Why are CPUs becoming important for AI?

CPUs suit agentic AI because they handle sequential workflows, scheduling, and tasks, unlike GPUs which process in parallel. Arm's CEO compares GPUs to dump trucks.

What is the SAKURA-II chip's key spec?

It delivers 60TOPS at 8-bit precision while consuming only 10W, and supports both deep learning and LLMs.

How much does SiMa.ai's edge AI device cost?

The device with MLSoC Modalix is sold for about 20万円 (approximately 200,000 yen), and it consumes only 13W.

Large Language ModelsTop Companies' AI MovesAI Business & IndustryTop Companies AISep 2, 2026Sep 3, 2026, 06:31 JST

AI chips diversify beyond GPUs for agents and edge

Send on LINE

3 Key Points

What happened
AI accelerators that once centered on GPUs now include CPUs, LPUs, and more. Companies like AMD, Google, and startups such as Cerebras and Groq are developing cloud AI chips, while EdgeCortix and SiMa.ai target edge AI.
Why it matters
As AI evolves from deep learning to generative and agentic AI, chips must handle varied tasks. CPUs suit agentic workflows, and LPUs like Groq's speed up token generation. NVIDIA's CUDA software strength keeps it leading in semiconductor sales.
What to watch
AMD's Helios with 6thGen EPYC and MI400 GPUs, paired with Cerebras's Wafer Scale Engine 3, achieved 5x faster token generation speed (Tokens/second/W). EdgeCortix's SAKURA-II offers 60TOPS at 10W, and SiMa.ai's Modalix device costs about 20万円 (Japanese yen).

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Get these free every morning

Context & Analysis

The article traces AI chip evolution from 2012, when Geoffrey Hinton's lab showed GPUs cut image recognition errors, a moment NVIDIA's CEO calls the 'big bang of AI.' That breakthrough led to deep learning and later generative AI, but training took hundreds of days even with thousands of GPUs. This demand for performance drove innovation beyond NVIDIA, yet NVIDIA's lead persists partly because its CUDA software ecosystem makes its chips easier to use. Now, with agentic AI, the article argues CPUs are better suited for workflow-driven tasks, and with physical AI, edge devices need on-device generative capabilities. This diversity means no single chip type dominates. However, designing advanced AI chips at 2nm is so complex that only a few design houses may handle it. That could limit which companies can produce data-center AI chips, potentially benefiting firms like Broadcom, Marvell, MediaTek, and Japan's Socionext, possibly aiding efforts to revive Japan's semiconductor industry.

FAQ

Why are CPUs becoming important for AI?: CPUs suit agentic AI because they handle sequential workflows, scheduling, and tasks, unlike GPUs which process in parallel. Arm's CEO compares GPUs to dump trucks.
What is the SAKURA-II chip's key spec?: It delivers 60TOPS at 8-bit precision while consuming only 10W, and supports both deep learning and LLMs.
How much does SiMa.ai's edge AI device cost?: The device with MLSoC Modalix is sold for about 20万円 (approximately 200,000 yen), and it consumes only 13W.

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.