Meetups provide an enjoyable, educational atmosphere for Nervana to connect with fellow AI enthusiasts.

Screen Shot 2016-04-28 at 4.06.50 PM

On April 26th, our cofounder and VP of Algorithms, Arjun Bansal, presented “Nervana and the Future of Computing” at Paris Machine Learning Applications Group’s  Meetup #12 Season 3 highlighting ML Hardware. Paris ML has been commended as one of the world’s largest machine learning Meetups with over 3.5k active followers.

Arjun introduced his talk by reporting how The Economist has identified cloud computing, deep learning, and custom hardware as the fundamental elements for the future of computing. He explained how Nervana’s technology delivers fast, scalable AI on demand and pushes the frontiers on each of these components.

Nervana was joined by passionate machine learning speakers from Mozilla,  Mathworks, and The slides for “Nervana and the Future of Computing” can be viewed on our Slideshare page.

On April 21st, we held our second San Diego Deep Learning Meetup sponsored by Teradata in Rancho Bernardo. Over 100 people attended the event to learn about deep learning, neon, and Nervana Cloud.

The event included dinner, one deep dive talk, two lightning talks, and time to network with the machine learning aficionados of San Diego.

Will Constable, one of Nervana’s machine learning engineers, started the evening by explaining the field of deep learning and Nervana’s acclaimed products.



Scott Clark, CEO of SigOpt, explained the Bayesian Optimization methods used by SigOpt. He walked through the technique
of coupling these methods with the “highly scalable deep learning architecture” provided by the Nervana Cloud and neon. He concluded with explicit examples, explaining that SigOpt allows users to easily tune their models to quickly achieve higher accuracy.




Yinyin Liu, one of Nervana’s machine learning engineers, presented the last lightning talk. She started by presenting Fast-RCNN, a model for object detection and localization. After describing the model, she described how to add  a new dataset, PASCAL VOC into neon, to interface with model training and inference.  She continued with a tutorial on how to add a ROI pooling layer into neon and how to build the needed network architecture.


Stay connected with Nervana:

Missed these talks and want a more complete recap? The presentations are on our Youtube, Github and Slideshare pages! You can also follow us on  Twitter to prevent missing our next event. 🙂

If you would like to be personally invited to our Meetups, please be sure to join our Silicon Valley Deep Learning Meetup and San Diego Deep Learning Meetup groups! Have a speaking opportunity for Nervana? Contact us at