Network Simulation

Simulation provides valuable insight into how your real world networks behave and can predict the performance of networks or components not yet deployed or available for testing. This understanding can be used to debug intricate network challenges or optimize overall performance. Theoretical technology enhancements, proposed transports and advanced Al workloads can all be simulated and studied at data center scale.

Why Scala Computing?

Gain Insights

Simulate real-world network behavior and predict your network and workload performance.

Debug complex network challenges and enhance overall performance.

Study theoretical enhancements, proposed transports, and advanced AI workloads, at data center scale.

Avoid Lab Issues

Large scale physical labs are prohibitively expensive, power hungry, and require IT to setup.

Shared resources often means long queues and limited capacity.

Unlock the Full Potential

Simulation Set Up

The user interface is designed for the setup of network simulations, including laying out or modifying the networking equipment topologies, workloads, traffic and stack parameters. Engineers can easily modify configurations and rerun to test what-if scenarios. Automated configuration validation guides the user to a successful simulation every time.

Performance

Our unique solution is optimized for massive parallelization, enabling simulations of large Al clusters with LLM workloads.

Deploy to Cloud Resources

Our automated deployment infrastructure enables easy elastic access to vast multi-instance cloud resources that provide the compute power your topology requires, on demand with no queues.

Results

The Ul provides a birds eye view of network performance metrics, with the ability to drill down into the data analysis to understand any single component endpoint. Our data analytics engine can process terabytes of results in real-time.

Compare

Side-by-side corparison of simulation results, with automated quantification, makes it easy to understand sweep test sensitivities.

AI/ML Workloads

Scala computing is focused on driving innovation around Al/ML workloads and next generation hardware. We provide Chakra analysis, visualization and trace synthesis for capturing realistic traffic, including Al traffic.

Collectives

Scala provides a library of traffic patterns that you can configure. Develop and experiment with your own collective operations to create custom traffic, replicate vendor traffic libraries or conduct Al research.

Consortium

The consortium ecosystem makes it possible for industry giants to work together by sharing models and workloads through the Scala Compute Platform without compromising their intellectual property.

Scala Compute Platform

This is all built on the Scala Computing Platform so it is easy to manage users, usage, models and results.

Get Started Quickly

Includes cutting-edge features like RoCE, Falcon and UEC transports for pre-production testing.

Ready-to-use generic options for immediate simulation.

Develop outside the box

Test experimental features, like congestion control algorithms and protocol enhancements (e.g., Selective ACK to RDMA).

Advanced environment for building custom ns-3 models and charts.

Scala Computing is comitted to being on the leading edge of networking technology.

Features such as RoCE, Falcon and UEC transports are included so your engineers can experiment with these technologies and understand their performance before they even become available for production testing.

With simulation it is possible to evaluate experimental features including congestion control algorithms and protocol enhancements, such as adding Selective ACK to RDMA.

Scala provides generic options so you can simulate on your first day. Our development environments are available to empower users who want to build their own advanced custom ns-3 models or charts.

Explore Our Comprehensive Development Environments

The Scala Development Environment can be used to develop ns-3 simulation models. These ns-3 models can be created to simulate capabilities of existing networking hardware, or to experiment with future enhancements, such as transport protocols and Al collectives. Once created, these models can be used exclusively for your own simulations or securely listed and shared with partners via the consortium ecosystem.
Scala Computing
VSCode
Use VSCode for editing/debugging models
Secure
Your Source code is always secure in the development environment
Unlimited
Environment includes a single core instance for unlimited simulation testing
Topology
Test your model in a topology along with provided generic models from Scala Computing
The Scala Analytics Environment for ns-3 allows users to go beyond the included set of charts and tables to build highly scalable and customizable analytics solutions. Custom charts and visualizations can be created for any data in the result set. The Analytics Environment can be used to create the charting to display novel results from custom models created in the Development Environment.
Scala Computing
Python
Python based data science toolkit with Jupyter Notebook IDE
Charting
Integrated and flexible graphics and visualization tools
Tightly Integrated
Tightly integrated with active ns-3 simulations to provide insight in real time
HPC
Efficient paralleled data processing of large data sets
Collaborative
Environment allows for multiple users to build insight reports together
Examples
Training examples are provided to reduce the learning curve for new users
The Scala Wireshark Environment is a powerful network protocol analyzer that allows you to capture and interactively browse the ns3-simulation traffic. It is typically used for advanced analysis of network traffic. Wireshark is used by industry professionals for network troubleshooting, protocol analysis, software and communications protocol development and training. It is fully integrated with the rest of the Scala Computing platform so that users can easily view their simulation packet capture data (PCAP) for running simulations.
Scala Computing
Integrated desktop
Decide which models produce PCAP results for analysis and evaluate those PCAP files in real time during the simulation
Full Capabilities
Wireshark allows for deep inspection of multiple protocols and includes features such as advanced packet statistics, highlighting potential issues and graphical representations of network traffic patterns for intuitive analysis
Collaborative
Platform supports a virtually unlimited number of users with securely managed access to shared simulations and results