Quote from: mi6ee on Today at 03:23:35 AM

SmartMinerPRO 1.3.0

Add Miners & Update

GMiner 2.96
TRM 0.10.0
T-Rex 0.26.1
xmrig-6.7.0
cpuminer-opt-3.19.7

Updates SMP:

Check generated DAG, a warning at very high GPU overclocking for Ethash, Etcash, KAWPOW and ProgPoW algorithms, helps overclock GPU without errors
checking Shares on CPU, a warning when the GPU is overclocked too high for the Ethash, Etcash, KAWPOW and ProgPoW algorithms, helps overclock the GPU without errors
DAG caching, if the GPU has enough memory, DAG files are not recalculated when switching to another algorithm when mining Ethash + Zilliqa or Nicehash, which has a positive impact on user profitability
auto-selection of optimal cores for each device on Ethash, Etcash, KAWPOW
ability to manually select the core on each device for Ethash, Etcash
GPU temperature control and shutdown in case of overheating
watchdog - a process-watchdog that monitors the state of key systems of the miner, and will restart the miner if it crashes or hangs
Mechanism to recover lost connections to the pool
Failover support - the Miner uses failover pools until the connection to the main pool is restored
support for secure connections
SOCKS5 proxy support
informative and readable display of tabular statistics on the console
detailed information per device (temperature, power consumption, load on the cooler, memory frequency, CPU frequency, power efficiency)
parallel output of the information to the console and to a file on the disk
built-in statistics server - remote monitoring of the miner in a web browser
memory settings for Nvidia GPUs with GDDR5X and GDDR5 memory
Overclocking core clock, memory clock, core voltage, memory voltage, fan speed, power limit for Windows
Safe DAG generation for Nvidia GPUs
auto fan speed control to achieve target temperatures

Download Link: https://github.com/MiningSoftware/SmartMinerPRO/releases/

If I run both cards together it will still say that all the cards are being selected, but there is no apparent impact on the 3090Ti, it will run at its normal 135-137 MH/s as a miner reported speed, but if I leave the 3090Ti idle, it will corrupt (maybe due to the P-state not switching to P0, or some fan monitoring that is performed by HiveOS when a card is engaged in mining vs. idle?) and the computer will reboot; I will have to try and confirm the actual poolside speed on the card to see if it is reflecting the 135+ from the miner.

As for the speed: the miner is reporting about 85.6 average, but the 'realtime' fluctuates a lot, as compared to the latest t-rex, which gives 83 MH/s from the miner and is very stable with that value after about the first 5 minutes, I think the pill is cycling too often (maybe there are conditions which make it cycle more often or less?) and when it does a cycle, the hashrate drops and reduces the overall reported rate for the next 60-90 seconds, which will reduce the average to 82 or 83 and possibly the overall poolside hashrate (although 2MH/s isn't really significant on a shares found metric as it's less than .25 SD, which means luck has a far greater impact on overall average share rate than 2 MH/s).

My next test will be to run the 3090Ti in another instance outside of Hive, or with another miner (since HiveOS won't allow the same miner to be running twice within its management software) and see if the Pill hurts the 3090Ti in that case, and also to get single card speed and variability metrics for the 3070Ti LHR card.

xolling

Posts

Merits