Archive
Analysis by inference startup Neuralwatt indicates that client-side software harnesses significantly affect GPU energy consumption and prompt caching efficiency during model inference.
Reports suggest significant price hikes for graphics cards are imminent, with a limited opportunity to purchase high-end Nvidia GPUs at their suggested retail price at a gaming convention.
The AI cloud provider is adding the software behind Ray to build a broader infrastructure stack.
The chipmaker's strategic partnership with Ilya Sutskever's research lab couples new capital with access to Vera Rubin computing systems.
The AI chip startup plans to scale production and customer deployments as it builds specialized hardware for high-volume inference.
The model infrastructure company says annual recurring revenue has passed $1 billion as customers deploy more specialized AI systems.
The open-model infrastructure company plans to expand training and inference services as enterprises seek alternatives to closed AI platforms.
The Berkeley company will accelerate work on fault-tolerant machines while pursuing additional federal support for U.S. quantum infrastructure.
Richard Socher's new laboratory is pursuing AI systems that can improve their own research process, backed at a reported $4.65 billion valuation.
The AI chipmaker nearly doubled its offering price as investors sought another route into compute infrastructure.
The six-person startup is training AI to learn digital tasks from video, with efficiency and safety becoming the key tests for its approach.
The company is assembling distributed GPU capacity into a platform that lets developers choose models, hardware and deployment locations.
New developer primitives address memory, subagents, execution and recoverable long-running work.
The first model from Meta Superintelligence Labs powers a redesigned Meta AI experience and can coordinate parallel subagents.
The financing gives the San Francisco AI company an extraordinary capital base, while placing equal pressure on revenue growth, infrastructure execution and product focus.
The Mountain View startup is combining expert labor, evaluations and software infrastructure to improve AI systems after their initial training.
The San Jose chip startup is betting that a shorter, vertical path for electricity can remove a growing bottleneck inside high-performance AI systems.
OpenShell, AI-Q and new Nemotron models aim to give companies more control over autonomous software.
The Santa Clara company reached a $4.2 billion valuation as large AI systems create demand for faster, more flexible connections between processors.
The Mission Bay startup is building laboratory systems that use living neurons for AI workloads, an ambitious alternative to conventional silicon.
The 102.4-terabit chip is designed to move data across increasingly dense computing clusters.
.png)
