Originally posted by mdedetrich
View Post
One thing to note is that the new M2 Macbook's with 256GB storage is half the speed of the original M1's. So you're losing some value even compared to older M1 devices. Unless you get a 512GB model, which you're paying $200 more for another 256GB of storage. Storage which you can't upgrade or replace. Part of the reason I think Apple did this is because the expense that goes into making the M series chips. Compared to AMD and Intel chips, the M series is a good 4x to 5x larger in transistor count, and it's all on one single die. What that means is that these chips are expensive to manufacture, especially on 3nm. The defect rate must be high, as the bigger the chip the more likely a defect will occur. This is why AMD went chiplet in order to improve their yields to lower cost and improve performance, something Apple isn't doing. What this means is that Apple has to charge a higher price and offer less in storage in order to maintain profits, compared to going with Intel chips. By the end of the year AMD will have their Zen4 chips running on 5nm, but their laptop versions will be out in 2023 running on 4nm. Intel will be using 3nm next year as well, though for what I don't know. Whatever the case is, Apple won't have the power advantage anymore by next year. AMD and Intel's x86 will have caught up in power efficiency, and will be a lot cheaper as well. Technically AMD and Intel chips are already a lot cheap but don't have the power advantage, though I would argue that really depends on what you do with these laptops as the M series is only really good at video editing and single threaded applications in terms of power efficiency.
Remember that AMD, Intel, and Nvidia have a much larger market share to spend R&D on, and have years in engineering their products. Apple though has basically been taking ARMv8 and made some tweaks and "borrowed" Imaginations GPU tech to make their SoC's. All for what is essentially 10% market share. Eventually Apple will have to shift gears and either pay Nvidia or Qualcomm to make them ARM SoC's and even maybe go AMD for discrete GPU's, or go right back to Intel and x86. We Linux users have a hard time getting Windows applications working on x86, so imagine what Apple M1 users have to deal with on Mac OSX, with ARM, without functional OpenGL or Vulkan. Unless you have a very specific reason to use an Apple M based product like video editing, you're going to have a bad time.
Leave a comment: