Hence why I conditioned it:
> If Apple is going to renege their stupid decisions, now is a great time to find a place for Apple Silicon
Apple's B2C obsession isn't smart. Their service revenue is quite literally deemed illegal and exploitative in several courts, they cannot rely on the App Store and iCloud-style tie-ins forever. The progress that they've made on the A-series chips goes largely unnoticed to a cohort of users that scrolls TikTok and Polymarket. The progress of the M-series chips has not moved the needle on macOS' market share even with 7 years of sales marked down. Right now, they're lining themselves up for the worst of both worlds; being shackled to low-margin B2C products that aren't legally allowed to ammortize costs with lock-in "ecosystem" services. Apple will be forced to make changes if the status-quo continues to punish their business strategy.
So... why not lean into their strengths? iPhone customers will rarely notice fast single-core performance, but devops and enterprises will. Apple fundamentally supports expansion via PCI on Apple Silicon, and their SOCs demonstrably handle Linux, CUDA and Vulkan without any significant bottlenecks. The Mac Pro was 90% of the complete package for a particularly rich type of customer, and Apple refused to invest in the last 10% to uphold a pointless grudge. This stupid decision is why terrible ARM SOCs like Ampere and Grace have dominated the AI buildout even when Apple Silicon destroys them on-paper. Depreciating the Mac Pro is a solipsistic, insane decision that formally abandons Apple's attempts at dominating anything except the prosumer market. This isn't even an Xserve style slip-up, it's a radical steering decision that prevents customers like Pixar from even considering Apple Silicon for a server farm.
> As far as the Mac, it's the truck that does the heavy work
For now. Apple has laid the groundwork for what comes after, though: https://developer.apple.com/xcode-cloud/
> If Apple is going to renege their stupid decisions, now is a great time to find a place for Apple Silicon
Apple's B2C obsession isn't smart. Their service revenue is quite literally deemed illegal and exploitative in several courts, they cannot rely on the App Store and iCloud-style tie-ins forever. The progress that they've made on the A-series chips goes largely unnoticed to a cohort of users that scrolls TikTok and Polymarket. The progress of the M-series chips has not moved the needle on macOS' market share even with 7 years of sales marked down. Right now, they're lining themselves up for the worst of both worlds; being shackled to low-margin B2C products that aren't legally allowed to ammortize costs with lock-in "ecosystem" services. Apple will be forced to make changes if the status-quo continues to punish their business strategy.
So... why not lean into their strengths? iPhone customers will rarely notice fast single-core performance, but devops and enterprises will. Apple fundamentally supports expansion via PCI on Apple Silicon, and their SOCs demonstrably handle Linux, CUDA and Vulkan without any significant bottlenecks. The Mac Pro was 90% of the complete package for a particularly rich type of customer, and Apple refused to invest in the last 10% to uphold a pointless grudge. This stupid decision is why terrible ARM SOCs like Ampere and Grace have dominated the AI buildout even when Apple Silicon destroys them on-paper. Depreciating the Mac Pro is a solipsistic, insane decision that formally abandons Apple's attempts at dominating anything except the prosumer market. This isn't even an Xserve style slip-up, it's a radical steering decision that prevents customers like Pixar from even considering Apple Silicon for a server farm.
> As far as the Mac, it's the truck that does the heavy work
For now. Apple has laid the groundwork for what comes after, though: https://developer.apple.com/xcode-cloud/