Tenstorrent has launched its Galaxy superclusters, a networked AI hardware architecture designed to improve inference speed and capacity for complex AI workloads. The system targets agentic workflows ...
Kopin Corporation and Fabric.AI announced a microLED-based optical interconnect technology at the 2026 AWS conference, aiming to boost AI data center performance by increasing data throughput and redu...
Broadcom has secured a long-term supply deal with Google for next-generation AI chips through 2031, while Marvell is in talks to develop complementary custom AI silicon. These moves highlight intensif...
Marvell Technology Group's stock price has rallied in early 2026 due to its strategic focus on custom AI chips and optical networking hardware. However, the stock faces resistance near $170 amid stron...
FortifAI launched the Nol8 FPGA on April 28, 2026, claiming it can replace up to 60,000 CPUs in AI data centers, offering significant improvements in performance and energy efficiency. The device’s pr...
In March 2026, Microsoft announced a strategic shift to prioritize AI compute resources on its Azure cloud platform, tightening supply and increasing prices for non-AI workloads. This move responds to...
sllm offers a novel approach to making large AI models affordable by sharing GPU nodes among small developer cohorts, promising cost savings, privacy, and compatibility with OpenAI APIs. This model co...
In 2026, AI infrastructure growth is increasingly constrained by energy supply challenges, prompting over $100 billion in investments into energy technologies and new utility models like 'bring your o...
OpenAI announced Symphony, an open-source specification to orchestrate Codex AI agents within issue trackers and development tools. Symphony enables continuous, coordinated AI agent operation to reduc...
Anthropic’s choice to deploy AI models across multiple cloud providers, combined with investor demands for transparency on data center resource use, signals a strategic shift toward more flexible, cos...