ARM vs RISC-V for IoT
A 25-page feasibility report on whether IoT deployments should be built on ARM or RISC-V, backed by real benchmarking on QEMU and Spike rather than vendor claims, plus a total-cost-of-ownership analysis framed around deployment in Lebanon specifically. It lands on a blended recommendation, not a clean winner.
The numbers behind the recommendation
Real figures pulled from the report’s TCO model and the QEMU/Spike benchmark runs
…
At a glance
- Format
- ~25-page feasibility report
- Tooling
- QEMU (ARM), Spike (RISC-V)
- Team
- 4 authors
- Course
- ENGL 206 Technical English
Benchmarks over brochures
Methodology
Rather than comparing ARM and RISC-V on paper specs, the report runs both architectures under emulation, ARM workloads on QEMU, RISC-V workloads on Spike, and measures what actually comes out. That's paired with a total cost of ownership analysis covering licensing, toolchain maturity, and hardware availability specifically in the Lebanese market, where import costs and supplier access change the calculus that a generic global comparison would miss.
The conclusion is a blended recommendation rather than a single winner, because the two architectures actually trade off differently depending on the deployment scenario, and the report says so instead of forcing a cleaner-sounding answer than the data supports.
Companion to the CPU build
Where this connects
This report and the Dual-Issue RISC-V CPU project sit on the same architecture from two different angles, one extends a RISC-V core at the RTL level, the other evaluates RISC-V as a deployment choice against its main competitor. Read together they cover both how the ISA works and why you'd pick it.
Inside the repo
Structure
- □report.docxfull 25-page feasibility report
- ▸visuals/TCO, execution-time and market-forecast charts
- □visuals.pymatplotlib, generates all 6 report figures
- □Hadi_Visual1_TCB.pngmy chart, total cost breakdown
- □README.mdteam notes and file guide
Skills, in context
Where each one actually showed up