Nvidia has introduced DSX POD, a full AI factory reference design built on the Vera Rubin architecture. The system is assembled by partners inside modular containers, with integration and validation completed before delivery. Once the container reaches a site, it can be put into operation after being connected to pre-prepared power, cooling, networking and related infrastructure. Nvidia said that setup time can be reduced from six months to as little as 14 days.
The company said DSX POD shifts the integration and installation of compute, networking, cabling, rack-level power and liquid cooling into manufacturing facilities rather than leaving that work to on-site construction. Nvidia said this reduces the planning, construction, specialized labor and logistics typically required in traditional deployments. It also said factory-tested containers lower the risk of delays tied to component issues and costly, error-prone on-site work, shortening the path to model training and production inference. CEO Jensen Huang said AI factories create value when they start producing tokens, not when construction is finished.
Nvidia on Tuesday introduced DSX POD, a full AI factory reference design based on the Vera Rubin architecture.
According to the company, DSX POD is built by partners inside modular containers, with integration and validation completed before delivery. After the container arrives on site, it can be put into use immediately once it is connected to pre-prepared power, cooling, networking and other supporting infrastructure. Nvidia said that cuts deployment time from six months to as little as 14 days.
Integration work moves into manufacturing facilities
Nvidia said DSX POD moves the integration and installation of compute, networking, cabling, rack-level power and liquid cooling into manufacturing facilities. The company said that reduces the extensive planning, construction, specialized labor and logistics work usually required for traditional on-site builds.
Nvidia also said factory-tested containers reduce the risk of disruption from component delays as well as costly and error-prone on-site operations. In its description, that shortens the path to model training and production inference, allowing customers to begin generating tokens and revenue sooner.
Jensen Huang on deployment speed and value creation
Nvidia CEO Jensen Huang said, 「AI factories create value when they start producing tokens, not when construction is complete.」
He added, 「With DSX POD, we are bringing factory precision and speed to AI infrastructure deployment, reducing deployment time from months to weeks so customers can scale capacity faster and bring revenue-generating production online sooner.」
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
