Google splits its eighth-generation TPU into two chips for training and inference

Poin penting

• Google has separated its eighth-generation Tensor Processing Unit into two distinct chips dedicated to training and inference tasks.
• This hardware split mirrors the broader evolution of artificial intelligence toward multitasking capabilities.
• The strategic shift highlights a growing requirement for specialized processing units to handle complex workloads efficiently across both model training and real-time inference environments.

Diringkas otomatis oleh Kryptokite dari laporan Crypto Briefing. Baca sumber aslinya untuk detail lengkap. Untuk saat ini berita dan ringkasan masih dalam bahasa Inggris.