ASRock has confirmed the release of the Radeon RX 9050 Challenger graphics card, which has become the youngest model in AMD's line of graphics accelerators based on the RDNA 4 architecture. The final technical specifications turned out to be more modest than the early leaks: the device received 16 computing units and 1024 stream processors, which is half the May forecasts. At the same time, the clock speeds are in line with expectations: the base frequency is 1920 megahertz with acceleration up to 2600 megahertz. The theoretical performance in floating-point computing of the FP32 reaches 5.3 teraflops, which is approximately 40 percent of the performance of the reference model RX 9060 XT.

Despite the initial level of positioning, the graphics card retains the full range of features of the RDNA 4 architecture, including third-generation ray tracing accelerators and second-generation AI accelerators. The Challenger model is equipped with 8 gigabytes of GDDR6 memory with a clock speed of 18 gigabits per second and a 128-bit bus, providing a bandwidth of 288 gigabytes per second. At the same time, information appeared on the official website about a version with 4 gigabytes of memory designed for ready-made OEM assemblies, which is likely to receive a 64-bit bus and half the bandwidth.

The cooling system design includes two fans with a card size of 249 by 132 by 41 millimeters and a weight of 645 grams. The power supply is provided via one 8-pin connector, the recommended power of the power supply is 450 watts. Two DisplayPort 2.1a and one HDMI 2.1b ports are available from the video outputs. The exact value of the power consumption has not been disclosed, but the recommended power of the power supply indicates the positioning of the new product in the entry-level segment for mass users.