Qubole

Where


Red Bull Media House 1740 Stewart St. Santa Monica, CA 90404

When

Thursday, September 13, 2018
8:55 AM - 7:30 PM PST

 

Schedule a Meeting

Fill out the form below to request a meeting with Qubole onsite

Meet with Qubole at Data Science Salon LA

Qubole's Director of Data Science PM, Piero Cinqugrana, will be onsite to talk about comparing scalability of deep learning frameworks to enable use cases such as voice identification and facial recognition. Catch him onstage at 5:05 PM.

 


Looking for Concrete Ways to Make Your Big Data Initiative A Success? Meet with us at Data Science Salon LA to learn how hundreds of data-savvy organizations are activating their big data strategies. Discover how to: 

  • Activate data for all users -- data engineers, data ops, data analysts, and data scientists
  • Extract more value from your data by adding Machine Learning or Artificial Intelligence
  • Be more efficient by taking advantage of cloud technologies that match your workflows
  • Reduce overall costs (Qubole customers saved $140M in 2017)

 

Agenda

AGENDA

8:55 AM - 9:00 AM    

Registration 

9:00 AM - 1:00 PM 

Speaker Sessions + Coffee Breaks + Schedule a Meeting with Qubole 

1:00 PM - 2:00 PM 

Lunch

2:00 PM - 5:00 PM

Speaker Session Continued + Schedule a Meeting with Qubole

5:05 - 5:35 PM 

COMPARING SCALABILITY OF DEEP LEARNING FRAMEWORKS—A BENCHMARK STUDY presented by Piero Cinquegrana

5:35 PM - 6:00 PM

Schedule a Meeting with Qubole

6:10 PM - 7:30 PM

Closing Reception + Networking


SPEAKER

 
Thumbnail

Piero Cinquegrana

 

Director of Data Science PM, Qubole


Piero is the Data Science Senior PM at Qubole, a cloud-hosted data science platform that scales with the size of your data. Prior to this position, Piero was a Senior Data Scientist for over 5 years at Marketshare/Neustar, a leading marketing cloud firm in Los Angeles. In that role, Piero was a key contributor in launching Marketshare’s TV app and was instrumental in reducing deployment time of the Strategy app by over 30%. Piero Cinquegrana holds a M.A. in Political Science at UCLA where he acquired his passion for data and statistics.

Presenting on: COMPARING SCALABILITY OF DEEP LEARNING FRAMEWORKS—A BENCHMARK STUDY

Deep learning works on large volumes of unstructured data such as human speech, text, and images to enable powerful use cases such as speech to text transcription, voice identification, image classification, facial or object recognition, analysis of sentiment or intent from text, and many more. However, with so many deep learning frameworks available, selecting the right framework is challenging. In this talk, we will present a benchmark study comparing different deep learning frameworks. The purpose of the benchmark is to assess how well these frameworks scale when training on large amounts of data using multiple nodes. Using the ImageNet dataset, we present scenarios both involving CPUs and GPUs for native TensorFlow, TensorFlowOnSpark, Horovod and MXNet and compare how these perform using Synchronous Stochastic Gradient Descent (SGD) either in a parameter server or ring all-reduce architecture.
 

Thumbnail

Mr. Lorem Ipsum

 

Consectetur Adipiscing Elit


Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed at arcu ut leo ornare ullamcorper. Aenean in arcu dictum, feugiat massa nec, cursus turpis. Nulla facilisi. Sed vehicula dictum accumsan. Nam euismod a turpis id pharetra. Etiam pharetra, ex ac placerat porttitor, nunc metus tincidunt tortor, vitae placerat tellus tortor quis mi. Sed vel tellus magna. Fusce vel malesuada magna. Donec lacinia mollis rutrum. Nullam ac augue orci.

Dolor Sit Amet:


Nulla vehicula rutrum sapien, in tincidunt metus tempus ut. Etiam accumsan viverra porta.

  • Nulla pellentesque est eu massa suscipit lobortis.
  • Maecenas lobortis quis urna sit amet auctor.
  • Maecenas rutrum hendrerit enim.
 
 
 
Thumbnail
 
Thumbnail
 
Thumbnail
 

 

Brands that depend on Qubole


     



 

Follow Us!


   


 
 

Invite a Friend