New information is beginning to circulate about the future NVIDIA GeForce RTX 60 graphics cards and their Rubin architecture. While the gains in “raw” performance remain relatively modest, the real generational leap could be in ray tracing and path tracing, where NVIDIA is aiming for a spectacular jump.
RTX 60: modest gains in rasterization, but massive in ray tracing
According to the latest leaks from the industry, the RTX 60 would offer an improvement of between 30 and 35% in traditional rasterization compared to the RTX 50 based on the Blackwell architecture. A significant gain, but far from revolutionary.
On the other hand, the story completely changes when it comes to ray tracing. NVIDIA is reportedly preparing a much more ambitious leap, with performance that could simply be doubled in this area. This progression is made possible by the introduction of 5th generation RT cores and 6th generation Tensor Cores, tailored for AI and advanced rendering workloads.
Higher specifications, but no revolution in memory
Regarding technical specifications, several models are beginning to take shape. The GeForce RTX 6090 would be based on a 3nm TSMC GR202 GPU, with 24,576 CUDA Cores, approximately 13% more than the RTX 5090. It would retain a similar memory configuration with 32 GB of GDDR7 on a 512-bit bus.
The RTX 6080 and RTX 6070 would also see improvements in bandwidth. The former would come with 20 GB of GDDR7 on a 320-bit bus, while the latter would offer 16 GB on a 256-bit interface. While the number of cores remains unknown at this stage, the increase in memory bus and capacities hints at an interesting bandwidth gain.
An architecture designed to support DLSS 5
Beyond the numbers, these evolutions are part of a larger logic: to accompany the arrival of DLSS 5. This new generation of technology would no longer be limited to image reconstruction but would focus on AI-assisted rendering.
In this context, the RTX 60 GPUs appear to be the first GPUs truly designed to effectively harness this technology. With an IPC gain, better Tensor Core management, and increased power in ray tracing, they should make DLSS 5 viable on a single graphics card.
An uncertain launch
The question of the timeline remains. While some sources suggest a launch in the second half of 2027, others do not rule out a delay until 2028. This is due to a tense industrial context, marked by memory shortages and the priority given to AI infrastructures.
In any case, this initial information outlines a generation of GPUs less focused on brute force and more on algorithmic efficiency. A logical transition as AI gradually becomes the core of graphic innovation at NVIDIA.
