New AWS SimSpace Weaver–Run Large-Scale Spatial Simulations in the Cloud

Today, we’re announcing AWS SimSpace Weaver, a new compute service to run real-time spatial simulations in the cloud and at scale. With SimSpace Weaver, simulation developers are no longer limited by the compute and memory of their hardware.

Organizations run simulations on situations that are rare, dangerous, or very expensive to test in the real world. For example, city managers can’t wait for a natural disaster to hit a city to test the response systems. Event planners don’t want to wait until a large sporting event to start to understand the impact the games will have on traffic. Scenarios like these need to be simulated in a safe environment in which planners can test different situations and tune each system.

Until today, spatial simulations were generally confined to being run on a single piece of hardware. If developers wanted to simulate a bigger and more complex world with lots of independent and dynamic entities, they needed to provision a bigger computer. Simulation developers were forced to make trade-offs between scale and fidelity, in other words, deciding how big the world is and how many independent entities there are.

The world we live in is complex, and the scenarios that developers want to simulate are very complex as well—for example, how traffic will be affected by a large concert or sporting event. Simulating these events requires modeling hundreds of thousands of independent dynamic entities to represent the people and vehicles. Each entity has its own set of behaviors that need to be modeled as it moves throughout the world and interacts with other entities. Simulating this at a real-world scale requires CPU and memory beyond what you can have in one instance.

With SimSpace Weaver, you can run simulations at scale across multiple Amazon Elastic Compute Cloud (Amazon EC2) instances. It supports simulating upwards of a million independent and dynamic entities.

When to Use SimSpace Weaver
Use SimSpace Weaver when you need to increase the scale or complexity of your simulations. SimSpace Weaver is great at simulating crowds. This is very useful, for example, when you’re planning large events or planning to build infrastructure like a new stadium. It is also ideal for simulating smart cities, complete with vehicles, inhabitants, and other objects.

AWS SimSpace Weaver lets you connect external clients to your simulations so that you can interact and view the simulations with multiple users in real time.

How SimSpace Weaver Works
When using SimSpace Weaver, you can parallelize your spatial simulations workloads across multiple instances. Scale your simulations across up to 10 EC2 instances by specifying the compute capacity needed for the simulation and how it should be split into partitions. SimSpace Weaver handles the provisioning of the EC2 instances, launches the simulation applications, and cleans the environment after the simulation ends.

In the following image, you can see a representation of how a spatial area, in this case, a city, is spatially partitioned across different instances. Each row represents an instance. The example simulation in this image contains 10 instances, and each instance handles 16 partitions.

Map courtesy of Amazon Location Service

When working with multiple partitions, you don’t need to worry about the complexities of transferring entities between partitions. The SimSpace Weaver data replication system handles the networking and memory management for doing the transferring, regardless of whether the partitions are in the same EC2 instance or in a different one.

Another important feature that SimSpace Weaver provides is the scheduler. The SimSpace Weaver scheduler keeps all the distributed partitions synchronized at a set simulation tick rate (10, 15, or, 30 Hz), so the simulation behaves as if it was run on one machine.

SimSpace Weaver provides the infrastructure to weave together a simulation across multiple instances, but it is not a simulator. Build your simulations by integrating the AWS SimSpace Weaver C++ SDK with your code. Integrating with the SDK allows your applications to interface with the SimSpace Weaver software running in your instances. This allows SimSpace Weaver to track the global state of all your simulated entities and facilitates the transfer of entities between simulation applications. Developers building with Unreal Engine 5 or Unity can take advantage of the SimSpace Weaver out-of-the-box plugins to jump-start their projects.

Getting Started
You can get started with SimSpace Weaver from the AWS Management Console or the AWS Command-Line Interface (AWS CLI).

From the console, use our one-click sample to quickly launch your first simulation. This is a simple example of a simulation divided into four different partitions. This simulation involves spherical entities that move freely throughout the world, avoiding each other and static objects.

The wizard guides you through the main steps for running a demo simulation:

1. Download the client demo application. This is a prebuilt application that you use later to view the simulation running in the cloud. You can only run this demo application using a computer with Windows operating system.
2. Start the simulation infrastructure in the cloud. SimSpace Weaver takes care of deploying all the infrastructure you need in order to run this simulation.
3. View the simulation using the demo application you downloaded in the first step. The following image shows the result of running this simulation. Each color represents a different partition.

Available Now
Developers using SimSpace Weaver pay for the number of instances they use for the length of their simulation, with no up-front costs or licenses.

SimSpace Weaver is available in the US East (Ohio), US East (Northern Virginia), US West (Oregon), Asia-Pacific (Singapore), Asia-Pacific (Sydney), Europe (Ireland), Europe (Frankfurt), and Europe (Stockholm) AWS Regions.

You can get started with SimSpace Weaver today from the console and the AWS CLI. Learn more about SimSpace Weaver on the service page.

— Marcia