Models & Labs · United States

Microchip, Navitas Unveil 800V Power Reference Design for AI Servers

Credited to TheElec · thelec.net · 1d ago

Microchip Technology and Navitas Semiconductor have unveiled a power-conversion reference design for artificial intelligence (AI) data centers. The design converts 800V direct current supplied to a server rack directly to 6V inside the server.

Microchip Technology and Navitas Semiconductor announced Oct. 6 that they introduced the power-conversion reference design, which combines two stages into one to reduce conversion losses compared to conventional two-stage conversion designs that step 800V down to 50V and then convert it again to 6V. The 6V output is ultimately converted to less than 1V by a voltage regulator module (VRM) on the graphics processing unit (GPU) board.

The board uses 16 gallium nitride (GaN) power semiconductors from Navitas. GaN can switch electricity on and off faster than silicon while generating less loss in the process, allowing power-conversion components to be made smaller while maintaining efficiency. The board is about 20% thinner than a smartphone, allowing it to be positioned close to the GPU board.

Microchip supplies the digital signal controller (DSC) dsPIC33AK, which controls power conversion, and the TA100 security chip. The TA100 handles functions including secure boot and firmware authentication.

Navitas unveiled the board's basic power architecture at Nvidia's GTC 2026 developer conference in March. This time, the companies combined Microchip's controller and security chip to offer a joint reference design. A reference design is a baseline design provided for server and power-equipment companies to use in product development. The two companies are providing the board, software, and design documentation together.

AI data centers are seeing a sharp increase in power consumption per rack as large numbers of GPUs are deployed. Nvidia is building an ecosystem around designs that deliver 800V direct current all the way to the rack, an approach that can reduce power losses and the scale of wiring required.

Nvidia expects the approach to gain wider adoption starting with shipments of its next-generation rack system, "Kyber," in 2027.

Models & LabsAsia-Pacific

As China’s AI race accelerates, ‘model fatigue’ becomes the next challenge

Credited to South China Morning Post$ · scmp.com1d ago

Useful

On September 22, the global artificial intelligence landscape spun on its axis within a matter of hours. In Beijing, Xiaomi engineers went live with an unusual broadcast, streaming the training run of their new MiMo-V2.6 model. In San Francisco, Anthropic unveiled its frontier…

View the article · $South China Morning Post

Front Method

Models & Labs · Americas

Cohere offers to put agents in lockdown mode with strict ACLs

Credited to The Register · theregister.com · 2d ago

Photograph · The Register

While OpenAI and Anthropic stoke fears over rogue agents, Canadian model dev Cohere promises enterprises a safer and more consistent experience. On Monday, the company said it revamped its North agent harness, which launched early last year, to improve collaboration, access control, token spend, and overall security of the platform.

As you'd expect from an enterprise focused agent framework, Cohere's North 2 is designed to tie into existing knowledge bases and SaaS apps in order to automate various business processes.

North 2 introduces a variety of features and functions that should be familiar to anyone who's used Claude Code or OpenClaw. Among them is the concept of skills, libraries, and automations.

Skills are arguably the simplest, and are essentially standard operating procedures documenting how the agent harness should complete tasks. The skills are often generated by working with the agent to complete a task and then documenting which steps achieved the best results.

With North 2, Cohere says that it has made it easier for these skills to be shared so that users aren't wasting time reinventing the wheel. Presumably, this should improve consistency by avoiding having the agent take different routes to achieving its goal, something that might miss key information or nuance.

To facilitate the creation of new skills, North 2 also introduces the concept of automations, generic templates that can be customized using a drag-and-drop graphical interface, not unlike LangFlow, which we looked at earlier this year.

Alongside skills and automations, Cohere's North 2 does something similar for shared knowledge and assets, which it calls libraries, to ensure that agents are working off of the same datasets. Finally, North 2 adds memory, a core feature of most modern agent harnesses, to ensure that agents can maintain context across multiple sessions.

In some sense, these features are a bit like self-generating macros, but rather than being limited to Excel, they work across multiple applications and data sources.

Sovereign AI remains a major talking point for Cohere, which says enterprises can deploy North 2 where they like and use their preferred models. If enterprises want to deploy the harness on-prem in an air-gapped environment, they can. If they prefer to run it in the cloud, that's also an option.

Along with new features aimed at collaboration and automation, North 2 implements new guardrails designed to help enterprises get a handle on their token spend and access controls.

Flow control functionality, meanwhile, promises greater visibility into which departments are hogging all the tokens, while access control functionality offers admins the ability to assign access control lists to various libraries and skills so that interns can't get their hands on proprietary data just because they asked for it.

North 2 is only the latest in a slew of enterprise focused agentic AI platforms which have been rolled out over the last year. Essentially every major and minor cloud provider has introduced some kind of agent harness tailored to enterprise needs. The competition has only gotten hotter over the last few weeks as OpenAI, Meta and others have set their sights on the enterprise. ®