NVIDIA is reportedly steamrolling ahead with its plans of getting the GeForce RTX 5050 out by this July. Recently, a rumour from a prominent leakster suggests that the entry-level GPU will not be using GDDR7, but GDDR6 memory.
More specifically, and according to MEGASizeGPU, the RTX 5050 could end up being fitted with GDDR6 that runs at a frequency of 20Gbps. That’s the same kind of memory used by AMD’s RDNA4-powered Radeon RX 9000 Series GPU, and has actually proven to run rather fast with the 9070 XT and 9060 XT.
However you may view it, we don’t exactly find this decision to be all that surprising – the RTX 5050 is, by any and all assumptions, set to be the entry-level, barebones GPU of the Blackwell architecture. It would make more sense for NVIDIA to fit the card with GDDR7 memory. What is arguable till today is why NVIDIA is still adamant on making 8GB of memory the floor capacity.
At 8GB, and the memory standard notwithstanding, the RTX 5050 would almost certainly be relegated to Full HD gaming at most. Of course, we’re just making speculations here, and as with all rumours, we’re asking you to moderate your consumption of Ajinomoto for now, or at least until NVIDIA makes an official announcement about this.
(Source: X, Videocardz)
