Skip to main content
SEGGER emRun / emRun++ / emFloat
SEGGER SEGGER

emRun / emRun++ / emFloat

Embedded Software

High-performance C/C++ runtime and floating-point libraries for Arm and RISC-V, delivering up to 4x smaller footprint and 4x faster execution than newlib. Available in India from GSAS.

Components

emRun (C), emRun++ (C++), emFloat (FP)

Architectures

Arm Cortex-M/A, RISC-V (RV32/RV64)

ROM Savings

Up to 2x smaller than newlib-nano

Speed

Up to 4x faster integer math vs newlib

IEEE 754

Fully compliant (emFloat)

License

Source code, royalty-free

Authorized SEGGER partner
Local FAE & application engineering
Hands-on training & integration
Manufacturer warranty

About emRun / emRun++ / emFloat

emRun / emRun++ / emFloat

SEGGER emRun is a complete C runtime library specifically designed and optimised for GCC/LLVM-based toolchains and embedded systems, delivering dramatically smaller code and faster execution than the standard newlib/newlib-nano libraries that ship with most Arm and RISC-V toolchains. emRun++ extends coverage to C++ runtime needs, while emFloat provides an IEEE 754 compliant floating-point library that achieves FPU-like performance in pure software, critical for microcontrollers without a hardware floating-point unit. Together, these three libraries form a drop-in replacement for the default runtime that can cut ROM usage in half and multiply execution speed without any source code changes.

Performance Benchmarks

Independent benchmarks on RISC-V 32-bit targets show the scale of improvement over GCC/newlib:

BenchmarkemRunnewlibImprovement
Floating-point arithmetic1,097 cycles1,907 cycles1.7x faster
Integer arithmetic251 cycles1,042 cycles4.2x faster
printf ROM (simple test)4,392 bytes9,552 bytes (nano)2.2x smaller
printf RAM (simple test)184 bytes416 bytes (nano)2.3x smaller

emRun achieves these gains through hand-optimised assembly routines for Arm (Cortex-M and 32-bit Cortex-A) and RISC-V (RV32M, RV64), customisable formatted I/O that excludes unused conversion specifiers, full localization and UTF-8 support linked only when used, and zero heap requirements. A generic ANSI C implementation covers additional architectures.

Adoption and Quality

emRun and emFloat have been adopted by leading silicon vendors including SiFive, Microchip, and Nuclei, and are integrated into SEGGER Embedded Studio and Rowley CrossWorks. The libraries are verified with over 450 test suites comprising millions of test cases, including corner-case validation for every function. Source code delivery and royalty-free licensing, with no viral open-source clauses or attribution requirements, make emRun suitable for commercial products in any market segment. Silicon vendor buyout licences are also available for companies that want to bundle the runtime with their own toolchain distribution.

Interested in emRun / emRun++ / emFloat?

Get pricing, an evaluation unit, or a technical consultation from our application engineers.

Request a Quote Demo