Cortex X4 Part 1.5

I wish I had followed up on my previous post. Now I completely forgotten what part 2 was suppose to be. So I have this Part 1.5 as an update for myself.

We now have rumours and hints that Cortex X4, on 3nm may be able to get 15% Clock Speed increase, and possibly 15% or 20% IPC increase. On a 15%/15% increase that would give GB5 1950. That is faster than the current M2 and A16 @1880. And if it was closer to 20% IPC increase, it should break the GB5 2000 barrier.

Now of course it is competing against A17. So I would not be surprised if X4 still end up being slightly slower. May be within 10% range but it becomes somewhat irrelevant.

The good thing is that X4 is coming directly from ARM. Anyone could buy the design and ship a SoC with it. Mediatek will use it. And it is only a matter of time before it filters down to other market segment. No longer do you have to buy an Apple devices for good ARM CPU. Which is what 99.99999% of the internet have been insisting for the past three years.

Another interesting observation is the Qualcomm / Nuvia Core, was originally targeting @ GB5 2000. Now this could be achieved with ARM Cortex X4. And by 2024, the release date of Nuvia Core, it could be up against a Cortex X5 design. Does it make sense for Qualcomm to continue to invest in their own CPU uArch design? This isn't the first time Qualcomm ends up in a similar scenario. Where they decide to invest and ultimately ARM has ready to ship solution available at a lower TCO.

Cortex X4 is not just exciting for Smartphone, but also for ARM on PC.