CGMiner 4.11.1: Features, Setup, and Optimization Tips for Crypto Mining
Version: 4.11.1
License: free
OS: Windows
Sooner or later, every cryptocurrency miner faces a crucial decision - whether to build their mining farm using AMD or Nvidia graphics cards. Let's explore some guidance on selecting GPUs and understand how different manufacturers' characteristics affect mining performance.
ALGORITHMIC PERFORMANCE
The primary factor in choosing between manufacturers should be the cryptocurrency's mining algorithm, as different cards perform differently with various coins.
For instance, AMD cards typically excel with DaggerHashimoto and CryptoNight algorithms, while Nvidia cards show better results with Equihash and NeoScrypt. However, this isn't a universal rule. The most reliable way to determine performance is through practical testing or using mining profitability calculators.
To illustrate the differences, here's a comparative analysis of hash rates between NVIDIA GTX 1060 and AMD R9 380, two cards with similar stated performance levels:
| algorithm | NVIDIA GTX 1060 | AMD R9 380 |
| NeoScrypt | 0.7 | 0.27 |
| Lyra2REv2 | 21.9 | 6.2 |
| DaggerHashimoto | 20.13 | 18 |
| Decred | 1.82 | 0.66 |
| CryptoNight | 0.42 | 0.48 |
| Lbry | 0.19 | 0.04 |
| Equihash | 281.69 | 168.45 |
| Pascal | 0.66 | 055 |
ENERGY EFFICIENCY
Power consumption is another crucial factor to consider alongside hash rates. Comparing our example cards, the GTX 1060 uses around 120W at idle, while the R9 380 consumes about 190W. This demonstrates Nvidia's advantage in both hash rate and energy efficiency.
Historically, older AMD cards were known for higher power consumption compared to Nvidia, though recent AMD generations have achieved parity or even superiority in some cases.
MEMORY SPECIFICATIONS
While memory capacity isn't critical for mining most cryptocurrencies, it's essential for certain algorithms like DaggerHashimoto. Our example GTX 1060 features 3GB memory, while the R9 380 has 2GB, giving Nvidia another advantage.
CONCLUSION
While this analysis provides general insights, neither manufacturer offers a perfect solution. Nvidia cards offer versatility and efficiency but at a higher price point. Some AMD models, like the Radeon Vega 64, deliver competitive performance across various algorithms while typically being more budget-friendly.
A balanced approach might be running separate farms with a 50/50 split between manufacturers. This provides flexibility for mining different algorithms. However, mixing card types within a single farm isn't recommended as it may cause compatibility issues and performance problems.
