Yesterday, we wrote about Dave Lee’s experiment with his high-end Intel i9, showing that under load, it got hot enough to trigger throttling, which slowed it down enough to be a problem.
One of the arguments made about Dave’s experiment is that it was specific to Adobe Premiere and possibly due to Premiere not being optimized for macOS.
Jeff Benjamin, writing for 9to5Mac, put his i9 MacBook Pro on the line to test this theory. Jeff did his testing with Apple’s own Final Cut Pro X (and a different model freezer).
I won’t spoil the conclusion (follow the headline link and read through Jeff’s well written walkthrough, don’t miss the update with John Poole’s take at the very end), but this line made me laugh:
The freezer is good for short term performance on the MacBook Pro, but may prove to be an inconvenient/hazardous working environment (condensation is bad, folks).
Well said.
I believe the update pretty much invalidates the entire story.
The time is still interesting, but his graph where the CPU was throttled actually shows where the CPU is idle because an on-chip special device is being used.
It makes the graph pretty much pointless.
Oh, yeah, the clock rate argument is indeed pointless.
Unfortunately, Reddit is afire now of people tracking clock rates.
Can you explain?
What I get from the story is that the 6-core config is not necessarily better than 4-core for sustained high-CPU workloads. I’d guess the main performance benefit for 6 cores would be for bursty use. For example, compiling in Xcode.
That doesn’t invalidate the fact that the i9, 6-core, 2.9GHz is being heavily throttled when at full tilt 🙂
I’d also like to see benchmarks of other laptops.
I’m hopeful one of the hardware sites go deeper into this, but this article was a fun and informative read.
If true (and I suspect that to some extent this situation is real), then I think we need to revisit the argument that a lot of people wanting high-performance in a laptop would be willing to trade size for power. The weight of my 17″ MBP never bothered me. It probably would now, because I’ve gotten used to a new 15″ model… but I could renormalize quickly if it meant more screen and more power. Hell, I used to have a 14 pound Toshiba. It made me laugh to hear people complaining that a 6 pound MBP was too heavy. As they later complained about the weight of the first iPad. Yeah, light gadgets are fun to tote. But engineering is about trade-offs, a tug of war so to speak; and weight shouldn’t always win every tug of war.
I think this states the issue, which I’ve heard more than once. https://www.computerworld.com/article/3291760/apple-mac/about-that-macbook-pro-core-i9-throttling-story.html