Into the BLACKWELL
What lurks at the bottom of Nvidia’s deepest GPU technology? Jarred Walton is our man for technological spelunking.
As you read this, Nvidia will be well on its way to failing to supply demand for its latest RTX 50-series of GPUs. The most capable and performant GPU on the market, it’s also the most expensive and questionably a lacklustre replacement for the 2022 RTX 40 series.
The January 2025 launch of the top-end RTX 5090 (see page 19) and RTX 5080 saw boards selling out instantly despite higher-than-recommended retail pricing. The more affordable and available RTX 5070 will be available as you read this and we’ll have thoughts on these models next issue.
Before looking at actual cards, we want to look at the architectural changes in Blackwell RTX 50-series GPUs. We also have a neural rendering and DLSS 4 deep dive, the RTX 50-series Founders Edition cards, RTX AI PCs and generative AI for games, Blackwell for professionals and creators, and Blackwell testing.
The official Nvidia release on the architecture didn’t provide a ton of detail on some aspects, but there are a lot of things that don’t seem to have changed much from the RTX 40-series Ada Lovelace architecture. Most of the upgrades and enhancements tend to be around AI and various Neural Rendering technologies. The main goals for Blackwell were: optimise for new neural workloads, reduce memory footprint, new quality of service capabilities, and energy efficiency. Those all sound like good things, but outside of the RTX 5090 with its significantly larger GPU die – 744mm2 compared to 608mm2 on the 4090 – a lot of the upgrades feel more incremental.
That’s not to say things haven’t changed at all. The 4th-Gen RT cores have twice the ray triangle intersect rates of Ada. They’re also built for Mega Geometry, which could help future Unreal Engine 5 games run better. The GPU shaders are also enhanced for Neural Shaders, and there are some other new additions.