Google Unveils Two Next-Gen TPUs to Speed Up AI Inference and Boost Security

At a glance

Google Cloud has added two new systems to its 8th-generation TPU lineup. One adds hardware-based confidential computing to protect enterprise data, while the other expands memory so AI agents can handle large volumes of material at once.

Key points

1. 1

   TPU 8tx is the first Google TPU with hardware-based confidential computing.

2. 2

   TPU 8tx cuts inference latency by up to half compared with TPU 8t.

3. 3

   TPU 8i Max combines two chips and increases memory capacity by 33% over TPU 8i.

4. 4

   Google says more than 100 customers already use TPUs.

Why it matters

As AI moves from answering questions to carrying out multistep tasks, competition over AI infrastructure is shifting toward security and memory efficiency, not just raw speed.

Also reported by

• The DecoderGoogle ships four TPU variants in one year as it ramps up competition with Nvidia
• TechCrunchGoogle broadens its AI hardware portfolio with two new specialized chips
• 중앙일보“AI 에이전트 시대, 작업량 수천배” 인프라 특화 나선 구글 [팩플]
• 연합뉴스구글, 신형 AI 칩 'TPU 8i Max' 공개…연산 성능 2배
• 이데일리구글, AI반도체 4종으로 확대…엔비디아 추격 '가속도'
• 조선일보구글, 자체 칩 TPU 기반 제품 2종 추가 공개… 한 해 4종 쏟아냈다
• 한국경제구글, 보안 강화한 새 AI칩 공개…TPU 8tx·TPU 8i MAX 발표
• 연합인포맥스구글, 엔비디아 추격 박차…AI 반도체 'TPU' 제품군 확대