While everyone agrees that NVIDIA completely botched its RTX 50 launch, the greens are communicating on the issue, providing some comparisons that will leave you scratching your head. In the first 5 weeks since the launch of the RTX 5090 & 5080, you couldn’t miss it: availability has been at the heart of tensions. All the indications are that low production of the new RTX 50 cards is at the root of this frustration and soaring prices. But NVIDIA has a different interpretation of the launch. They claim to have shipped twice as many RTX 50 GPUs as RTX 40s over the same 5-week launch period.
RTX 50: Are we living in a parallel world?
To back up its assertion, Nvidia delivers its figures with in-house graphics. In particular, there’s a chart showing cumulative GPU shipments in the first 5 weeks after launch.
Obviously, many people will fall off their chairs at the sight of this graph (we don’t know if Nvidia’s data is boosted by DLSS 4.0). But in reality, given the lack of detail on the figures (there’s no information other than the number of weeks on the x-axis…), NVIDIA isn’t lying to us, but it’s still confusing us a little.
Let’s put this comparison into context:
In the space of 5 weeks, NVIDIA literally threw down the gauntlet, launching no fewer than 4 references (RTX 5090, 5080, 5070 Ti and 5070). The first two on January 30, 2025 and the last two on February 20 and March 5, 2025 respectively.
By way of comparison, for RTX 40, the 5-week launch took place with only RTX 4090 on October 12, 2022, and 4080 on November 16, 2022, i.e. almost 5 weeks later. So in reality, the 4090 was the only reference available for 5 weeks. Factually, all the GPUs delivered by NVIDIA were 4090s vs. 4 different GPUs for the RTX 50 launch. This is, I’m sure you’ll agree, another way of reading this graph, and one that’s undoubtedly more objective, don’t you think?
