Introducing the Inaugural Ampere Developer Summit!

As @pjbaker wrote a couple of weeks ago, we are hosting a virtual Ampere Developer Summit in a few weeks, on September the 26th. You can see the event page and register here - there is a lot of information about sessions and speakers on the page, if you click through to see them.

In the meantime, I wanted to share a high-level overview of the event, the guest speakers we are inviting to join us, the topics we will cover - and hopefully give you some more reasons to sign up today! Incidentally: there is a “care package” of conference swag coming for everyone who registers and attends on the day.

We just posted a blog post to share the content line-up - Unveiling the Ampere Developer Summit Schedule - but here, I want to leave a high-level summary here and invite you to find out more over at the blog post!

The idea of this year’s event is to tell the story of migrating Ampere (or any Arm64 processor) for your applications. After an opening keynote sharing how far we’ve come, and where we see the industry going in the next few years, we will delve into why you should consider Ampere as a deployment platform, which applications are a good fit, and how you can make that change painlessly.

Because we know that the architecture is something that is not top of mind for application developers, we will share ways in which high-core-density and low power matter to application performance. Sean Varley, Chief Evangelist at Ampere, will learn from Kate Goldenring of Fermyon how new Compute footprints like WASM, combined with Ampere, can enable you to get scalable, high performance, and power efficient serverless workloads. Dor Laor, CEO and co-founder of ScyllaDB will talk about how performance single-threaded cores with fast memory I/O are critical to workloads like cloud native databases like ScyllaDB, and ways in which the architecture and CPU design drive web scale performance.

We also know that you have questions about “easy wins” - applications that will perform better, for less, on Ampere CPUs “out of the box”. These “Workloads that Shine” include AI Inference, Compute-bound applications like reverse proxy cache and web serving, latency-sensitive workloads like in-memory databases, and stateless web services written in portable languages like Golang and Java. But don’t take our word for it.

My colleague Naren Nayak, VP, Application Engineering at Ampere will talk with Vikrant Soman, Senior Staff Solutions Architect at Uber about their journey to adopt Arm64 as a deployment architecture, and the work they have done with Ampere along the way. Then, he will speak with Dmitry Polyakovsky, Consulting Software Engineer at Oracle Cloud and Tech Lead for OCI Cache Service on Oracle’s experience deploying and managing a high volume “Redis as a Service” offering on OCI, using Ampere-powered A1 instances under the hood.

After that, Victor Jakubiuk, Head of AI at Ampere, will focus on efficient AI inference, and why Ampere CPUs are a great fit. Victor will welcome Amar Gowda, OCI’s AI/ML Senior Product Manager, on the work that OCI has done with Ampere to enable AI inference workloads. Victor and Amar will be joined by Steve Notley, Field Engineering at Wallaroo.AI, to share how their products ease the management and operation of trained models in the cloud.

And finally, yours truly will look at changes that you will need to make in your build, test, and deploy infrastructure. We will learn from Maximilian Wittich of Stackable how Stackable added Arm64 as a supported platform for their product with the help of Ampere, and from Andreas Lykke of Uber what it looks like to add a new deployment architecture to large-scale application infrastructure.

Register for the Ampere Developer Summit!

If any of that sounds interesting to you, you’re in luck! The Ampere Developer Summit is virtual and free, and by registering now, you will receive reminders about the event and more “sneak peaks” into the content as we get closer to the event, and you jump to the head of the line for a “swag pack” after attending the event on September 26th.

If you are unable to attend, if you register for the event, we will notify you when session recordings are available.

We will be sharing deeper dives into each session as we get closer to the event - in ways that processor architecture and design matter for application performance, which applications you might want to move first, operationalizing your MLOps workoads on Ampere, and how to make the move in your CI/CD’s build/test/deploy pipelines. Stay tuned for more!

We look forward to seeing you then!

4 Likes