Newswise — Upgrading to a new smartphone is exciting—until you realize your old apps no longer work. Many people have faced this frustration, but imagine if businesses had the same problem with the technology running their smart home devices, medical equipment, or factory automation systems. When companies switch to newer, more efficient processors, they often find that their existing software is incompatible, forcing them to rewrite or modify programs—an expensive and time-consuming process.
Researchers from the National University of Defense Technology (China) have developed RVAM16, a new low-cost processor that acts like a built-in translator, allowing older software to run seamlessly on next-generation hardware. By integrating a hardware-based translation system, RVAM16 enables smooth execution of programs originally designed for ARM Thumb chips on RISC-V processors, ensuring high performance without the usual compatibility headaches or extra costs.
Why Software Compatibility is a Challenge in Embedded Systems
Many of the electronic devices we rely on—from smart home assistants to car infotainment systems—run on embedded systems, which are specialized computers designed for specific tasks. These devices operate using Instruction Set Architectures (ISAs), which define how a processor understands and executes software instructions.
For years, many embedded systems have relied on ARM Thumb, a compact version of the widely used ARM processor design, known for its power efficiency and small size. However, more companies are now shifting toward RISC-V, an emerging open-source processor architecture that is cost-effective, flexible, and highly customizable. The challenge? Software designed for ARM-based chips won’t naturally work on RISC-V processors without modification or performance loss.
This is where RVAM16 comes in. Instead of requiring companies to rewrite their software from scratch, it automatically translates ARM Thumb instructions into RISC-V commands on the fly—directly in hardware—ensuring a smooth transition without expensive redevelopment.
Faster, More Efficient Processing Without the Slowdown
Most existing solutions rely on software-based translation, which can significantly slow down performance—similar to running an old computer game on an emulator. RVAM16 avoids this pitfall by performing the translation inside the processor itself, meaning it can run legacy software at up to 65% of the original ARM speed, with performance gains of up to 5.86× compared to traditional software translation methods.
Additionally, RVAM16 is designed for power efficiency, requiring minimal extra hardware. This makes it an ideal solution for devices where battery life and low power consumption are critical—such as smartwatches, medical sensors, and industrial automation systems.
A Practical, Cost-Effective Path Forward
By removing the compatibility barriers between different processor architectures, RVAM16 offers a future-proof solution for embedded system developers and policymakers looking for cost-effective, open-source alternatives to proprietary systems.
“Integrating multiple ISAs through hardware-based translation allows embedded systems to evolve without the high cost of rewriting software,” said Prof. Yuanhu Cheng, the corresponding author of the research.
With RVAM16, upgrading hardware no longer means leaving behind years of valuable software investments. Just like a universal charger works across different phone brands, RVAM16 ensures that software can run smoothly across different processors, keeping costs low and performance high. The complete study is accessible via DOI: 10.1007/s11704-023-3239-x.