There’s a buzz around NVIDIA planning a late-year release for their much-anticipated RTX 50 series of GPUs. The company had previously put the spotlight on its Blackwell GPU for data centers, but a release date remains up in the air. Now, the conversation is shifting towards a launch that’s got gamers talking: the GeForce RTX 50 series.
Earlier in the year, NVIDIA unveiled a trio of new GeForce RTX 40 SUPER graphics cards, providing an update to their lineup for serious gaming aficionados. Presently, the RTX 40 series boasts 12 different models, which is a smaller variety compared to the 20 different models from the RTX 30 series. However, if we set aside the LHR (Lite Hash Rate) variants, the diversity in the range remains consistent. This points to the possibility that NVIDIA might have wrapped up development on the RTX 40 series.
According to Money UDN, Industry insiders and NVIDIA’s hardware partners are keeping their eyes peeled for a potential launch of premium RTX 50 models later this year, with special attention on the RTX 5080 and RTX 5090 cards. Yet, it’s important to bear in mind that the company is known for its eleventh-hour changes and has a history of adjusting specifications—or even scrapping products altogether—just before they’re slated to hit the shelves.
Board manufacturers anticipate the launch of NVIDIA’s latest RTX 50 series graphics cards in the fourth quarter, initially focusing on high-end products like the 5090 and 5080. Analysts are optimistic that NVIDIA’s biennial major GPU upgrades will boost shipment volumes for board manufacturers this year and drive up the average selling price (ASP) of graphics card products.
– Money UDN (translated)
A prominent source known for NVIDIA leaks concurs with the prediction of a RTX 50 series debut later this year. This leaker, Kopite, was notably the first to suggest that the Blackwell GPU is expected to integrate GDDR7 memory and feature an expansive 512-bit memory bus, which would be a first for the GB202 GPU model.
Speculation is mounting that NVIDIA’s GB202 GPU could be equipped with as many as 192 Streaming Multiprocessors, whereas the GB203 might have 96 Streaming Multiprocessors. If NVIDIA maintains its current FP32 (CUDA) cores configuration, it would mean the GB202 could have 24,576 cores and the GB203 would have 12,888 cores. To ensure high production yields, it’s anticipated that neither the RTX 5090 nor the 5080 will offer these maximum configurations.
