lolMiner-144 0.6 alpha release, Equihash 144.5 for AMD and Nvidia, Linux and Windows

[mlgswag160]

Latest version of lolMiner-144 0.6 alpha release, Equihash 144.5 for AMD and Nvidia, Linux and Windows

Hi people,
This is an announcement for a public preview of lolMiner-144 0.6 alpha release, Equihash 144.5 for AMD and Nvidia, Linux and Windows.

So, what is this miner for: at the moment it runs Equihash 144.5 (BTG, BTCZ, LitecoinZ, Safecoin, Snowgem) OpenCL, therefore for Nvidia and AMD.

Release:

Windows
Linux

Compatibility:

Nvidia:

For Nvidia, the miner works on any Maxwell GPU and a newer generation with at least 3G RAM and CUDA 8 or newer. However, there is a limitation: due to some hardcoding that does not work in Windows with 3 / 4G Nvidia cards, and in Linux it is not a problem when the cards are in computation mode. I will be working on fixing a Windows problem soon.

AMD:

For AMD, we need any 4G / 8G GCN2 and a new GPU, so Hawaii (R9 290/390), Tonga (R9 285/380), Fiji (R9 Fury series), Baffin (RX 460/560), Ellesmere (470, 480 , 570 and 580) and the Vega series graphics processor. Most of the cores (exceptions below) are built from 17.40 blockchain driver.

For Vega: the Windows kernel for Vega is built with a very new version of AMD Drive (Adrenalin orders 18.5.1). It is unknown whether the old drivers are supported. For Linux, the AMD Vega kernel is available only for AMD ROCM-CL drivers! I have to admit that we have not tested the Vega GPU yet, so crossing fingers will work. I will also try to make it work as soon as possible.


Miner speed:


Expected speeds

Equihash 96/5:

Nvidia:

GTX 1080: 24,000 sol / s

Equihash 144/5:

AMD:

RX 580: 27-29 sol / s
RX 560: 14 sol / s

Nvidia:

GTR 2080: 75 sols / s GTX 1080 ti: 85-90 sols / s GTX 1080: 53-56 sol / s
GTX 1070: 40-45 sol / s
GTX 1060: 26-30 sol / s

more values ​​added if available

Equihash 192/7:

AMD:

RX 580: 17 sol / s

Nvidia:
GTX 1080: 29-30 sol / s

Changes in the new version:

0.6 alpha 2 (November 2, 2018)

Equihash 192/7 and 96/5 added (96/5 for Nvidia only, AMD will be added next week)
Faster production by 192/7 compared with a 0.5 stable release
Faster production at 96/5 for Nvidia
192/7 memory consumption reduced to 2860 MB / GPU
Updated stratum interface: more fault tolerant and time and time statistics
Improved API module: GPU name, PCIE topology, new statistics from the stratum module

- New development code for 192/7 (Zero, GenX) -
This is based on the best code 144/5. Compared with 0.5 code Nvidia 25% faster, AMD - about 15%.
Also, the code should now work again on 3G cards (~ 2860 MB required)

- New mining code for MNX on Nvidia -
AMD code will follow next week - I had problems with a means of complete stability.
Nvidia is 20% faster than 0.5 code. This will most likely not be the final code for a full 0.6,

it exceeds ewbf 0.5, but not 0.6, so they have enough motivation to further improve xD

- Massive rework in the stratum interface -
I fixed a lot of mistakes there. The most important are the correct routine processing miner, when the first attempt to connect
does not work and corrects "the received new task, but the error during the extraction stops".
Also as a new feature, lolMiner now detects which GPU received received balls, and it measures the delay
pool response.

- Improved API / Statistics module -
All statistics are now collected in one position in the miner
The API now provides information on pool latency, image and issue promotions for each GPU, GPUs,
the PCIE address of the graphics processor (to connect the miner with external tools) and so on.

[monika 576]

very interesting topic