You are viewing content from a past/completed QCon

It’s Not Your Machine; It’s Your Code.

What You'll Learn

1Get to see why the standard Linux kernel can hurt your performance at scale, and can prematurely force you into horizontal scaling.

2Learn about the dangers of extensive abstractions in high throughput systems, and how to get around them.

Can we use information in tweets to decide how and when to trade stocks? Yes. But can we do this efficiently at scale, and on a tight budget? Well, this talk will discuss what such a task entails. We will look at some of the performance issues encountered in a simulated high frequency trading scenario, and how we can solve them with a hardware-conscious approach.    

By peeling back layers of software abstractions, we will learn skills and tools to decide how to make the most of our hardware, and how to circumvent limitations imposed by commonly used software.

What are you working on these days?

Lately, I'm working on software for risk management where we try to predict and mitigate the risks associated with creating content at a given time. One such risk for example is what happens when we cannot meet certain production aspects due to COVID-19 restrictions. Based on such factors, we dynamically and intelligently decide how to proceed with the creation process.

And what is the goal of your presentation?

To examine a real-world problem where we might need to challenge our thinking by bypassing popular abstractions and hence regain otherwise hidden performance.

What would you like attendees to walk out with?

Horizontal scaling should not be the default solution to performance issues. Knowing the limitations of your tools and that those limitations can often be circumvented by putting in a little extra work, can usually save resources, time, and money.


Adekunle Adepoju

Senior Software Engineer @Netflix

Adekunle works on the Content Platform Engineering team at Netflix, which is tasked with optimizing and accelerating the process of content creation. This involves using engineering methodologies to turn the complex human processes of content creation, into a more deterministic process in order to deliver appropriately to Netflix's stakeholders (viewers and creators). Currently Adekunle is building systems which find and analyze the risks caused by disruptions in the content creation process.  

Before Netflix, Adekunle worked in various roles where he was involved in software for data center interconnectivity, robotics for semiconductor testing, hardware for oil/gas drilling, and iOS development.  

He holds a degree in Electrical Engineering, and has published research papers in robotics controls systems, and intelligent power systems. His interests involve building cars, teaching, and making complex systems work more efficiently, no matter the domain.

Find Adekunle Adepoju at:

Wednesday Nov 18 / 09:50AM PST (40 minutes )

TRACK Mechanical Sympathy: The Software/Hardware Divide ADD TO CALENDAR Add to calendar

From the same track

View full Schedule

3 weeks of live software engineering content designed around your schedule.

Don’t miss out! Save your seat now