NVIDIA Hikes Jetson Pricing by Up to 101%, As the Company Gets Caught Out By Its Own AI Bubble

Driven, most likely, by growing shortages in memory components, NVIDIA passes the pricing pain on to buyers.

Gareth Halfacree
4 months ago β€’ HW101 / AI & Machine Learning

NVIDIA is, apparently, beginning to feel the pain of component shortages caused by the very bubble it is currently in the process of inflating β€” and, as a result, has quietly increased the prices of its Jetson computers-on-modules, in some cases by more than double.

First spotted by VideoCardz, the price hikes have not been communicated via the usual channels but in an unpublicized update to NVIDIA's Jetson frequently-asked questions page. The price hikes, which follow the discontinuation of selected models some weeks ago, see prices rising from 33% on the Jetson TX2i module (which goes from $749 to $999) to a whopping 101% on the Jetson Nano module (going from $99 to $199.) Developer kits, which pair Jetson computers-on-modules with a suitable carrier board and cooling system, have jumped between 57% (Jetson AGX Thor Developer Kit, from $3,499 to $5,499) to 75% (Jetson AGX Orin Developer Kit, from $1,999 to $3,499), with the popular Jetson Orin Nano Super Developer Kit (pictured top) rising 60% (from $249 to $399.)

NVIDIA has not disclosed the reason behind the price increases, nor the previous decision to discontinue selected models in the Jetson embedded compute line-up, but it comes as the company boasts record revenues and provides from the generative artificial intelligence (gen-AI) bubble. Demand for hardware for training new models and running inference continues to climb, even as concerns are raised about the models' usability for real-world tasks, their environmental impact, and the ethics surrounding the wholesale plagiarism of copyright material for their training, leaving memory-makers unable to meet demand β€” and NVIDIA, along with everyone else, having to pay ever-inflating prices for DRAM parts.

The company is far from alone in this β€” Raspberry Pi has been steadily increasing the cost of its eponymous single-board computers over the last year for the very same reason, while coming up with workarounds including the launch of models with lower-capacity RAM and a redesign that allows for the use of two smaller DRAM modules in place of one bigger one β€” but is one of the few companies to be benefiting from the bubble's continued inflation. The companies providing generative AI services themselves, like Meta, Google, OpenAI, and Anthropic, are making massive losses; NVIDIA, which is in the business of selling shovels during a gold rush, is on the other side of the coin with record-breaking profits. With the company accepting some questionable circular financing deals that are dependent on one or more of the AI companies figuring out how to turn gen-AI fervor into bankable profit, though, it remains on shaky ground.

The full list of new price points for the current Jetson line-up is available on the NVIDIA Developer portal; note that entries marked "1KU+" are based on buying tray quantities of 1,000 or more units at a time, with lower volumes demanding a correspondingly higher price.

Gareth Halfacree
Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire: [email protected].

Comments

Please log in or sign up to comment.

Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles