AMD's Taalas Acquisition: 48x Faster Inference, 20x Cheaper, Reshaping AI Chip Landscape
adrianscottcom · x · 2026-08-08
AMD has agreed to acquire Toronto-based AI startup Taalas, potentially disrupting the economics of generative AI. Taalas' HC1 chip etches models like Llama 3.1 directly into silicon, achieving 17,000 tokens per second (48x faster than legacy GPUs), 20x lower build costs (no HBM or liquid cooling), and 10x power efficiency. The 6nm chip has 53 billion transistors, dubbed a 'Hardcore Model.'
The acquisition signals a shift from general-purpose GPUs to specialized ASICs for inference. AMD projects 80% annual growth in AI inference and plans to integrate Taalas tech into its Instinct roadmap, pairing with EPYC CPUs to challenge Nvidia. Industry impacts: inference costs could drop by an order of magnitude, hardware becomes model-specific (e.g., 'Llama chips'), and Google is reportedly exploring similar etched chips. Founder Ljubisa Bajic, former AMD lead architect and Tenstorrent co-founder, returns as a 'homecoming' of elite talent.
Related event: AMD Acquires AI Inference Chip Startup Taalas to Boost Inference Computing(10 posts)→
More from Infra
- GLM-5.2 Hits 180 tok/s in Local Inference, Showcasing Massive Edge Potential — SIGKITTEN · 2026-08-08
- ai& and Voltaiq Partner to Secure Power Reliability for AI Data Centers — DavidBennett__ · 2026-08-08
- Floating Nuclear Reactors Proposed to Power AI Data Centers — markjeffrey · 2026-08-08
- Running Minimax I2V Locally on RTX 5080: Generates Video in 3 Mins — BackgroundAd5676 · 2026-08-08
- Local Qwen Coding Agent on MacBook: Tackling Context & Output Bottlenecks — Techngro · 2026-08-08
- Big Tech's AI Revenue Surges, but Capex Remains Higher Fueling Suppliers — Beth_Kindig · 2026-08-08