
tech
Google unveils two new specialized AI chips to reduce external dependence
Google unveiled TPU 8tx and TPU 8i Max chips on Monday to reduce external dependence for AI workloads.
Key points
- Google launched TPU 8tx and TPU 8i Max chips on Monday for AI workloads.
- TPU 8tx was designed with AMD to secure sensitive data within the hardware layer.
- TPU 8i Max doubles memory capacity and processing power by combining two previous units.
Why it matters. This development signals a shift in the AI hardware landscape as companies seek to secure their data and reduce dependence on a single dominant supplier for specialized computing chips.
Google unveiled two new specialized AI chips, the TPU 8tx and TPU 8i Max, on Monday to reduce reliance on external manufacturers for its cloud computing business. The TPU 8tx, designed with AMD, integrates security directly into the hardware to protect sensitive data and model weights during active use. Meanwhile, the TPU 8i Max combines two previous inference-focused units to double memory capacity, processing power, and bandwidth in a single package.
Latest updates
Google unveils two new specialized AI chips to reduce external dependence
Google on Monday unveiled two new variants of its in-house-developed Tensor Processing Units (TPUs) to broaden its AI hardware portfolio. These specialized chips are designed for additional use cases, including confidential model training and providing greater memory capacity. The move aims to reduce Google's dependence on chips made by other companies for its cloud computing operations. This expansion signals Google's continued investment in its own semiconductor capabilities to support growing AI workloads.
Sources
AI-assisted summary of the reporting above. Quotes are kept to one sentence. Copyright concern? See our DMCA takedown page.