AMD has quietly introduced a new way to process shaders in its RDNA GPUs that can slash game load times by up to 95%, but it comes with a catch: more heat and power draw. The feature, called Advanced Shader Delivery, is now baked into every RDNA GPU, from the desktop RX 7900 XTX down to mobile parts.
The improvement isn't just about faster load screens—it's about how data moves through the GPU pipeline. Instead of relying solely on memory bandwidth, AMD is pre-processing shaders in a way that reduces the amount of work the GPU has to do at runtime. That means less stuttering when jumping between scenes or levels, but it also means more computational effort before the game even starts.
For users who prioritize speed over efficiency, this could be a game-changer. A 95% reduction in load time means skipping through menus and getting straight into gameplay with barely a pause. But for those running tight power budgets—whether on a laptop or a silent desktop setup—the extra heat and performance drain might not be worth it.
AMD hasn't shared exact power consumption numbers, but early benchmarks suggest that this feature could add anywhere from 5% to 10% more load to the GPU under heavy workloads. That's enough to push some mobile GPUs closer to their thermal limits or require a desktop card to ramp up its fan speeds noticeably.
What's still unclear is whether this will be an opt-in feature in games, meaning users could disable it if they prefer cooler, quieter operation. AMD has also been tight-lipped about whether future updates or driver tweaks will let users adjust the balance between speed and efficiency. For now, the focus seems to be on raw performance gains.
The real question is whether this will become a standard in the industry or if it'll remain an AMD-specific quirk. If other GPU makers see similar benefits, we could start seeing faster load times across the board—but for now, it's just another reason to choose an RDNA card if speed is your priority.
