Intelligence Processing Unit (IPU) // CPU, GPU, TPU, IPU, QPU, ?PU

NoAILabs

--

Chips designed for AI

Actually it’s just a huge field of MAD-blocks + on-chip memory)

https://www.graphcore.ai/products/ipu
https://www.graphcore.ai/hs-fs/hubfs/Chip%20Mk.2-1_compressed.png?width=1920&name=Chip%20Mk.2-1_compressed.png

What’s the problem? // make this IPUs real performance close to peak one.

But all good GPUs is the main hardware for AI anyway.

The GPU market is experiencing a shortage, causing significant impact on the AI industry. NVIDIA, the market leader, is releasing annual updates to its AI chips, including the H200 and B100. Intel has announced their Core Ultra processors with integrated neural processing units for AI acceleration. AMD has introduced the MI300 chip for AI training. There are also emerging players like Cerebras challenging conventional chip designs, and projects like NorthPole and NuPIC developing innovative neural architectures. NorthPole integrates computing and memory on a single chip, while NuPIC enables efficient deployment of AI models on accessible CPUs.

https://www.numenta.com/press/2023/09/11/numenta-transforms-the-ai-landscape-with-nupic/

Sign up to discover human stories that deepen your understanding of the world.

Free

Distraction-free reading. No ads.

Organize your knowledge with lists and highlights.

Tell your story. Find your audience.

Membership

Read member-only stories

Support writers you read most

Earn money for your writing

Listen to audio narrations

Read offline with the Medium app

--

--

NoAILabs
NoAILabs

Written by NoAILabs

Tech/biz consulting, analytics, research for founders, startups, corps and govs.

No responses yet

Write a response