AI Infrastructure news and analysis on Web Pulse: 35696 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
Researchers proposed FIBER, a thread-register decoupled GPU execution model that extends the SIMT architecture to improve tensor computation efficiency, achieving a 2.25x end-to-end speedup on Ampere, 1.8x on Hopper, and…
Samsung presented a 16 GB LPDDR5X-PIM memory package at Hot Chips 2026 that delivers 614 GB/s of internal bandwidth to its own compute units, an eightfold increase over the 76.8 GB/s available through external pins, enab…
Token prices could be significantly lower, but inference companies are inflating them by renting GPUs and adding margins of over 40%, according to an unnamed analyst. These companies have current business because they st…
Tailscale announced the general availability of Aperture, its AI gateway for agentic AI, which now includes initial tokens for new instances, MCP endpoints for tailnet control, and an upgraded chat experience with Projec…
Google Cloud has launched Gemini Enterprise Agent Platform, a unified developer platform for building, deploying, and governing AI agents in business environments. The platform integrates agent creation tools, runtime in…
Gartner predicts that by 2028, 33% of enterprise software applications will include agentic AI, up from less than 1% in 2024, enabling 15% of day-to-day work decisions to be made autonomously. Computerworld and Foundry's…
OpenAI announced its first custom inference chip, Jalapeño, on August 25, 2026, claiming it outperforms Nvidia's GB200 and GB300 accelerators by 1.5 to 1.9 times more AI work per watt and 1.7 to 3.6 times lower latency, …
ForgeMesh, a firm that tracks timelines across rail telemetry, grid filings, and catalog censuses, has opened a case file on a twelve-day period in August when a federal banking regulator, a major cloud provider, the Whi…
Cerebras Systems introduced the CS-4, the fastest AI accelerator in the industry, at Supernova 2026, and detailed its architecture at the Hot Chips conference in Palo Alto. The CS-4 is the first system built on the Cereb…
Agentmuxer, a Y Combinator-backed platform from Amorphic Labs, launched on August 25 as an open router for AI agent capabilities, offering unified access to tools like web search, data scraping, and image generation thro…
Nvidia's earnings report on Wednesday is expected to influence tech debt markets, as investors grow wary of bonds issued by major technology companies. The cost of credit-default swaps (CDS) against tech debt has risen, …
Anthropic has agreed to spend roughly $45 billion over six years renting AI computing capacity from Nscale's planned data center campus in Mason County, West Virginia, covering about 460 megawatts of power, according to …
More than 80% of enterprise data is unstructured and 99% is dark to AI, prompting Advanced Micro Devices Inc., Super Micro Computer Inc. and MinIO Inc. to target the data pipeline bottleneck with lakehouse architectures …
AMD announced that its upcoming Versal adaptive SoCs will natively support Universal Chiplet Interconnect Express (UCIe) 1.1 interfaces, with the Versal RF Series first to feature up to four UCIe-SP and two UCIe-AP links…
Perplexity launched Portable Computer on August 25, moving its Computer agent to run fully local on Nvidia's DGX Spark desktop, which costs $4,699, with zero per-token costs for local jobs. The initial release supports Q…
Google released Gemini 3.5 Transcribe on August 26th, a speech-to-text model for live and recorded audio that handles 85+ languages, custom vocabulary, and up to three speakers, with streaming and non-streaming endpoints…
IREN Limited (NASDAQ:IREN) stock fell 6% to $39.85 midday Wednesday as traders de-risked ahead of its fiscal fourth-quarter 2026 earnings report Thursday, while Core Scientific (NASDAQ:CORZ) declined 4% to $17.40 in symp…
Google Cloud on Tuesday launched Gemini Enterprise for Legal, a purpose-built AI platform for law firms and corporate legal departments, now available in preview. The platform integrates with existing legal software such…
OpenAI restored a five-hour usage allowance for ChatGPT Plus subscribers using Codex and ChatGPT Work on Aug. 25, ending a temporary period when only the weekly quota applied. Thibault Sottiaux, an OpenAI engineering lea…
Jim Cramer questioned Nvidia's roughly $30 billion investment in OpenAI ahead of Nvidia's second-quarter earnings report on Wednesday, August 26 at 4:20 PM ET, arguing that the capital is financing a competitor, Jalapeño…