If you start a big thread about this, then you should also be the first to share results instead of expecting everyone else sharing results.
Maybe start with why you like that 5700XT so much; e.g. whatās the wattage/hashrate ratio? What did you compare it with? ā¦
For example, I get these WHrās (lower=better):
GTX 1070: ~4
GTX 1080ti: ~3.6
RX580 Samsung: ~2.3
Vega 56 Hynix: ~2.4
Vega 56 Samsung: ~2.4
Vega 56 Gigabyte: ~3.3
Vega 64: ~2.6
AMDās power consumption numbers are completely false. Donāt be fooled.
You need to buy a meter (for example, a āKill-a-Wattā) and measure the actual consumption at the wall outlet in watts.
Typically, my RX5700 and RX5700 XTs pull about 170-200 watts each mining at 57Mh/s when measured at the wall with a 1100 watt 80-plus Gold rated PSU.
HiveOS is simply reporting the driverās power estimation for the GPU chip only, not including the entire card (VRAM, voltage regulators, etc). Plus, there are losses at the motherboard, power supply and cooling you need to account for when running a mining rig.
Donāt get me wrongā¦ youāre probably very profitableā¦ RX 5700s are amazing cards that have the BEST hashrate-to-dollar values. (Hashrate-to-watts is another stupid metricā¦ Watts are cheap for most peopleā¦ the cards are NOT cheap. If a card takes 8-12 months to pay itself off, it might as well be a boat anchor)
Hypothetically, if the driver says 74 watts, but the actual power consumption of the card is the same (or only a few watts different) as when the software said 94 watts, is it really 20 watts more efficient??
No.
You can find people on other forums who found that AMDās numbers didnāt actually show the same change in power consumption. Iām fact, in many cases, the actual power consumption didnāt change at all!
Iām a proud owner of an AMD GPU household, with only one Nvidia card in service, but their drivers are really bad when it comes to their GPU power estimates.
Thatās why this is an AMD thread to ensure comparability, even if the AMD GPUS are supposedly all wrong, it doesnāt make any difference in efficiency. If all are wrong, you can still compare them.
There is an NVIDIA thread and an AMD thread and I donāt want to compare the efficiency between Nvidie and AMD GPUS or drivers here.
This thread is just for MHs / watt comparisons that are displayed in Hiveos. Everyone can decide for themselves whether that makes sense.
But this way we get at least a few numbers that we can compare and estimate with each other.
We donāt have to make rocket science out of it here.
Im also thinking about this, do you believe the RX 6000 Version can beat the Boss 5700 XT in future? Tomorrow there will be released a new Card from AMD. Very excited about that.