Here’s a fact about boutique laptops that doesn’t get nearly enough attention: Generally speaking, the manufacturers of these systems will cheerfully outfit them with CPUs and GPUs that demand more cooling performance than the laptop can provide. As a result, the highest-end systems available often throttle back their own performance to avoid damage. How much of an impact this has on overall system performance varies from vendor to vendor and system to system, but it’s a rare laptop that avoids this issue altogether — and if they do avoid it, it’s often by cranking system fans up to decibel levels that could cause long term hearing loss.
Manufacturers have a perverse incentive to engage in these practices. Enthusiasts are believed to buy systems that advertise the most powerful hardware, which means systems that don’t ship these high-end configurations are theoretically at risk of losing sales to those who do — even if the lower-end configuration actually promises more consistent performance and better overall results.
That brings us to our current situation. YouTuber David Lee has created a video showing how the new MacBook Pro’s Core i9 performance lags well behind last year’s Core i7 CPU, as well as behind an equivalent PC system.
The gap between Premiere Pro optimizations on the PC versus the Mac make the comparison against the Gigabyte Aero 15X less accurate than we might wish; the two shouldn’t be directly compared. But the MacBook Pro 2017 quad-core is 11 percent faster than the 2018 system, despite the fact that the Kaby Lake CPU has fewer CPU cores. Lee’s tests show the reason readily enough — the MacBook Pro Core i9 is stuck at 2.2GHz, despite the advertised 2.9GHz minimum clock speed. Drop the system in the freezer, and the render time drops to 27 minutes, 18 seconds. That’s a roughly 31 percent improvement, and it corresponds exactly to the minimum specified clock the 2018 MacBook Pros should be holding in the first place.
Apple Insider first broke the story, but the site is unable to contemplate the idea that Apple is engaging in exactly the same type of bait-and-switch bullshit that has typified the PC industry’s approach to this topic for years. Instead, AppleInsider notes “it is highly unlikely that the company would ship a flagship product without first rigorously testing its performance.”
Of course, they’re quite right about that. There’s no chance at all that Apple shipped the MacBook Pro without knowing exactly about how it performed. This is the same company that chose to choke iPhone performance rather than fix its battery life, ship larger batteries, or design multi-core CPUs with more than two cores. It’s the same company that chose to ship iPhones with broken antenna configurations. It’s the same company that shipped iPhones with touch disease, after knowing they were more likely to bend. It’s the same company that shipped a 2015 keyboard with a failure rate 2x higher than its predecessor, failed to completely address that problem in a 2016 redesign, and just had to deploy a third redesign to fix an issue that never should’ve happened. It’s the same company that built an iPhone X with a glass back so expensive and difficult to replace, the cost of doing so is more than half the cost of the phone.
It is beyond ridiculous to look at Apple’s current round of product designs and say “Well gosh, who would’ve thought they’d pull a stunt like that?” Pulling “stunts like that” is what Apple does these days. And here it’s done nothing more than take a page from boutique manufacturers, which have been stuffing more hardware into mobile systems than those systems could handle for as long as I’ve been reviewing hardware.
In this case, it’s also possible that this behavior will impact the second-highest MacBook Pro SKU as well. The fact that the CPU clock is pulling down as far as 2.2GHz — well below the baseline 2.9GHz clock — suggests that the Core i7-8850H in the 15-inch refresh could also be impacted. That CPU has a 2.6GHz base clock, which is still higher than we know the Core i9 can maintain. This point is unconfirmed, however, and should be treated as supposition until someone with the laptop can check directly.
The only surprising thing about Apple’s behavior is that anyone, at this point, would expect anything different. And despite Apple Insider’s assertions to the contrary, there’s every reason to believe David Lee and his performance metrics. They line up, exactly, with what we’d expect to see in a scenario in which a six-core chip is put under all-core load in a thermally constrained versus thermally unconstrained environment.
And you watch — if Apple chooses to discuss this at all, it’ll be with remarks similar to the following: “Apple has designed the 2018 MBP to meet the most demanding needs of content creation professionals. The new six-core CPU provides better performance in many circumstances or scenarios.” As well it may. Just not, you know, when you actually try to use the entire CPU you paid for.
Intel’s Raja Koduri to Present at Samsung Foundry’s Upcoming Conference
Intel's Raja Koduri will speak at a Samsung foundry event this week — and that's not something that would happen if Intel didn't have something to say.
Samsung May Build $10B Foundry in Austin, Texas
Samsung may be planning a new $10B foundry in Austin Texas, with an aggressive plan to challenge TSMC.
Researchers Found Another Major Security Flaw in Intel CPUs
Security researchers have found another flaw in Intel CPUs — this time related to Intel Active Management Technology. Once again, this flaw can be leveraged to take complete control of a system, regardless of any security measures the user might employ.
GlobalFoundries to Soon Pilot 7nm and EUV Production
GlobalFoundries is pushing ahead with a plan to pilot EUV and 7nm production. We recently got a chance to sit down with the company and talk about its plans for the future.