AMD Radeon RX 7600 Challenges RTX 4060 Ti For Budget GPU Dominance – SlashGear

AMD has fitted 32 compute units inside the Radeon RX 7600, ticking at a peak frequency of 2.6 GHz, alongside 32 ray-tracing acceleration units and 64 AI processing units. AMD touts 21.75 TFLOPs of peak single-precision compute performance, and the shaders on Nvidia’s GeForce RTX 4060 Ti can deliver 22 TFLOPs. You get 32MB of AMD’s second-generation Infinity Cache, and there’s 8GB of VRAM inside, the same as the base model of Nvidia’s aforementioned GPU.

The memory bandwidth on this one peaks at 288 GB/s, and it also serves the usual set of AMD tech such as FidelityFX Super Resolution, Smart Access Memory, and Readeon Boost. On the creative side of things, AMD says the Radeon RX 7600 GPU offers up to 32% in certain scenarios like video encoding. AV1 encoding and decoding is also part of the package, while AI-based tasks also get a sizeable boost in performance.

Based on raw capabilities, it looks like the Radeon RX 7600 is a repackaged Radeon 6600 XT, but it’s disappointing that the VRAM is capped at 8GB. It’s surprising that AMD has not only managed to fit respectable hardware inside the Radeon RX 7600 but has also dropped its price significantly vis-a-vis its predecessor. AMD charges $269 for its latest entry-level GPU, while the Radeon RX 6600 hit the shelves priced at $329. AMD is also undercutting Nvidia, which is putting the GeForce RTX 4060 series on the market starting at $299.