NVIDIA Consumer GPU Market Update: RTX 5090 Asking Prices Up 60% Since August
NVIDIA consumer GPU prices have moved sharply higher over the past two months, and the trend extends well beyond the latest RTX 50 series. A comparison of the lowest qualifying marketplace asking prices on August 8 and October 11 found that none of the 60 price floors tracked had meaningfully declined.
The RTX 5090 saw the largest increase. Its lowest used asking price climbed 60%, from $3,680 to $5,879.80, while the lowest new-condition asking price rose 64%, from $3,004 to $4,920.99. The new-condition price floor is now about 146% above NVIDIA’s $1,999 launch MSRP.
The increases are not limited to flagship cards. Used RTX 40-series asking-price floors rose by a median of around 30%, with the RTX 4070 recording the largest increase at 43%. Even older RTX 30-series cards have become more expensive in the used market, suggesting that the change extends across several generations of NVIDIA hardware.
Why are prices rising? Several factors may be contributing. Reports that NVIDIA has redirected RTX 5090-class chips toward professional AI products could mean less supply for consumers, although the reported production changes remain unconfirmed. Rising memory costs, higher component expenses and demand from buyers building AI workstations may also play a role. However, memory costs alone do not explain the scale of the RTX 5090’s price increase.
One of the report’s most interesting findings is that used asking prices now exceed new-condition asking prices for five GPU models, including the RTX 5090 and RTX 4090. This can happen when new stock is scarce and individual sellers set prices based on expected replacement costs. It does not mean buyers are actually paying those amounts.
That distinction is essential: the report tracks the lowest qualifying advertised prices, not completed transactions or confirmed resale values. A higher asking price signals a change in seller expectations and market conditions, but it cannot establish what a GPU will ultimately sell for.
For businesses and individuals holding retired graphics cards, the findings suggest that older valuations may need to be revisited. In particular, GPUs with 16GB or more of VRAM may deserve closer attention, as memory capacity can remain valuable for certain AI and workstation workloads.
Full report:
https://www.buysellram.com/blog/nvidia-consumer-gpu-market-update/
#NVIDIA #GPU #RTX5090 #RTX4090 #AIInfrastructure #GPUMarket #Hardware
