NVIDIA suppliers solve AI rack issues, says next-gen GB300 AI racks will launch in Q3

NVIDIA supply chain partners are accelerating the production of its flagship AI data center racks, after technical issues had delayed shipments, but now these issues have been resolved and supply of GB300 AI server racks will now flow.

2

In a new report from The Financial Times, we're hearing that NVIDIA supply chain partners including Foxconn, Inventec, Dell, and Wistron, have "made a series of breakthroughs" that have allowed them to start shipments of NVIDIA's high Blackwell AI servers. NVIDIA experienced technical problems that started at the end of 2024 that disrupted production, and their annual sales targets.

This new advancement is also seeing NVIDIA laying the groundwork of its next-gen GB300 NVL72 AI server racks, featuring 288GB of HBM3E memory per GPU and offering even more compute power than GB200 NVL72 AI servers, which will handle even more complex reasoning models including OpenAI O1 and DeepSeek R1 models. NVIDIA CEO Jensen Huang said last week that GB300 will launch in Q3 2025.

In order to speed up the deployment of its AI servers, NVIDIA has reportedly compromised on aspects of the GB300's design, where the company had planned a new chip-board layout known as "Cordelia" which allowed for the replacement of individual GPUs. However, in April 2025, NVIDIA told its partners it would be reverting to its earlier "Bianca" board design -- the one in the current GB200 AI servers -- due to installation issues, according to two suppliers that talked to The Financial Times.

NVIDIA isn't moving away from the new Cordelia board completely, as it has told suppliers that it will be implementing the new Cordelia board redesign with its next-gen AI GPUs -- the new Rubin GPU architecture and its new R100 and GR100 AI servers that will be unleashed later this year, and pushed into the market in 2026 and beyond.