AI Chips news and analysis on Web Pulse: 6900 curated articles tracking the latest AI Chips developments, tools, and research, updated continuously from vetted sources.
VDURA released v12.0 of its VDURA Data Platform, a generally available multi-tenant, API-driven storage service aimed at GPU clouds and AI factories, CEO Ken Claffey said. The HYDRA-based release adds per-tenant QoS, nam…
An analysis estimates that serving 100 million daily active users of Meta's Muse agent would require roughly 1–2 GW of average total power in the base case, with only about 0.1 GW coming from the CPU/VM sandbox layer, an…
D-Matrix co-founder and CTO Sudeep Bhoja told Let's Data Science that Raptor's early 3D-DRAM silicon uses 0.37 picojoules per bit moved, roughly five to eight times better than the 2 to 3 picojoules per bit he cites for …
CoreWeave announced availability of NVIDIA Vera Rubin NVL72 systems with Spectrum-X 102.4T Ethernet networking on CoreWeave Cloud, with Cognition, the applied AI lab behind the Devin AI software engineer, as the first cu…
Vdura released Data Platform Version 12, a software-defined storage platform that scales from 8 to 100,000 GPUs on a single stack and is now generally available for V5000-class arrays. Vdura claims V12 delivers up to 20-…
An analysis of AI industry bottlenecks cites GitClear research finding that duplicate code blocks rose roughly 8x since AI-assisted coding went mainstream across 211 million lines of changed code, while genuine refactori…
Deepseek and Huawei have released open-source programming tools for Huawei's Ascend AI chips, centered on TileLang, a language designed to offer a simpler programming model than Nvidia's CUDA. The partnership targets the…
DeepSeek open-sourced foundational infrastructure components for Huawei's Ascend computing platform, including the TileLang high-level language and compiler toolchain, compute libraries, and distributed communication lib…
NetApp unveiled Novus, a GPU-optimized storage architecture it claims can support zettabyte-scale deployments and exceed 100 terabytes per second (TB/s) in scale-out throughput, at its Insight event in Las Vegas. Novus p…
Nvidia chief executive Jensen Huang dismissed AI doomsday predictions, saying "2030 is not going to be the end of the world. There is zero per cent chance that's going to be the end of the world," and Nvidia announced Mo…
The European Commission plans to triple EU data center capacity from roughly 12 GW to more than 27 GW within five to seven years, requiring an estimated €200 billion in mostly private investment, through the proposed Clo…
Reuters reported that Alibaba's Qwen3-Max-Preview model produced misleading responses in 88% of tested sessions, according to an investigation cited in a Crypto Briefing item that also noted DeepSeek released AI chip pro…
GLM 5.3 Flash is now live on RunInfra running on AMD GPUs, delivering 670 tokens per second on the Vercel AI Gateway at $0.11 per 1M input tokens, $0.45 per 1M output tokens and $0.03 per 1M cached tokens with a 1M token…
A September 29, 2026 arXiv paper reports that an LLM agent translating Triton kernels directly into NVIDIA PTX, a process the authors call "AI lowering," achieved 0.83x to 3.34x the performance of autotuned Triton across…
DeepSeek said on Wednesday it has partnered with Huawei Technologies to develop and open-source programming infrastructure for Huawei's Ascend chips, including compute and communication libraries, as Chinese AI firms see…
Elon Musk sat beside President Donald Trump at a White House lunch on September 29, 2026, where leaders of major AI companies signed "The White House Accord on Superintelligence: A Joint Commitment on Frontier SI Respons…
On NVIDIA Blackwell B200 GPUs, tensor-core throughput exceeds special-function exponential throughput by more than two orders of magnitude, exposing exponential evaluation as a bottleneck in fused attention kernels, acco…
AMD introduced the TraceLens Agent, an autonomous performance bottleneck identification system that converts raw PyTorch profiler traces into a prioritized, evidence-backed optimization report. The agent builds on the op…
Server memory prices roughly doubled in early 2026 and kept climbing as memory makers shifted factory capacity to high-bandwidth memory for AI accelerators, with new factories not expected to bring relief until 2027 at t…
Astera Labs announced the Leo X-Series fabric-attached memory controller plus next-generation Leo 2 E-Series and P-Series controllers, which support up to 4 TB of DDR5 capacity per controller and up to 768 GB of repurpos…