Tools & Solutions
Overview
Understand Your Software.
Unlock Your Hardware.
Software tools for embedded analysis, code optimization, and AI deployment. Find the right starting point for your code, model, and target platform.
Go directly to:
What Do You Need to Achieve?
Choose your task to find the relevant tool.
Edge AI Compiler
Optimize AI Models for Embedded Deployment
Turn ONNX and PyTorch models into C code for embedded processors. Apply target-aware optimizations and integrate the generated implementation into your existing embedded software workflow.
Performance Estimator
Estimate Execution Times Early in Development
Estimate execution times from C code or generated assembly without running the application on the target. Compare implementation alternatives and monitor timing budgets as the software evolves.
Dependency Analyzer
Understand Dependencies and Event Chains
Analyze data dependencies, signal relationships, and event chains in embedded software. Check specified relationships and generate documentation to support verification, impact analysis, and reviews.
Parallel Studio
Parallelize Code for Multicore and DSPs
Interactively parallelize handwritten or generated C code for multicore processors and DSPs. Support validation of the transformed code with the Qualification Kit.
Code Vectorizer
Generate SIMD Code for Vector Hardware
Automatically vectorize C code into SIMD implementations for vector-capable processors. Use target-specific instructions to process data in parallel on the selected hardware.
C++ to C Compiler
Translate C++ into Portable C Code
Convert supported C++ code into portable C for integration into C-based environments. Use the resulting code in existing embedded toolchains and software analysis workflows.
■ Code Generation & Conversion
■ Analysis
■ Optimization
emmtrix Studio
Explore updates to our analysis and optimization tools in the emmtrix Studio release notes.
From Your Input to the Result You Need
Edge AI Compiler
Input: ONNX / PyTorch models
Processing: Model-to-C generation, automatic code transformations, and target-specific vectorization
Output: Target-optimized C code
Performance Estimator
Input: C code / generated assembly
Processing: Static execution-time estimation without running the application on target hardware
Output: Timing estimates and reports
Dependency Analyzer
Input: C code; AUTOSAR information where applicable
Processing: Analysis of data dependencies, signal flows, and event chains
Output: Dependency graphs and reports
Parallel Studio
Input: Sequential C code
Processing: Interactive parallelization and mapping to processor cores
Output: Parallelized C code for multicore processors and DSPs
Code Vectorizer
Input: Sequential C code
Processing: Code transformations and target-specific SIMD vectorization
Output: Vectorized C code
C++ to C Compiler
Input: Supported C++ code
Processing: Translation into C for C-based toolchains and analysis workflows
Output: C11 code
Your Input
Choose Your Tool
Your Result
ONNX / PyTorch models
Deploy AI
Edge AI Compiler
Model-to-C generation + optimization
Target-optimized C
C / assembly
Analyze
Performance Estimator
Timing estimates and reports
C / AUTOSAR information where applicable
Analyze
Dependency Analyzer
Dependency graphs and reports
Supported C++ code
Convert
C++ to C Compiler
C11 code
Sequential C code
Optimize
Parallel Studio: multicore
Code Vectorizer: SIMD
Parallelized or vectorized C
Simplified overview. Supported inputs and available features vary by tool and target platform.
How do Parallel Studio, Code Vectorizer, and Edge AI Compiler differ?
Parallel Studio distributes work across processor cores. Code Vectorizer uses SIMD instructions to process multiple data elements in parallel. Edge AI Compiler starts with an AI model and automates a target-oriented optimization flow that includes Code Vectorizer technology.
Start with C Code from Your AI Model
emx-onnx-cgen
Convert ONNX models into portable C for embedded integration, analysis, and further optimization.
emx-pytorch-cgen
Generate analyzable C from PyTorch workloads for C-based development workflows.
Integration Into Your Development Workflow
Model-Based Development
Assess execution times in generated C code to guide model-based implementation decisions.
AUTOSAR Analysis
Trace dependencies between signals, runnables, and components to understand change impact.
CI/CD Integration
Track timing and dependency changes with repeatable analysis in automated workflows.
What People Are Saying
We have been working with emmtrix for a couple of years now and we find their technology of great interest. Their expertise in the field of source-to-source compiler technology and their tool suite emmtrix Parallel Studio help us in developing and improving our high-performance hardware solutions.
ePS shows where and how performance can be increased. As a ‘side-effect’ of using ePS the developer quickly learns to design his application to be suitable for multicore HW.
Some Supported Platforms
R-Car
DSP
STM32
iMX.8
The platforms shown above are just a selection – we support many more for embedded systems.
Reach out to us to discuss how we can support your specific platform needs.
Which Tool Fits Your Project?
Tell us about your task, input code or model, and target platform. We can discuss a suitable tool and evaluation scope.
