Open Source · MIT License

GPU-Accelerated
Embedded Infrastructure

Open source tools for embedded control systems, real-time simulation, and edge computing. From benchmarking and diagnostics to flight software frameworks, command & control, and 3D visualization — built on modern C++ and CUDA.

Explore on GitHub → View Products

The Stack

V
RELEASED
Vernier
Performance & Energy Benchmarking

Benchmarking framework extending Google Test with statistical analysis, multiple profiler backends including energy measurement, CUDA/multi-GPU support, and CI/CD regression detection. Measure performance and energy consumption of your code with statistical rigor.

C++23 Multi-platform Energy profiling
CUDA gtest CI/CD Multi-GPU
View on GitHub →
S
RELEASED
Seeker
System Diagnostics & RT Validation

System diagnostics library for real-time Linux systems. Covers CPU, memory, storage, network, GPU, and more with RT-safety annotations and CLI tools for introspection and validation. Know your system before you benchmark it.

C++23 Multi-domain RT-safe
GPU support CLI tools Diagnostics Validation
View on GitHub →
A
RELEASED
Apex CSF
Control & Simulation Framework

Real-time component framework for embedded and edge systems. One codebase targets POSIX hosts and six bare-metal MCU families. You write components — models, drivers, services — and the framework handles scheduling, lifecycle, communication, logging, and deployment with RT-safe guarantees. Built for flight software, hardware-in-the-loop test rigs, industrial controllers, and edge compute pipelines.

C++23 POSIX + bare-metal CUDA accelerated
Jetson STM32 ESP32 Monte Carlo Hot-swap
View on GitHub →
Z
RELEASED
Zenith
Web Command & Control

Real-time operations interface for Apex CSF applications. Provides a REST API for commanding and telemetry, and serves a web UI for visualization, configuration, and system management. Fully configuration-driven — no hardcoded component knowledge.

Rust backend React frontend WebSocket real-time
REST API WebSocket Multi-target Telemetry
View on GitHub →
H
IN DEVELOPMENT
Horizon
Apex CSF ↔ Unreal Engine 5 Bridge

Real-time 3D visualization of Apex CSF applications. Renders live and simulated state in Unreal Engine 5, with an optional HIL/SIL back-channel that feeds synthetic sensor data back into the simulator. Also includes a procedural geometry kit for fully code-defined, reproducible scenes. Shipping soon.

C++20 UE 5.7 plugin HIL / SIL ready
UE5 Procedural geometry Sensor simulation HIL / SIL

In Action

Apex CSF

Apex CSF Edge Compute Demo
GPU-accelerated workloads under deterministic real-time scheduling on NVIDIA Jetson AGX Thor.
Apex CSF HIL Flight Demo
Closed-loop hardware-in-the-loop flight simulation across Raspberry Pi and STM32 bare-metal.

Zenith

Zenith Dashboard
Real-time dashboard with component health, executive stats, and multi-target management.
Zenith Telemetry
Live telemetry visualization with configurable plots, channel picker, and data export.

Vernier

Vernier GPU Time Breakdown
GPU kernel vs transfer time breakdown across CUDA benchmarks.
Vernier Measurement Stability
Measurement stability analysis with CV% thresholds for CI/CD regression gating.

Seeker

Seeker CPU RT Check
CPU real-time readiness validation with pass/warn/fail checks and remediation steps.
Seeker GPU RT Check
Per-GPU RT readiness checks for persistence mode, thermal throttling, and process isolation.

Horizon

Apex CSF + Horizon + Zenith Ecosystem
Ecosystem view: one Apex CSF application runs on the host (control/simulation); Horizon (UE5 plugin) and Zenith (browser operator interface) both attach live as clients.
Procedurally generated quadcopter flying a programmed trajectory inside an asymmetric sensor-test volume. Motion streams from Apex CSF at 50 Hz; six lidar rays report distance readings back each frame.
Live-tuned airfoil in a virtual wind tunnel, shape driven by conformal mapping. Angle of attack, chord shape, and freestream velocity all stream in at 50 Hz; analytical lift coefficient recomputes every frame and streams back to the simulator.
Built With
C++23
Core language
CUDA
GPU acceleration
Rust
CLI tools & backend
CMake
Build system
Docker
Cross-compilation
GitHub Actions
CI/CD

About

Apex Edge Systems builds open source developer infrastructure for embedded control, real-time simulation, and GPU-accelerated edge computing.

The stack covers the full lifecycle: benchmark and profile your code with Vernier, validate your system configuration with Seeker, build your application on Apex CSF, command and monitor deployments through Zenith, and visualize live or simulated state in 3D with Horizon. One ecosystem, from microcontrollers to GPU-accelerated SBCs.

Everything is MIT licensed.

Platforms x86, ARM, RISC-V, bare-metal MCU
GPU CUDA / NVIDIA
Contact kgomez@apexedge.dev