Xeolot
Invented Xeolot, a crypto miner written in C++, cross-platform for Linux, Windows, and macOS, including support and binaries for Raspberry Pi and OpenWrt routers. It was 28% more efficient than traditional miners.
Most crypto miners are hungry programs that assume they are running on a beefy rig with a fast, reliable connection. Xeolot was built on the opposite bet: that a miner written carefully enough in C++ could squeeze more work out of less hardware and shakier networks. The result is a lean, cross-platform mining engine that runs everywhere from a full desktop down to a Raspberry Pi or a home router, and it does so while running about 28% more efficiently than conventional miners. That number is the whole point of the project: same silicon, more useful work, less waste.
### One codebase, everywhere
Xeolot compiles and ships for Linux, Windows, and macOS, with dedicated binaries for resource constrained hardware like Raspberry Pi boards and OpenWrt routers. Rather than treating those tiny devices as an afterthought, the whole design leans into them: a low memory footprint and efficient CPU use mean the miner behaves the same on a cluster of cheap nodes as it does on a high end machine. That makes it a natural fit for edge computing and distributed setups where every watt and every megabyte counts.
### Talking to the network
Underneath, Xeolot speaks the languages the mining world actually uses. It supports the Stratum and GBT pooled mining protocols, plus ultra low overhead solo mining for anyone pointing straight at their own node. A scalable networking scheduler lets it handle any hashrate without stalling while keeping connection overhead to a minimum, so communication stays steady even under heavy load. It also uses long polling to react the instant new work appears, and it is clever enough to borrow longpoll data from any available pool even when the primary one does not offer it.
### Never sitting idle
The efficiency gains come from a simple obsession: never let the hardware wait. Xeolot is heavily threaded, so fetching new work and submitting completed work happen in parallel without ever pausing the actual mining. It prefetches the next job before the current one finishes, caches submissions during brief network outages so nothing is lost, and can even generate valid work locally through Stratum, GBT, or ntime rollover when supported. To avoid wasted effort, it runs its own lightweight block detection database, spotting fresh blockchain blocks on its own when a pool's longpoll is slow or failing, which keeps it from grinding away on stale work that would only get rejected.
### Built to stay up
Pools go down, get slow, or start misbehaving, and Xeolot is designed to shrug all of that off. Its failover and multi-pool support automatically reroutes mining to backup pools the moment the primary one stumbles. A pool that starts rejecting every share is temporarily disabled so it cannot drag performance down, and when the main pool is merely sluggish, the miner trickles spare work to the backups so no computational cycle goes to waste.
### Full transparency and control
Xeolot keeps you fully informed. It tracks detailed per device statistics, including accepted, rejected, and hardware error rates alongside work utility metrics, and it prints a clean summary report on exit so you can see exactly how a session went. A menu based configuration interface lets you adjust most settings on the fly, with no restart and no hand editing of config files. Taken together, the threading, the smart network handling, and the caching logic make Xeolot a small but genuine demonstration of how thoughtful system design turns into real, measurable gains in the decentralized computing world.
At a glance
- Timeline
- Jan 2019 - Mar 2021
- Type
- Research & Development
- Efficiency
- 28% more efficient
- Other Contributors
- Ujwal Prakash
Technologies
- C++
- Cryptocurrency Mining Algorithms
- Cross-platform Development
- Embedded Systems (Raspberry Pi, OpenWrt)
Categories
Links
Machine-readable version of this page: https://www.sherwinvishesh.com/project/xeolot.md | site index: https://www.sherwinvishesh.com/llms.txt
Copyright 2026 Sherwin Vishesh Jathanna. Text may be quoted with attribution. The design and source code are not licensed for reuse.