Nvidia restricts consumer GeForce and Titan GPUs from commercial datacenter use via licensing terms.

Nvidia has updated the licensing agreement for driver software accompanying its GeForce and Titan graphics processing units, a change confirmed on Wednesday that is intended to deter the use of these GPUs inside data centers. The update was made when Nvidia introduced the Titan V GPU earlier this month. The change does not apply to Nvidia's Tesla line of GPUs, which are available from cloud providers like Amazon Web Services and Microsoft Azure. GPUs have driven an 85 percent surge in Nvidia's stock price this year and remain popular among artificial intelligence researchers.

The price difference between consumer and data-center options has sparked criticism. The Titan V starts at $3,000, while the new Tesla V100 is available for $8,000. Discussion forums on Reddit in recent days have tried to make sense of the new language in the software end-user licensing agreement. Some critical responses have alleged that Nvidia is making the change to compel customers to use the more expensive Tesla GPUs instead of lower-cost processors.

An Nvidia spokesperson, in an email to CNBC, highlighted several reasons that companies would prefer Tesla GPUs inside data centers, including support and longer life expectancy. The company indicated that customers wanting to use Titan or GeForce GPUs in data centers can contact Nvidia's sales department to discuss their needs. "We expect that, working together with our user base on a case-by-case basis, we will be able to resolve any customer concerns," the spokesperson wrote. Additionally, those who don't download new drivers won't be held to the new terms.

In its full statement to CNBC, Nvidia emphasized that GeForce and Titan GPUs "were never designed for data center deployments with the complex hardware, software, and thermal requirements for 24x7 operation, where there are often multi-stack racks." The company noted that Tesla products include a three-year warranty covering data center workloads, Nvidia enterprise support, guaranteed continuity of supply, and extended SKU life expectancy for data center components. Nvidia also stated: "We recognize that researchers often adapt GeForce and TITAN products for non-commercial uses or other research uses that do not operate at data center scale. NVIDIA does not intend to prohibit such uses."