Since the data is ranked by value in descending order, the lower the rank number is, the higher the daily profit is and longer it takes to earn back the money it costs to buy that gpu. As the visualization shows, the higher a gpu’s price is, the higher profit that gpu can earn each day and longer it takes to earn the gpu’s price back. Also the clustering result aligns well with the gpus’ gaming performance and gpu types. Gpus in the middle group are all high-end gaming gpus that are faster than RTX 3080 with a 3dmark timespy score of at least 19000 as website 3dmark.com shows. Gpus in the left group are basically mid-lower range gpus with the highest one as RTX 3080 and a maximum timespy score of 18000. The right group are either professional mining gpus or data center gpus. This is an additional finding showing that gpus’ mining characteristics align well with their gaming performance.
We can also see that, as the gpus go more expensive, the size of area it shows in the spider chart will get bigger. We can also say that the more expensive a gpu is, it will have higher mining performance.