Rumor: Switch makes use of NVIDIA’s Maxwell architecture
According to a GamesBeat report, Switch will be making use of NVIDIA’s Maxwell architecture for its graphics as opposed to the chipmaker’s newer Pascal technology. The site writes that Wii U’s lack of success prompted Nintendo to move quickly, and avoid waiting on the updated Pascal version of the graphics technology. Had Pascal been pursued, Switch would not be able to launch in March as currently planned. The chip would have required a redesign to fit within Nintendo’s “mobile, low-power constraints” as well.
GamesBeat goes on to say that “Switch doesn’t have as much visual horsepower as the PS4 when played on a television, and it may not be able to handle 4K graphics, either.” It’s unclear if Switch is weaker than the Xbox One since “performance may be close.”
Writing more about Switch’s power, GamesBeat anticipates “the Nintendo Switch to be more than 1 teraflop in performance, but far less than the 6 teraflops that Microsoft is promising for Skorpio. The PS4 is around 1.8 teraflops, and it has much better memory bandwidth performance as well compared to the Switch.”
While that may sound like a downer, going with Maxwell means Switch will initially be cheaper than other brand-new systems. NVIDIA is embedding the Maxwell graphics in the ARM-based Tegra processor. This single chip “consumes less space, uses less power, and costs less than having two chips in the system.”
Another important aspect of Nintendo preparing for Switch as soon as possible concerns other game makers. The big N wanted to ensure that other teams would have plenty of time to create projects with development kits. GamesBeat has heard that these kits feature “a Tegra X1-based system”, and “the final chip will be a custom version of the X1.”