Apple is a master of wordplay, which often makes even the most rational users think that their MacBook Pro from last year is suddenly hopelessly outdated. After all, even when introducing the new M5 chip, he once again brought out his traditional barrage of superlatives about "the biggest leap in history" Apple Silicon", "the fastest cores in the world" and "4x higher GPU performance". But as the first real-world tests show, the difference between the M4 and M5 generations is in practice much less dramatic than it might seem.
You could be interested in
Apple built the M5 on the third generation 3nm process and added a new 10-core GPU architecture with a neural accelerator in each core. It also increased memory bandwidth by about 30% to 153 GB/s and improved the 16-core Neural Engine. Benchmarks indeed confirm that this is a monster chip, at least on paper.
For example, according to the Black testmagic Disk Speed M5 can write data almost twice as fast as M4 and the difference in reading is even more than 130%. In Geekbench AI GPU then M5 achieved double the score of M4 and in 3DMark Solar Bay Extreme test managed to render scenes half as fast. So in numbers impressive difference, which sounds like a clear reason to upgrade.
Gallery
But once you get down to real-world use, the marketing magic quickly fades. Macworld ran a series of practical tests that show that in common creative tasks, the differences shrink to a matter of seconds. Exporting 562 photos from Lightroom took 66 seconds on the M4, compared to 56 on the M5. A large GarageBand project was also quite a feat – 65 vs. 56 seconds, to be exact. And even a complex render in Blender was only 10 seconds faster on the M5.
You could be interested in
In other words, unless you're exporting 4K videos ten times a day or doing complex AI calculations, the M5 won't offer you much extra in practice. Yes, the benchmarks look great, but for most users, the difference will be more noticeable on paper than in everyday use. Apple Of course, he can claim that every new chip is “revolutionary” because his own comparisons always refer to perfect laboratory conditions. But when the real-world time savings are barely a few seconds between two sips of coffee, it’s worth considering whether the upgrade makes sense.
Apple states that the M5 is 15% faster than the M4 in multicore. Macworld measured a 15% improvement in Lightroom and GaraBand. What kind of drama are you causing here?
Nobody is making any accusations here. The article just shows the difference between the machines with illustrative examples.
"Apple "The reality is exactly what it seems like, but the performance difference is a little different." Apple It's tempting. A 15% speedup for operations that take less than a minute will logically be hard to notice.
"When introducing the new M5 chip, he once again unleashed his traditional barrage of superlatives about "the biggest leap in history" Apple Silicon", "the fastest cores in the world" and "4x higher GPU performance"" - I think an uninformed user might have quite high expectations from this.
After all, those cores are the most powerful in the world. And the 4x higher GPU performance was supposed to apply to AI operations, which Macworld didn't test, so it didn't refute that claim.