Rumors Suggest AMD Will Launch 32 GB Radeon RX 9070 XT GPU This Year

Advanced Micro Devices (AMD) is gearing up to unveil a cutting-edge GPU, the Radeon RX 9070 XT, potentially equipped with a whopping 32 GB of VRAM. Although it may initially appear to be a professional-grade card, this model is primarily designed with gaming enthusiasts in mind.

The Anticipated Launch of the AMD Radeon RX 9070 XT with 32 GB VRAM

Despite AMD’s existing lineup lacking any RDNA 4 GPU releases, whispers about upcoming models continue to circulate. The RX 9000 series, which is based on the RDNA 4 architecture, is expected to debut with the Navi 48 and Navi 44 variants, starting with the RX 9070 XT and RX 9070 set for March.

Initial reports suggest that these GPUs will be equipped with 16 GB of GDDR6 VRAM. However, speculation arises that AMD is preparing to launch another RX 9070 variant boasting double the memory at 32 GB. This information comes from a user on the Chiphell Forum, zhangzhonghao, who has indicated that this version may become available in the first half of 2025.

According to the leaker’s clarification, this advanced GPU may launch before the end of Q2 2025. Although there’s initial speculation about the RX 9070 XT being a professional-oriented card, the leaker asserts that it will serve mainly gaming applications while also catering to AI tasks requiring substantial VRAM.

zhangzhonghao RX 9070 32 GB 1 — Image Source: Chiphell Forums

To achieve the ambitious 32 GB VRAM capacity, AMD would likely need to implement 16 dual GB GDDR6 memory modules, a challenging endeavor that may necessitate placing some of them on the underside of the card. Additionally, current market conditions do not offer 4 GB GDDR6 modules, which would likely inflate the GPU’s production costs due to these technical hurdles.

It’s noteworthy that the memory bus width is expected to remain at 256 bits, and the memory speed of 20 Gbps will stay consistent, ensuring that the total memory bandwidth does not fluctuate. Although the RX 9070 XT might appear impressive on paper, its capability to effectively utilize such significant VRAM for gaming may be limited. Nonetheless, this GPU could excel in memory-intensive applications like large language models (LLMs), enhancing performance in AI-related tasks.

Source & Images