NVIDIA launches the DSX platform, continuing to advance into AI factory infrastructure

Jin10 Data

NVIDIA (NVDA.O) announced the NVIDIA DSX platform at the NVIDIA GTC Taipei conference held in Taipei, China, further extending its business layout into the AI factory infrastructure sector.

Unlike the past focus primarily on GPU sales, DSX is aiming to provide enterprises with a complete AI factory solution spanning design, simulation, deployment, and operations management.

As AI model sizes continue to grow, the challenges facing data centers have gone beyond chip performance to include power supply, cooling capacity, resource scheduling, and overall operational efficiency. NVIDIA believes that in the future, the key metrics for competition in the AI industry will gradually shift from individual chip performance to overall infrastructure efficiency—namely, 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 building and operating an AI factory. By using a unified technology stack for compute, software, and facilities, the platform helps customers improve deployment speed, reliability, and operational efficiency, while also reducing the cost of generating tokens during AI inference.

Huang Renxun said:

“ We are not only delivering chips—we are providing every infrastructure builder with a complete methodology system for building AI factories. With the DSX platform, you can simulate the entire factory at no cost, verify performance before installing the first rack, and operate with the reliability required for production-grade AI.”

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

Among them, DSX MaxLPS uses 45°C liquid cooling and rack-level power optimization technologies to increase token output per megawatt of electricity. NVIDIA states that this technology allows up to 40% more GPU deployment with extremely minimal impact on performance, thereby further reducing computing costs under a fixed power budget.

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 a modular software library, APIs, reference designs, and accelerated computing platforms to build a unified software architecture.

In addition to the core software, DSX also integrates multiple existing capabilities. DSX Reference Design provides reference architectures covering compute, network, storage, power supply, and cooling systems; DSX Sim supports digital twin simulation and optimization across the entire process from planning to operations; DSX Flex can dynamically adjust workloads based on changes in grid load and electricity prices; and DSX Exchange enables data collaboration among the compute, network, energy, and cooling systems.

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

The hardware ecosystem is also expanding in parallel. Dell Technologies (DELL.N), Hewlett Packard Enterprise (HPE.N), Lenovo Group (0992.HK), Super Micro Computer (SMCI.O), ASUS, Foxconn, GIGABYTE, Pegatron, and Quanta Cloud Technology are developing NVIDIA DSX-ready systems to help customers build full-stack AI factories.

Meanwhile, DSX Flex has launched commercialization pilot projects together with Emerald AI and Silicon Valley Power to validate the ability of AI factories to dynamically adjust power consumption according to grid demand.

From a strategic perspective, DSX marks NVIDIA’s continued shift from being an AI chip supplier to becoming an AI infrastructure platform provider. By bringing chips, software, data center architecture, operations management, and energy scheduling into a unified system, NVIDIA aims to establish industry standards covering the entire AI factory lifecycle and further strengthen its leading position in the global AI infrastructure market.