Artificial Intelligence | News, how-tos, features, reviews and videos
Financial pressures of partnering with Nvidia for semiconductors capable of running hyperscale workloads have forced Google, Amazon, Meta and Microsoft into the processor market.
Trillium, the sixth iteration of Google’s Tensor Processing Unit (TPU), is nearly five times more efficient than its predecessor, TPUv5, in peak compute performance and memory bandwidth, Google said.
Reported $700 million acquisition is aimed at helping AI users with workload management and resource allocation.
New edge-optimized processors and FPGAs will power AI-enabled devices in vertical industries including retail, industrial and healthcare.
Plus, Google unveils Axion, its custom Arm-based chip for data centers, at Google Cloud Next 2024.
Infrastructure enhancements targeting AI workloads include updates to compute hardware, new Nvidia GPU offerings, and storage optimization.
At the Intel Vision 2024 event, Intel unveiled the Gaudi 3 accelerator along with AI-optimized Ethernet networking tech, including an AI NIC and AI connectivity chiplets.
Other updates to the service includes new racks, Apigee integration, and survivability features.
When data centers are designed with sustainability in mind, IT teams need to consider the network implications, cooling challenges and performance requirements of AI workloads, a Cisco exec notes.
AI, machine learning, and high-performance computing are creating cooling challenges for data center owners and operators. As rack densities increase and temperatures rise, more data centers are finding ways to add liquid cooling to their facilities.
Sponsored Links