Organized by: Golden Ten Data

NVIDIA (NVDA.O) held the NVIDIA GTC Taipei conference in Taipei, China, where it released the NVIDIA DSX platform, further extending its business layout into the AI factory infrastructure field.

Different from the past focus on GPU sales, DSX aims to provide companies with a complete AI factory solution from design, simulation, deployment to operational management.

As the scale of AI models continues to expand, the challenges faced by data centers are no longer just about chip performance, but also involve power supply, cooling capacity, resource scheduling, and overall operational efficiency. NVIDIA believes that the key indicators for future competition in the AI industry will gradually shift from single chip performance to overall infrastructure efficiency, which is how to produce more computing power and intelligent services under limited power, space, and resource conditions.

To this end, the DSX platform integrates NVIDIA's chips, systems, software, reference architectures, and partner technologies, covering the entire lifecycle of AI factory construction and operation. The platform helps customers improve deployment speed, reliability, and operational efficiency through a unified technology stack of computing, software, and facilities, and reduces the cost of generating tokens during AI inference processes.

Jensen Huang stated:

“We are not just delivering chips—we are providing a complete methodology for every infrastructure builder to create an AI factory. With the DSX platform, you can simulate the entire factory without spending a penny, validate performance before installing the first rack, and operate with the reliability needed for production-level AI.”

The software ecosystem released this time mainly includes DSX MaxLPS and DSX OS.

Among them, DSX MaxLPS uses 45 degrees Celsius liquid cooling and rack-level power optimization technology to increase the token output per megawatt of power. NVIDIA stated that this technology can deploy up to an additional 40% of GPUs with minimal impact on performance, further reducing computing costs under fixed power budgets.

DSX OS is an open-source software platform for AI factory operations, supporting features such as lifecycle management, intelligent scheduling, health automation, multi-tenant operations, and platform services. NVIDIA will also open source modular software libraries, APIs, reference designs, and accelerated computing platforms to build a unified software architecture.

In addition to core software, DSX also integrates multiple existing capabilities. The DSX Reference Design provides reference architectures covering computing, networking, storage, power supply, and cooling systems; DSX Sim supports digital twin simulation and optimization throughout the planning to operation process; DSX Flex dynamically adjusts workloads based on grid load and electricity price changes; DSX Exchange achieves data collaboration between computing, network, energy, and cooling systems.

In terms of commercial landing, cloud service providers such as CoreWeave, Crusoe, IREN, and Lambda have deployed core components of DSX to improve GPU utilization and shorten the time to launch AI cloud services.

The hardware ecosystem is also expanding simultaneously. Companies such as Dell Technologies (DELL.N), HPE (HPE.N), Lenovo Group (0992.HK), Supermicro (SMCI.O), ASUS, Foxconn, GIGABYTE, Pegatron, and Quanta Cloud Technology are developing NVIDIA DSX-ready systems to help customers build full-stack AI factories.

At the same time, DSX Flex has conducted commercialization pilot projects with Emerald AI and Silicon Valley Power to verify the capability of AI factories to dynamically adjust power consumption based on grid demands.

From a strategic perspective, DSX marks NVIDIA's continued transformation from an AI chip supplier to an AI infrastructure platform provider. By integrating chips, software, data center architecture, operational management, and energy scheduling into a unified system, NVIDIA aims to establish industry standards covering the entire lifecycle of AI factories and further consolidate its leading position in the global AI infrastructure market.