Nvidia · Tom's Hardware
Nvidia accelerates end-of-life for some Jetson AI processors due to memory shortages
Compiled by KHAO Editorial — aggregated from 1 outlet. See llms.txt for citation guidance.
◌ Single Source
That's according to Connect Tech, a Canadian supplier and system integrator for AI systems that says Nvidia has moved these devices to Non-Cancelable, Non-Returnable (NCNR) status due to the change. (Shout out to CNX Software for spotting this.)
Key facts
- In any case, it's not that surprising; all of these are older models, with the TX2 having been introduced in 2017 and the Xavier parts originally introduced in 2018, though some of the specific
- That's according to Connect Tech, a Canadian supplier and system integrator for AI systems that says Nvidia has moved these devices to Non-Cancelable, Non-Returnable (NCNR) status due to the change
- Moving from AGavier to AGX Orin is even more straightforward since both use the same 699-pin connector family, although power delivery and thermals still need to be validated
- If you're building projects based on Nvidia Jetson modules, don't panic; it's fairly likely that you're working with the newer Orin and Thor modules
Summary
If you're building projects based on Nvidia Jetson modules, don't panic; it's fairly likely that you're working with the newer Orin and Thor modules. The specific devices affected at Connect Tech are the Jetson TX2 NX, the Jetson TX2i (all SKUs), the Jetson AGavier 32GB Industrial variant, and the Jetson Xavier NX in 8GB and 16GB versions, but the supplier says that Nvidia has marked all TX2 and Xavier models as NCNR on its side. Strictly speaking, those timelines are from Connect Tech, not Nvidia itself, but the company does say that they are "based on Nvidia timelines. Given that, it's probably time to start thinking about moving to newer hardware if you're working with these parts. These are hardly the first casualties of the RAMmageddon, but they do complicate the idea that only DDR5 is under pressure.