API Comparison

Table of Contents

• • 🔹 Overview
  • 🔹 Performance
    • RapidCode (Direct)
    • Real-Time Tasks
    • RapidCodeRemote
    • Synchronization
    • Determinism
  • 🔹 Ease of Use
    • Development Complexity

RMP API performance, ease of use, & requirement comparison.

🔹 Overview

API Name	Language	Supported Platform	Real-Time Capable
RapidCode	C++	Windows	❌
RapidCode	C++	Linux	✅
RapidCodeRT	C++	INtime RTOS	✅
RapidCode.NET	C#, VB.NET	Windows, Linux	❌
RapidCodePython	Python	Windows, Linux	❌
Real-Time Tasks	C++ (task functions)	Linux (PREEMPT_RT), INtime	✅
RapidCodeRemote	gRPC (C#, Python, C++, Go, etc.)	Client: Any	❌

🔹 Performance

C++ generally delivers the best raw speed, making it ideal for performance-critical motion control. RapidCode.NET (C# or VB.NET) and RapidCode Python are thinner wrappers over the C++ library, offering simpler development at the cost of a small performance penalty.

RapidCode (Direct)

Using RapidCode directly on the PC running RMP provides the best performance, since it operates directly on the RMP firmware's shared memory. Even in C#, which is wrapped over the C++ API, reading values typically takes around 300 nanoseconds, only marginally slower than C++ calls.

Real-Time Tasks

Real-Time Tasks (RTTasks) use RapidCode C++ inside conveniently managed real-time threads. The RTTaskManager handles thread creation, scheduling, and synchronization, allowing you to write deterministic logic without deep real-time programming expertise.

RapidCodeRemote

RapidCodeRemote operates over gRPC, so every call goes through rapidserver, adding serialization and network transport overhead. Depending on your application, this may or may not be significant.

If you need to operate the RMP from a separate machine, RapidCodeRemote is the right tool; RapidSetupX uses it for this reason. Otherwise, most customers use the RapidCode libraries (C++, C#, Python) directly on the RMP host for better performance.

Synchronization

On an RTOS (e.g., Linux real-time threads, INtime), synchronization mechanisms introduce overhead but enable deterministic execution. Hence, while real-time threads may run slower, they ensure consistent timing.

Determinism

Most deterministic to least:

1. Real-Time Tasks – Deterministic scheduling alongside RMP firmware
2. C++ (Linux/INtime, when using real-time threads)
3. C++ (Windows)
4. .NET (C# / VB) and Python
5. RapidCodeRemote – Network latency makes timing unpredictable

🔹 Ease of Use

Development Complexity

Easiest to most complex:

1. .NET (C# / VB.NET) – Easiest setup, extensive tooling.
2. Python – Also easy but less widely used; good for quick tests or scripts.
3. C++ (Windows) – Requires more low-level setup.
4. Real-Time Tasks – Write C++ task functions; RTTaskManager handles the real-time complexity.
5. C++ (Linux / INtime, Real-Time) – Manual real-time thread management adds complexity.
6. RapidCodeRemote – Requires understanding gRPC, protobufs, and rapidserver setup.