AI Infrastructure news and analysis on Web Pulse: 32753 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
PyTorch 2.11 now enables direct installation of CUDA-enabled PyTorch wheels on aarch64 Linux from PyPI, eliminating the need for custom package indexes and workarounds that previously complicated deployment on NVIDIA GH2…
Apple has largely missed out on the artificial intelligence boom that has lifted sales of its technology peers, yet the company's profits and stock value continue to grow. Unlike competitors funneling cash to Nvidia for …
Dell Technologies announced major updates to its AI Factory product set at Dell Technologies World, including a new AI Data Platform with three internal storage engines—PowerScale, Lightning, and ObjectScale—designed to …
Marc Andreessen told a16z's January 2026 LP meeting that the competitive advantage in artificial intelligence no longer comes from the underlying model, but from the infrastructure and applications built around it. The v…
The ExecuTorch MLX delegate now enables GPU-accelerated inference for PyTorch models on Apple Silicon Macs through Apple's MLX framework. The new backend achieves 3-6x higher throughput on generative AI workloads compare…
Western Digital has introduced the Ultrastar DC HC6100 UltraSMR hard disk drives with post-quantum cryptography (PQC) to protect against "harvest now, decrypt later" attacks, where adversaries collect encrypted data toda…
ChatFeatured, a Toronto-based startup that helps brands influence how they appear in AI search engines, migrated from PlanetScale Postgres to Postgres managed by ClickHouse to unify transactional and analytical workloads…
Redis launched Iris, a Context Engine that provides a single context and memory layer to make enterprise data usable by AI agents, alongside a Flex SSD-based version of Redis. The offering includes three core components …
Vercel engineer Chris Tate developed Zero, a new systems programming language designed specifically for AI agents rather than human developers. The language treats AI agents as first-class users of the entire toolchain, …
A developer created a local proxy server that routes Claude Code traffic to NVIDIA NIM, enabling use of NVIDIA's AI models through the Claude Code interface. The proxy tool, available on GitHub, requires an NVIDIA API ke…
President Trump hosted Nvidia CEO Jensen Huang, Citigroup's Jane Fraser, and Boeing's Kelly Ortberg in the Oval Office ahead of a China summit, arranging for Huang to fly on Air Force One. Trump outlined his dealmaking s…
A 70-billion parameter model can cost less per useful token than an 8-billion parameter model in production inference, because the real cost driver is GPU utilization rather than parameter count. Teams that switch to sma…
Three AI coding tools — OpenAI's Codex CLI, Anthropic's Claude Code, and the Hermes Agent orchestrator — have converged on a shared command format called /goal that lets developers define a "done" state and walk away whi…
Canonical released an experimental Ubuntu 26.04 LTS concept build for CIX P1 Armv9 devices, including the Radxa Orion O6 and Orange Pi 6 Plus. The image uses a Linux 7.0 kernel with open-source drivers and ACPI hardware …
Productivity and infrastructure companies must now design for AI agents as their primary users, not humans, as agents increasingly handle tasks that people once performed manually. Traditional metrics like DAU/MAU ratios…
Hippocratic AI partnered with Modular to integrate the MAX framework into its inference pipelines, achieving sub-500ms mean time to first token and approximately 30% faster P99 end-to-end latency for real-time patient co…
TanStack AI has achieved full compliance with the AG-UI open protocol for agent-to-frontend communication, enabling bidirectional interoperability between any AG-UI client and server. The update makes @tanstack/ai-client…
PyTorch introduced a "fake export" method that allows developers to verify the exportability of large deep learning models using `torch.export` APIs without requiring actual GPU memory. The approach uses fake tensors wit…
Claude is experiencing elevated error rates on requests to multiple models, with the company actively monitoring the incident. Users can subscribe to email and text message notifications to receive updates when incidents…
Cerebras completed its IPO this week, with shares closing at $280 for a $60 billion market capitalization, following a pulled S-1 filing and a major partnership with OpenAI. The company’s CFO Bob Komin stated Cerebras se…