Search

Firebird opened the CIS region's largest AI factory in Armenia, powered by NVIDIA accelerated computing and Dell high-performance infrastructure. The site adds a new regional hub for AI workloads.

Lightmatter CEO Nicholas Harris outlined the company's CPO roadmap after joining Nvidia's NVLink ecosystem and its deepening ties to Taiwan's supply chain. The comments come ahead of the 2026 OCP APAC Summit in Taipei on August 11-12.

AMD acquires Taalas, an AI chip startup whose chips etch model weights directly into silicon. Its HC1 test chip reached 16,960 tokens per second on Llama 3.1 8B, far outpacing Nvidia GPUs. The deal aims to lift AMD inference speeds and narrow Nvidia’s lead in high-performance AI services.

Nvidia is reportedly seeking Chinese telecom equipment partners to develop 6G AI-RAN base stations that extend its GPU platform into wireless networks and edge infrastructure.

NVIDIA's Vera Rubin NVL72 compute tray assembles in 1 minute. • 100% automated • No cables, hoses or fans • Faster ramp to revenue

Anthropic confirmed plans to build an in-house silicon team as it scales models and reduces Nvidia dependence, matching moves at OpenAI.

SpaceX is reportedly partnering with Nvidia to equip its Starmind satellites with orbital datacenters.

NVIDIA posts about RTX Spark, describing it as the start of a new era of personal computing.

SpaceX targets over 2 GW of compute by year-end, scaling closer to 10 GW than 5 GW by the end of next year. The company will build exclusively on Nvidia, with Vera Rubin as the planned platform.

SpaceX is partnering with Nvidia to equip Starmind AI1 satellites with Rubin GPUs and Vera CPUs for orbital datacenter compute. Each satellite is said to carry the full Nvidia stack for space-based workloads.

NVIDIA joins the NSF State and Regional AI Infrastructure Hubs program to supply advanced computing, data, software, and expertise for AI research and education across states. The effort launches today.

AI workloads are expanding context windows beyond system memory limits, shifting focus to storage architectures that deliver efficient, secure access to massive datasets for grounded insights.

Nvidia released Alpamayo 2 Super, an open model for robotaxis and autonomous vehicles. The company noted that the hardest problems are the rare, complex situations rather than everyday scenarios.

Over 120 members strong, the Open Secure AI Alliance released new open source security contributions, including proposed SAFE guidelines for converting confidential incident findings into shared standards.

Nvidia GeForce RTX 50 series prices in South Korea are expected to climb as much as 30% starting August 2026. Higher TSMC wafer costs and continued GDDR7 memory price increases are pushing supply chain expenses.

AMD is reportedly raising prices on its graphics cards and memory units, following Nvidia's increases.