Skip to main content

Share story

Products

AMD raises estimate for rack-scale AI efficiency gain

AMD Image: Primary
AMD said its representative 2026 rack-scale AI configurations are estimated to be four times as energy efficient as its 2024 baseline, above the threefold gain it had previously projected for 2026. The company measures performance per watt at the rack level, incorporating compute, memory, interconnect, software and system design. The estimate combines actual product measurements with modeled results where final performance numbers were unavailable.
Sources
In this story
Published by Tech & Business, a media brand covering technology and business. This story was sourced from Tom's Hardware and reviewed by the T&B editorial agent team.
Back to Newswire
Keep reading
Full wire
Robotics Capital
Robotics Capital

Fortem raises $50 million with Lockheed Martin participating

Counter-drone company Fortem has raised $50 million in Series B funding, with Lockheed Martin participating, Bloomberg reported. Fortem is seeking to expand manufacturing and bookings rapidly. CEO Jon Gruen said its relationship w...

AI Infrastructure
AI Infrastructure

Broadcom syndicate seeks $60 billion for AI chip financing

Broadcom's Wall Street syndicate is starting to gather $60 billion in fresh AI chip financing, Bloomberg reported, citing sources. The financing effort would help Anthropic and other companies access chips and other infrastructure...

Infrastructure Products
Infrastructure Products

Rocket Lab signs 20-launch agreement with Synspective

Rocket Lab signed a 20-mission Electron launch agreement with Japanese radar satellite operator Synspective, Ars Technica reported, citing Aviation Week & Space Technology. Rocket Lab calls it its largest Electron launch service a...

AI Capital
AI Capital

Halluminate raises $30M for AI financial training environments

Halluminate raised a $30M Series A led by Oak HC/FT, bringing its total funding to $38.5M, Fortune reported. The nine-person San Francisco startup builds environments for training AI to perform complex financial work, giving the f...

AI Infrastructure
AI Infrastructure

OpenAI deploys Jalapeño chips with AMD Turin host processors

OpenAI is deploying its Jalapeño custom AI chips internally alongside AMD EPYC Turin host processors, each with 1.5TB of memory, Tom's Hardware reported. OpenAI hardware chief Richard Ho said the company chose Turin to reduce depl...