Akash Systems Deploys Diamond-Cooled GPU Servers
Akash Systems announced the delivery of the first-ever NVIDIA GPU servers cooled with synthetic diamond. The company's "Diamond Cooling" technology is an additive solution for existing air and liquid cooling systems. It reportedly increases GPU compute performance by about 15% in high-temperature data centers by preventing thermal throttling.
- Akash Systems has prominent backing from investors such as Vinod Khosla and Peter Thiel. The company's technology is protected by over 50 patents, creating a significant competitive barrier. - The company's synthetic diamond material is five times more thermally conductive than copper, reducing GPU hotspot temperatures by 10°-20°C even in systems that already utilize liquid cooling. This can lead to a 25% overclocking capability and can double the lifespan of servers. - In December 2024, Akash Systems secured a $27 million contract with NxtGen, a major Indian AI compute solutions provider, to deploy its diamond-cooled servers across their data centers. This deal is projected to cut AI compute costs for NxtGen's customers by over 50%. - Prior to this enterprise expansion, Akash Systems was approved for up to $68 million in funding under the CHIPS and Science Act to support its semiconductor manufacturing. - Enterprise AI adoption faces significant hurdles, including data quality issues, integration with legacy systems, and a shortage of skilled talent. Nearly half of organizations report that concerns about AI accuracy and bias are a top barrier to adoption. - For Chief Revenue Officers (CROs), the primary considerations for adopting new sales technology are its integration with the existing tech stack (like CRM), robust data reporting and analytics, and the potential for a clear return on investment. - Agentic AI architectures, which enable autonomous agents to perform complex tasks, are structured around components for planning, memory, and tool use. Orchestration patterns like "supervisor" (centralized control) and "adaptive network" (decentralized collaboration) determine how these agents interact, impacting factors like cost and latency. - In 2025, the Bay Area's AI sector attracted over $122 billion in funding, representing more than 75% of all U.S. AI investment. However, by early 2026, there has been a 36.73% drop in the number of funding rounds for Bay Area AI companies compared to the same period in 2025.
