I compared the M5 MacBook Pro to older models, and it made my $2,500 M1 Max weep
Briefly

I compared the M5 MacBook Pro to older models, and it made my $2,500 M1 Max weep
"Of course, that last task was the least stressful, as Apple made the reason to upgrade to the hot, new MacBook Pro this year as straightforward as ever. If your current system isn't holding up, you can buy the trustee MacBook Pro, with its polished aluminum design, smooth-as-ever ProMotion display, and an updated M5 chipset, all for the same price as last year's model -- and the year before."
"For starters, beyond the M5 chipset, the new 14-inch MacBook Pro features the same core of internal hardware as the M4 model, from the cooling fans to the placement of memory and battery. However, subtle capacity changes make a noticeable difference to performance, from faster unified memory bandwidth (153 GB/s from 120 GB/s) to faster SSD speeds to up to twice the storage cap (4TB from 2TB). Each of the 10 GPU cores also receives Neural Accelerators for faster AI/ML processing."
"In fact, the M5 MacBook Pro looks identical to the M4 variant that I reviewed last year; as I was benchmarking the two, I had to look closely at the keyboards for any glistening left by millions of fingertaps to distinguish the M4 model. After several rounds of performance testing, here's how the M5 fared with older Macs, including my $2,500 M1 Max MacBook Pro from just three years ago."
The 14-inch M5 MacBook Pro keeps the same polished aluminum design, ProMotion display, cooling system and internal layout as the M4 model while upgrading the chipset. Unified memory bandwidth increases from 120 GB/s to 153 GB/s, SSD speeds improve, and the maximum storage doubles to 4TB. Each of the 10 GPU cores gains Neural Accelerators for faster AI/ML processing. The highest M5 configuration targets demanding users, though most workflows will be well served by the base model. Benchmarks compare the M5’s performance to older Macs, including recent M1 Max systems.
Read at ZDNET
Unable to calculate read time
[
|
]