“We don’t wait,” says Millet, “For someone to come and give us a list of things to do. We are busy in the background, looking for obvious ways that we can enhance and enable the capabilities of the device, but it’s never done in isolation.” He adds: “We didn’t invent the neural engine and wonder how it was going to be used. The concept starts years in advance. We work closely with our software team to make sure it’s all going to be lined up, make sure the camera system, all the APIs are available so that when it comes out, it’s a complete story, it’s a complete solution.”
“Putting a much, much bigger GPU in a device that doesn’t need it doesn’t help the battery life. So what we do with those devices is to make sure that we are putting in the right size GPU to enable the experiences that are appropriate for that device, that screen size and the software that runs on it, and the battery life. And then for the bigger devices, the Pro devices with the bigger displays and the more complex camera systems, we know we are going to need a little bit of additional performance, so that’s where we can extend the GPU,” says Millet, adding that it is his team’s job to build that scalability story so that the software team doesn’t see something different, but something very familiar.
“We spend a lot of time looking at what we call all-day battery life, that’s what really matters to people it’s not the capacity of the batteries,” he says, underlining how it is really about how much useful work do we get out of the battery.[Via]