70%
Probable
The "upgraded" GeForce RTX 5050 is expected to arrive in June this year at the Computex event.
NVIDIA Will Reportedly Debut its 9 GB RTX 5050 GPU at Computex; Same GPU, but Newer GDDR7 Memory With an Additional Gigabyte of VRAM
Instead of increasing the VRAM to 12 GB, NVIDIA is straight up reducing RTX 5050's memory bus width to offer 9 GB VRAM capacity. The original version comes with 8 GB GDDR6 memory and is the only RTX 50 series GPU in the series to run on previous-gen GDDR memory. That said, with the new GPU planned for launch, gamers might be interested to know when they can see it on the shelves.
It appears that we may see the RTX 5050 9 GB card in a few months. According to Benchlife, NVIDIA is reportedly launching the GeForce RTX 5050 9 GB GDDR7 card at Computex. Computex 2026 is going to be held from June 2 to June 5 in Taipei. So, we have nearly three months before the launch, but as of now, NVIDIA hasn't confirmed any such plans.
Within NVIDIA’s Blackwell GPU architecture, the GeForce RTX 5050 was originally the only model planned to use GDDR6 memory. However, it is now believed that NVIDIA could introduce a version of the RTX 5050 using GDDR7 memory around the time of Computex 2026.
The main reason for this change is reportedly GDDR6 supply and demand conditions, rather than a simple upgrade in capacity. This does not mean the memory would jump from 8GB to 12GB, but instead move to a 9GB configuration using the new 3GB GDDR7 memory chips.
- Benchlife (Machine Translated)
The variant is a weirdly interesting GPU that gets its memory bus downgraded from 128-bit to 96-bit to make sure gamers won't get a "much" higher VRAM capacity. The GPU is reportedly going to use the same GB207 die and will retain the same specs, including 2560 CUDA cores, but will utilize 3 GB GDDR7 memory modules to reach 9 GB memory capacity. If NVIDIA wanted, this could be a 12 GB GDDR7 card through the new configuration, but we know that's not something it wants at the moment.
Now, considering the usage of 3 GB GDDR7 memory modules, it appears that NVIDIA is now in a position to start deploying these on the existing cards. We have recently seen that some vendors have started listing RTX 5070 laptop GPU with 12 GB GPU, which is only possible if NVIDIA uses 3 GB modules. As of now, the RTX 5090 laptop GPU is the only card to have 3 GB modules, and seeing 3 GB modules on RTX 5070 laptop GPU is unlikely. Still, since the whereabouts of the RTX 50 Super series are unknown, existing cards might be the ones where NVIDIA may use higher capacity modules.
News Source: Benchlife
About the author: Sarfraz Khan is a hardware reporter with a focus on PC components and the builder community. With years of experience writing about PC hardware and laptops, his work has been featured on several reputable technology publications. Sarfraz's hands-on experience is demonstrated through his first-person accounts of using and comparing different hardware configurations, providing practical and relatable insights for everyday users. His technical analysis is respected by peers in the enthusiast community and has been cited by specialized hardware sites such as Germany's Igor's Lab.
Follow Wccftech on Google to get more of our news coverage in your feeds.
