Is the iGPU on AMD Ryzen 8000 series exceptionally powerful, comparable to RTX 3060, enabling AAA gaming without a discrete GPU?

Currently, the Ryzen 7000 APU 'Phoenix' series integrates a GPU with 12 cores, translating to 768 stream processors (referred to as CUDA by NVIDIA), with a clock speed of up to 2700 MHz, delivering computing power of up to 8.90 TFLOPs. However, it appears that AMD is not stopping there; on the AMD Ryzen 8000 'Strix Point' next-generation, the number of GPU cores is further increased for even more impressive graphics performance.

According to leaked information, the AMD Ryzen 8000 'Strix Point' will feature a single CCD with a dual CCX configuration. One of these CCXs will have 4 cores & 8 threads based on the core architecture.

Zen 5, while the other will feature 8 cores & 16 threads based on the Zen 5C CPU core. Zen 5 CCX comes with a full 16 MB L3 cache, whereas Zen 5C CCX has an 8 MB L3 cache, resulting in a total L3 cache of 24 MB on the chip. The clock speeds of both CCXs are expected to remain the same, but Zen 5C cores will provide slightly better energy efficiency.

On the iGPU front, the AMD Ryzen 8000 'Strix Point' APUs will be configured with the AMD RDNA 3.5 GPU core, featuring 8 WGP (Working Group Processor) and a total of 16 compute units for a maximum of 1024 stream processors. The computational power of this iGPU is estimated to reach ~11 TFLOPs, equivalent to an RTX 3060, capable of handling all current AAA games.

However, it's important to note that discrete GPUs, designed with higher bandwidth VRAM and higher power consumption, maintain high performance for a longer duration. iGPUs commonly share system RAM with lower bandwidth and are unlikely to sustain peak performance for extended periods.

The APU AMD Ryzen 8000 'Strix Point' will compete against Intel's Arrow Lake and Lunar Lake in the laptop segment.