NVIDIA is phasing out older Jetson modules as LPDDR4 supply gets tighter

VIDIA’s older Jetson developer modules are now being phased out as memory shortages spread beyond normal PC parts. The affected boards use LPDDR4 memory, which has become harder to source and more expensive as the wider DRAM market stays under pressure.

Jetson modules are small embedded computers used for robotics, edge AI, cameras, industrial systems, and developer projects. You can think of them as NVIDIA’s AI focused alternative to boards like Raspberry Pi, but with GPU acceleration and NVIDIA’s software stack.

Older TX2 and Xavier Jetson boards are being retired

The affected products are older Jetson models from around the 2017 to 2020 period. These include TX2 and Xavier based modules, all of which rely on LPDDR4 memory.

The phase out schedule has already started. New purchase orders for these products are now non cancelable and non returnable. Final purchase orders are expected by July 1, 2026, while last shipments are planned for July 15, 2027.

NVIDIA and its partners are pushing customers toward newer Jetson platforms such as Orin and Thor. Those newer modules use LPDDR5 memory instead of LPDDR4, but even LPDDR5 is facing pressure because AI datacenters are buying huge amounts of memory.

This matters most for companies that built products around older Jetson boards. Industrial and robotics deployments often need long hardware lifecycles. When a module reaches end of life, teams may need to redesign boards, retest software, update thermal designs, and revalidate systems before moving to a newer platform.

The bigger issue is that memory shortages are now affecting small embedded systems too. Samsung has already moved away from LPDDR4 and LPDDR4X production, which adds more pressure to lower cost and older hardware markets. Raspberry Pi and other single board systems have also felt pricing pressure.

For developers, the message is clear. If you depend on older Jetson TX2 or Xavier modules, migration planning should start now. Orin and Thor may offer better performance, but the shift will not be automatic for every project.

The memory market is forcing hardware makers to retire older designs faster than some customers would like. NVIDIA’s Jetson phase out shows how AI demand is reshaping even the embedded and robotics hardware space, not only GPUs and servers.