My guess is this has to do with core stability. Dropouts occur more easily on the newer chips because the cores behave more erratically, even though they’re faster overall.
The article pretty clearly states it seems to be a software issue with either the programs themselves, or the OS. Where are you getting suspicion about the processors running erratically?
Because this was reported last time with the M2 vs M1. The major performance improvements came from increased clocks, at the expense of some core stability.
These new cpus can bounce tracks faster, but are apparently worse at real-time processing. Most DAWs can only process individual tracks on a single thread.
If you’re monitoring while using heavy plugins then you may be more prone to dropouts on the newer chips.
The different # of P and E cores plays more into the multithreaded performance
Imagine 2 runners racing. One can trip and fall, get up and still win. Faster but less stable. For real time signal processing you have to maintain a minimum clock or dropouts and artifacts will occur.
My guess is this has to do with core stability. Dropouts occur more easily on the newer chips because the cores behave more erratically, even though they’re faster overall.
The article pretty clearly states it seems to be a software issue with either the programs themselves, or the OS. Where are you getting suspicion about the processors running erratically?
Because this was reported last time with the M2 vs M1. The major performance improvements came from increased clocks, at the expense of some core stability.
These new cpus can bounce tracks faster, but are apparently worse at real-time processing. Most DAWs can only process individual tracks on a single thread.
If you’re monitoring while using heavy plugins then you may be more prone to dropouts on the newer chips.
The different # of P and E cores plays more into the multithreaded performance
Doesn’t the article indicate the difference had to do with utilization of e-cores? Since the newer Apple silicon has more e-cores and fewer p-cores?
What does this even mean?
Are you trying to say something about latency versus throughput?
Imagine 2 runners racing. One can trip and fall, get up and still win. Faster but less stable. For real time signal processing you have to maintain a minimum clock or dropouts and artifacts will occur.
It means the clock frequency varies more than M1