iPhone 17 Pro Max On-device AI speed — does the processor and RAM matter?

[dmaxdmax]

I’m really concerned with AI translation speed and accuracy. I know that the better chip and extra RAM matter for gaming and Pro functions but what about AI? In 2 years will I notice the difference?

 

[Sheepish-Lord]

I’m really concerned with AI translation speed and accuracy. I know that the better chip and extra RAM matter for gaming and Pro functions but what about AI? In 2 years will I notice the difference?

Since you're talking about AI let's have ChatGPT Plus answer it ha:

"AI translation can run in two main ways. If it’s on-device (offline), your chip and RAM matter a lot, because faster processors and more memory make real-time translation smoother. If it’s cloud-based (like Google Translate or DeepL), most of the work happens on remote servers, so your device hardware doesn’t matter much except for internet speed.

Right now, most consumer translation apps still lean on the cloud, but phone and computer makers are adding NPUs (neural processing units) that will handle more AI locally. In two years, newer devices will probably be able to run bigger translation models offline with better speed. RAM will also help because AI models are memory-hungry.

Accuracy is mostly about better models, and those usually come from cloud updates, so even older devices will benefit. The main difference with new hardware will be speed, especially for real-time voice translation or offline use.

Bottom line: If you mostly use cloud-based translation, you probably won’t notice much difference over the next two years. If you want offline or real-time translation, a newer device with a strong NPU and more RAM will feel faster and smoother."

 

[NotApplicable]

RAM definitely, because these aren't discrete GPUs with their own memory. But the impact of the CPU cores relative to the GPU cores, that's another question. Note: these are both "processor" cores.