A NVIDIA logo is shown at SIGGRAPH 2017 in Los Angeles, California, U.S. July 31, 2017. REUTERS/Mike Blake/File Photo
SAN FRANCISCO (Reuters) -Nvidia and artificial intelligence search firm Perplexity on Wednesday said they are partnering with more than a dozen AI firms in Europe and the Middle East to refine those firms' AI technologies and distribute them to local businesses.
Nvidia said it will work with model makers in France, Germany, Italy, Poland, Spain and Sweden to help make AI models in local languages become what are called reasoning models, which are capable of carrying out more complicated tasks.
AI technologies built in English and Chinese have started to shift to that technology, but that transition is more difficult in languages where less training data is available.
Kari Briski, vice president of generative AI software for enterprise at Nvidia, said the company will help model makers generate new data - known as synthetic data - in local languages to help improve them.
"We're doing a lot of synthetic data generation to bring to these low-resource languages and translating our reasoning data so that they can train on it," Briski said in an interview Tuesday. "Europe needs strong models that reflect each nation's unique language and culture."
Once those local models are trained, Perplexity will help distribute them in Europe, where businesses can run them in local data centers and use them to carry out business tasks such as researching a new topic. Perplexity CEO Aravind Srinivas said Germany is already Perplexity's second largest market by revenue.
"That's the kind of system we are heading to in the future, where models are basically doing a few hours' worth of work in one single prompt," Srinivas said.
The deal was part of a number of announcements Nvidia made at an AI conference in Paris on Wednesday. Nvidia and Perplexity did not disclose any financial terms of the deal.
(Reporting by Stephen Nellis in San Francisco; Editing by Sam Holmes)
