Hi there,
this is the announcement thread for lolMiner 0.5, a multi algorithm Equihash miner with focus on AMD GPUs (OpenCL based).
Note it will run on most Nvidia GPUs, too but the speed and stability may vary. Hope to improve this in the future.

Currently with this miner you can mine 4 different algorithms:

Algorithm	Req GPU Mem	Coins that use it
Equihash96/5	500MB	Minex Coin (MNX)
Equihash144/5	3G	E.g. ANON, BTG, BTCZ,XSG, ZEL and many more!
Equihash192/7	4G	Zero (ZER), Genesis (GENX) & Zero Classic
Equihash210/9	1G	Aion Mainnet

Download:

From Dropbox:
[Lin 64] Click me (0.5 Linux)
[Win 64] Click me (0.5 Windows)
(More Mirrors to come)

Usage:

The first way to use the miner is from command line interface. The syntax is


Here the parameter is a short name for the coin you want to mine. For example inserting AION will configure the miner to mine AION, MNX for MinexCoin and so on.
For more shortcuts and more options please read the lolMiner_manual.pdf shipped with the miner.
lolMiner 0.5 supports fail-over pools, configuration files (easy sharing between Rigs and easy switching between coins), automatic worker names, a simple statistics reading API, logging to text file and many more features.


Estimated Performance:

Most of this measurements were reported by community members in this thread. Results may vary a bit depending on clocks, driver version and operation systems:

GPU	Equi 96/5	Equi 210/9	Equi 144/5	Equi 192/7
Vega64	~24.000 sol/s	~155.0 sol/s	~37 sol/s	~22.5 sol/s
580	~13.800 sol/s	~107.5 sol/s	~25 sol/s	~15.5 sol/s
560	~6.900 sol/s	~54 sol/s	~12.5 sol/s	~7.5 sol/s

(More data to come when available)

Known issues and troubleshooting:

Problem: The miner crashed at startup when creating memory buffers or miner ends up with 0 sol/s from start.
Solution: The AMD driver requires on start as many MB of system memory as it will need on all started GPUs together. Increase the fixed page file size (Windows) or swap partition size (Linux) to give the rig more virtual memory.

Problem: The miner gets unstable or slow when using more then 4-5 GPUs
Solution: Start two instances and use the --devices parameter (command line) or "DEVICES" (config file) to point one instance to the first n GPUs and the second to the next. For most systems that will work better.