Distributed Algorithms Developer

• Master’s or Ph.D. in computer science, electrical engineering or related fields (statistics, applied math, computational neuroscience)
• Strong skills in algorithms
• Strong programming skills (C++ and Python preferred)
• Experience writing software for GPUs, HPC or distributed systems is preferred
• Skills or experience in quantitative areas such as machine learning and computer vision is preferred

Nervana benefits

• Medical plans: Health, vision, dental
• Life insurance and 401k plans
• Flexible work schedule
• Great stock options
• Each engineer gets a top of the line Macbook Pro and access to the fastest GPU clusters
• Annual company retreat
• Small teams
• Fun office spaces with many toys!
• Our Palo Alto office is a part of Playground Global & our San Diego office is 10 minutes from beautiful beaches
• Free healthy snacks and beverages

About Nervana

Nervana provides “AI on demand”. Businesses use our Nervana Cloud platform to create and deploy solutions that include natural language processing, image recognition, computer vision and other types of artificial intelligence. We specialize in “deep learning”, which is the technology that powers Apple’s Siri, Facebook face recognition, and Google’s self-driving cars and AlphaGo. The space is white hot right now! A bit more about Nervana:

• Closed a $20M+ Series A round last year from DFJ, Allen & Company, and others.
• Offices in both San Diego and in the ultra-cool Playground space in Palo Alto.
• Current customers in industries such as agribusiness, financial services, health services, automotive, and the intelligence community.

Find out more about Nervana at www.nervanasys.com.

What you will be working on

• Developing a highly-scaleable distributed computing framework
• Implementing various distributed algorithms such as model/data parallel frameworks, parameter servers, dataflow based asynchronous data communication
• Developing a MPI-like communication framework for Deep Learning
• Developing distributed memory constructs
• Ensuring optimal performance: balance job distribution, minimize communication, balance computation vs. communication
• Optimizing code for various computing hardware backends
• Interacting with deep learning researchers to support different distribution frameworks
• Interacting with driver developers and hardware architects to influence low level features