Here are some more benchmark results for our Sapphire NITRO + Radeon RX 580 Limited Edition card:
- DCR: 1.202 GHS;
- XMR: 600 H / s;
- LBC: 0.165 GHS;
- PASC: 0.830 GHS;
- SIB: 8.2 MHS.
In terms of optimization, you can experiment with memory lines and get better results. In addition, lowering the GPU clock from the higher value (default) will also lower the voltage and achieve lower power consumption. This is especially true for optimizations for ETH, although the ZEC will require a slightly higher voltage (and clock speed).
Some miners also report that the RX 580 LE board can work fine on one 8-pin PCI-E connector. But to see for yourself, again you need to optimize the frequency and voltage to avoid power overload.
What are the conclusions? The Sapphire NITRO + Radeon RX 580 Limited Edition, while appealing to miners, is clearly designed for gamers.
The overclocking option is more suitable for gaming. Miners may also like the large cooler and additional fans in the package.
But two power connectors and increased power consumption are not something that can please. The tuning margins of the RX 580 LE are a bit disappointing, but you can see some variation on performance / power consumption (the same results could be obtained by tuning the RX 580 or even the RX 570).
Hashrate RX 580 8gb in ETH + DCR mining
The first screenshot shows the performance of the RX580 8GB video card with a core frequency of 1150MHz with a BIOS mod optimized for dual ether mining and decree . With these lower core frequencies, the video card produces 30 Mh / s on air + 1000MH / s on decree. Used miner from Cliemore version 9.
On the second video card RX580 8Gb, ETH mining was launched at stock settings with a result of 24.5 MH / s.
From the data obtained, we can conclude that the performance of the new RX 580 graphics cards does not significantly exceed the performance of the previous generation RX 480. But since the RX 480 has long been out of stores, the RX 580 will be an excellent choice, the most optimal that AMD Radeon has to offer.
