforming NVIDIA's A100-80GB GPU.

[{"selector":"#anim-17e35d94-43e5-4578-b023-bf022b1bd9c0","keyframes":{"transform":["translate3d(-113.94659%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5c1455db-b0dc-4b68-84f6-9e7ae60e6e4c","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-c16048c5-1c26-4f32-a639-3e5b58c238a4","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

IBM z16: IBM's Telum Processor-based system enhances AI and cyber resiliency for hybrid clouds.

[{"selector":"#anim-ce103134-2f88-4f15-98aa-6083b84faca7","keyframes":{"transform":["translate3d(-115.14196%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-4871b49f-b9df-44d4-8c0f-4c4c4d887a1f","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5ffa95f0-365a-40f7-8cee-16ff6ea3e138","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

AMD Zen4: AMD's new Zen 4 architecture hints at future AI hardware development.

[{"selector":"#anim-f071412c-40aa-49f9-9a27-6f28af9e0cf1","keyframes":{"transform":["translate3d(-115.14196%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-b1a9fb93-1474-4d57-ba25-8a348a52d23d","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-69075807-3101-48ab-b356-3d40ee3995b0","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Tesla Dojo: Tesla's Dojo chip revolutionizes self-driving car technology.

[{"selector":"#anim-af3b5fca-80f2-461d-bc93-1d889e82a484","keyframes":{"transform":["translate3d(-115.45741%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-27360d66-b42d-416e-a5c4-69b1af3a7c87","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-f9d58c1c-d328-438d-9340-3774244587ac","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Cerebras Andromeda: Cerebras' AI supercomputer, Andromeda, can perform one quintillion operations per second.

[{"selector":"#anim-1f689165-68d1-4182-b035-d98478f0d43d","keyframes":{"transform":["translate3d(-115.14196%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-49d4cf4e-5ad2-4c60-9bec-a61623bfb979","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-7afdde5c-0027-458e-9df8-9218f303c1d2","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

SambaNova SN30: SambaNova's SN30 Datascale offers 12.8X more memory capacity than NVIDIA's DGX A100 for AI and deep learning workloads

[{"selector":"#anim-188457d3-51af-497b-94b4-a29898ec8f96","keyframes":{"transform":["translate3d(-115.55555%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a931aec8-848e-4287-ba65-bd6a1c4c674d","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-e8492f95-d533-4e38-a670-ddaa98c4a7c6","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]