NVIDIA has selected Samsung’s GDDR7 memory solution as the preferred choice for its upcoming GeForce RTX 50 “Blackwell” gaming lineup. This partnership marks a significant win for Samsung, which had struggled to gain traction in the adoption of its AI-focused HBM solutions.
The adoption of Samsung’s GDDR7 memory is seen as a major success story for the Korean semiconductor firm, which competed with SK Hynix and Micron for the coveted spot. The latter, however, opted for NVIDIA’s existing GDDR5X and GDDR6X memories in its RTX 40 and RTX 30 GPUs.
NVIDIA plans to use both 28 Gbps and 32 Gbps GDDR7 memory dies across its next-gen lineup, with the RTX 5080 utilizing the fastest memory dies. The RTX 5090 will feature a slightly slower 28 Gbps solution due to increased memory chip usage. Samsung has already announced 16Gb and 24Gb memory modules based on the GDDR7 standard, which are expected to debut later.
The GeForce RTX 50 “Blackwell” lineup is set to launch at CES 2025, with initial products hitting shelves in January.
Source: https://wccftech.com/nvidia-geforce-rtx-50-blackwell-gpus-utilize-samsung-gddr7-memory-chips