As the GPU shortage continues, what constitutes “success” is quickly reworked. Several publications have recently run stories claiming that an uptick in Ampere GPU distributions according to the Steam Hardware Survey is evidence that these cards are making their way to gamers, and that miners are not sucking up all the demand. This is actually true, since cryptocurrency does not literally use every single GPU. When considered in a historical context, however, today’s SHS does not support an optimistic narrative about Ampere accessibility.
RTX 3070 gained 0.17 percent market share from February 2021 to March 2021. It is the largest market share some GPU got last month. But according to previous Steam data, a single GPU that tops out with 0.17 percent adoption is not very good at all.
I have researched several more data points in SHS in the last two years. In November 2019, no fewer than 9 GPUs gained more than 0.17 percent market share. For example, RTX 2060 picked up 0.42 percent that month. In February 2020, before the pandemic hit, the GTX 1660 Ti and GTX 1650 received 0.34 percent and 0.51 percent, respectively, with other cards above 0.17 percent. Even in March 2020, with the pandemic in place, as short as RTX 2060 (0.51 percent), RTX 2070 (0.31 percent) and RTX 2070 Super (0.28 percent) stronger growth than what was reported for Ampere in day.
Steam has only included data on the RTX 3070 for two months, but the RTX 3080 has been included longer. The trend is not encouraging:
Data on market shares above are for the period November 2020 – March 2021. Look at what happened to RTX 3080’s adoption rate after December. We see the share of gaming markets more than double in one month. Then the growth rate falls off a cliff. It took RTX 3080 one month to grow by 2.08 times, and then another three months to grow by 1.77 times. In a normal year, this may not reflect more than seasonality, but this is not it a normal year. There are still many potential players waiting for prices to fall.
This is reflected in how the numbers for all the other cards Stop moves. Anyone with a GTX 1070 Ti, RTX 2080 Super or RTX 2080 is a potential RTX 3070 customer (the 2080 Ti is a little too high to really see the RTX 3070 as a replacement card). In November and December, the percentage of users with each of these cards bounces around. From January onwards, the percentages have been almost static. RTX 2080 Ti customers do not upgrade to RTX 3090. RTX 2080 and 2080 Super owners obviously do not upgrade to Ampere. The Pascal players that Nvidia said were explicitly targeted with this launch, remain largely devoted to their hardware. GTX 1060 has fallen 1.14 percent since November, and 1050 Ti has fallen around 0.5 percent. RTX 1070 is 0.36 percent less common now than in November 2020. GTX 1080 has fallen even less.
One reason the RTX 3070 looks good is that Steam did not allow the GPU to track until it reached 1.12 percent. If we actually had the month-on-month report, however, I bet we would see exactly the same thing as with the RTX 3080 – a first shock, followed by slow growth for such an attractively positioned high-end part. RTX 3060 Ti entered the SHS at 0.27 percent in January and has risen to just 0.38 percent three months later.
Together, RTX 3080 and RTX 3070’s market share grows at the level of how RTX 2080 and 2070 performed back in 2019. At MSRP, Ampere is all Turing was not. It offers beam tracking performance we feel more comfortable recommending and much stronger AI and game performance. It is also much cheaper (theoretically) than the RTX 2070 and RTX 2080 were at launch back in 2018. In this context, an increase of 0.17 percent in the number of RTX 3070 GPUs in the market is not a ray of hope. It is a demonstration of how bad the market continues to be.
At the time of writing, anyone who needs a new GPU should consider AMD’s R9 290 and R9 290X as the best options for an affordable card. We continue to keep an eye on this situation, and it continues to offer the best price / performance ratio beyond being lucky. RDNA2 GPUs are not considered in the story above because Steam has not yet added these cards to the database. However, it can take Valve months to update the database with new cards, and it does not add them to a consistent market share level. In all cases, however, AMD availability is poor.
It is not clear whether these deficiencies are mainly driven by cryptocurrency-related demand, by low returns or by a mixture of both. We can get better data at that point when Nvidia finally gives Q1 results, but it will not happen in a couple of months. Several GPU manufacturers have indicated that they can not get sufficient GPU inventory, but the demand for cryptocurrency is also known at the moment.
We do not usually refer to SHS for hardware information due to doubts about its accuracy. But to the extent that we can trust that this data shows us something, it is not positive what it shows. Four to five cards with an increase of> 0.17 percent can be a positive sign that we are on our way back to normal. A single GPU illustrates just how far we have to go.