AI Infrastructure news and analysis on Web Pulse: 32757 curated articles tracking the latest AI Infrastructure developments, tools, and research, updated continuously from vetted sources.
AI inference has become the fastest-growing market in technology, projected to reach $250 billion within seven years, surpassing the database market. Companies like Anthropic, Google Cloud, Datadog, and Twilio are benefi…
OpenClaw Security is developing new protections for its AI personal assistant platform, including filesystem boundary controls called fs-safe and a network egress routing layer called Proxyline. The platform aims to addr…
Together AI and Pearl Research Labs have partnered to launch a new AI inference endpoint for the Gemma-4-31B-it-pearl model, offering over 25% discounts by offsetting costs through cryptocurrency mining. The Pearl Networ…
Kimi K2.6, the open-weight trillion-parameter model from Moonshot AI, now uses TiDB Cloud to power its agent website builder, supporting millions of agent-created tenants without per-tenant database instances. Moonshot A…
NVIDIA's Vera Rubin platform, powered by the Groq 3 LPX accelerator, addresses the scale-up challenges of agentic AI by delivering deterministic, low-latency execution across thousands of chips for trillion-parameter mod…
Kioxia and Dell have built a 9.8 petabyte all-flash storage server in a 2-rack-unit form factor by packing 40 of Kioxia’s 245.76 TB LC9 QLC SSDs into Dell’s AMD EPYC 9005-powered PowerEdge R7725xd server. The system, whi…
A new generation of enterprise software companies is shifting focus from owning the "system of record" to building "systems of intelligence," according to analysts. These systems sit above databases like CRM platforms, u…
Veeam has declared agentic AI the number one cyber threat, warning that autonomous AI agents now outnumber human employees 82 to 1 and that 97% of those agents carry excessive privileges. The data protection company anno…
HYCU launched aiR, a set of AI Agents that query backup data across more than 100 workloads to identify cyber-resilience risks no single-source tool can detect. The software enables natural language searches across backu…
A global DRAM shortage driven by AI demand is making ransomware recovery more difficult and expensive, according to Assured Data Protection chief strategy officer Stacy Hayes. Hayes argues that supply chain constraints o…
Anthropic and OpenAI have acquired package manager companies Bun and Astral respectively, signaling growing interest from frontier AI labs in these software development tools. Package managers are programs that handle th…
Neon and Databricks Lakebase are redesigning their cloud infrastructure to handle the surge in agent-driven workloads, which are creating databases and scaling compute at rates far exceeding human operators. The companie…
Ben Thompson discussed the implications of the compute shortage on Aggregation Theory and consumer AI during an interview at the MoffettNathanson Media, Internet & Communications Conference. The conversation focused on h…
Venture capital firm a16z is leading Stitch's Series A funding round, marking its first investment in Saudi Arabia. Stitch provides an API-first operating system for financial institutions, including a general ledger, co…
OpenAI Codex users can run multiple agents in parallel using git worktrees, terminal multiplexers, or the official Codex desktop app to avoid branch collisions and speed up workflows. The approach allows developers to ru…
A user's AI session history—containing their unfiltered reasoning, business context, and half-formed ideas—is stored on a vendor's server by default when using hosted AI agents, exposing sensitive professional judgment t…
A state-of-the-art AI email solution would cost between $22 and $130 per month in raw model expenses, with a software company likely charging around $500 per year for a fully agentic product. That price is roughly double…
The GGUF file format consolidates all necessary model components—including weights, chat templates, and special tokens—into a single file, offering a more ergonomic alternative to the scattered JSON files typical of safe…
OpenAI models now run agent turns through the native Codex app-server harness by default in OpenClaw, replacing the previous approach where OpenClaw drove the model loop itself. The change gives Codex ownership of the lo…
PyTorch 2.12 introduces a new device-agnostic `torch.accelerator.Graph` API that unifies graph capture and replay across CUDA, XPU, and out-of-tree backends. The release delivers up to 100x faster batched eigenvalue deco…