Algorithmia Blog

Emergent // Future – Cloud GPUs, Deploying Deep Learning Models, Applied ML, and more

Issue 43
This week we look at
Google’s new cloud GPUs, how to deploy deep learning models in the cloud, and what applied machine learning looks like at Facebook, Pinterest and others.

🚀 Forwarded from a friend? Sign-up to Emergent // Future here.

👋 Spread the love of E//F on Twitter and Facebook


Google Cloud GPUs

You Might Have Heard: The Google Cloud Platform got a performance boost with the public beta of NVIDIA Tesla K80 GPU virtual machines.

The new Cloud GPUs are integrated with Google’s Cloud Machine Learning service and the cost per GPU is $0.70 per hour in the U.S. and $0.77 in the European and Asian data centers.

So, not cheap. At the same time a Tesla K80 with two cores and 24 GB of Ram will set you back a few thousand dollars.

But did you know… it’s still not as easy as spinning up an EC2 instance or DigitalOcean Droplet. Deploying deep learning models in the cloud is still a challenge due to complex hardware requirements and software dependencies.


Applied Machine Learning

Every time you use Facebook, Instagram or Messenger, you’re experiencing artificial intelligence at work. Some within the social network say FB is so reliant on AI that they cannot exist without it.

So, while the Facebook Artificial Intelligence Research group has been publishing breakthrough after breakthrough on ways to improve the way computers see, hear, and converse, this research doesn’t just find its way into the products.

FB has dispatched an Applied Machine Learning team to create “magical” experiences by leveraging the machine learning algorithms to improve the results of feeds, ads and search results, and to create new text understanding algorithms that keep spam and misleading content at bay.

More importantly, however, because Facebook can’t exist without AI, it needs to empower all of the company’s engineers to integrate machine learning in their work. In other words, algorithm development is broken.

But, thanks to the algorithm economy, every company can now access algorithmic intelligence.

For instance, Google has been applying machine learning to everything from self-driving cards to language translation, and now to the comment sections on sites to make the internet a little less awful. While Pinterest has built a new visual search, Lens, which is basically a point and shoot search engine thanks to deep learning, lots of data, and GPUs now available on many phones.


What We’re Reading 📚


Things To Try At Home 🛠


Emergent // Future is a weekly, hand-curated dispatch exploring technology through the lens of artificial intelligence, data science, and the shape of things to come. 

🚀 Forwarded from a friend? Sign-up to Emergent // Future here.

Follow @EmergentFuture for more on frontier technology

Lovingly curated for you by Algorithmia