AWS Lambda Updates: High Resource Functions, Container Images & Millisecond Billing

Amazon Web Services revealed significant enhancements to its Lambda serverless function service today. Initially, Lambda will now support functions utilizing up to 10 GB of memory and 6 virtual CPUs (vCPUs). This expanded capacity will empower developers creating demanding functions to secure the necessary resources.

“As of today, Lambda functions can be configured with up to 10 GB of memory, representing an increase of over 300% from prior limitations. AWS Lambda proportionally allocates CPU power and other resources based on the configured memory. Consequently, each execution environment can now access up to 6 vCPUs,” the company detailed in a blog post outlining these new features.

The term “serverless computing” does not imply the absence of servers. Rather, it signifies that developers are relieved of managing compute, storage, and memory demands, as the cloud provider – AWS in this instance – handles these aspects, allowing developers to concentrate solely on application coding instead of resource deployment.

This announcement, alongside compatibility with the AVX2 instruction set, enables developers to leverage this methodology with advanced technologies including machine learning, gaming, and high performance computing.

A key benefit of this system is the potential for cost savings, as you only pay for the resources actively consumed. Billing occurs each time the application requires resources, and no longer than that. Further enhancing this advantage, the company also stated, “Beginning today, execution duration will be rounded up to the nearest millisecond, with no minimum execution time,” as announced in a blog post regarding the updated pricing structure.

Additionally, the company announced support for container image deployment for Lambda functions. “To facilitate this, you can now package and deploy Lambda functions as container images up to 10 GB in size. This allows for streamlined building and deployment of larger workloads dependent on substantial dependencies, like machine learning or data-intensive applications,” the company explained in a blog post detailing the new functionality.

Taken together, these updates mean that Lambda functions can now handle more complex operations than before, and the revised billing model is expected to reduce overall costs during the transition to these new capabilities.