I initially thought this could occur, but I really think that Apple will hold the price points. Maybe, we could see Apple knock $100 off the intel equivalent.
I suspect that Apple will use the margin to increase the content of the devices, like better screens, better cameras, 5G chips, etc. Premium prices with premium features.
Agree. In general, Apple doesn’t compete on price.
I think they will do both (like what happened with iPad) by expanding the market without many users moving to lower models. The people who buy the best to earn a living (like developers and creatives) isn't going to change. If they hit a lower price point they will get more enterprise use, casual users, and students. They can do Thunderbolt and micro-LED on the high end and LCD / iPad chips / slower storage on the low end. The lower end of the market could be much bigger than what Apple currently has. I think Apple has been unwilling to go in to the lower end in the past because they would need to sacrifice too much. I think we are at the point where Apple could easily do a $800 cost of entry. Maybe even down to $500 if they keep older chips in circulation like with the entry level iPad. If Apple can hit this point and run iOS software, they may be able to compete with ChromeBooks since dual ecosystem and better hardware can make up for the rest of the price difference. Apple already owns the top end of the market. Any future growth will need to make Mac more accessible to more people.
The performance + battery improvements alone are enough to fundamentally change Apple's unit market share in the PC industry. This analyst seems to be missing that fundamental point. A lower priced SKU variant is just the icing on the cake.
It'll be generations behind AMD in performance and won't gain much now that AMD will own Xilinx. Both will be introducing FPGAs and more DSPs to their CPU designs. AMD has the most advanced CPU designs in the world.
Prices will be the same but their profit margins will increase due to being the CPU designer.
I see you’re still butt-hurt that your favourite chip house didn’t figure into Apple’s plan.
You haven’t seen ASi, and you haven’t seen AMD’s new whatever, so you don’t know which is faster and which runs cooler.
You also don’t how much it cost to develop the new chips, or do you think Apple’s chip designers work for free?
Apple's expenses are huge. I see no reason Apple can or should lower prices.
The cost of creating a wholly new hardware system is exorbitant. The cost of the transition is exorbitant. The cost of enhancing and supporting two different platforms is exorbitant.
There must be roadmap to more features down the line is exorbitant -- and trying not to leave too far behind Intel platforms with its millions of users.
Licensing fees and manufacturing costs mentioned elsewhere are just rounding errors.
The performance + battery improvements alone are enough to fundamentally change Apple's unit market share in the PC industry. This analyst seems to be missing that fundamental point. A lower priced SKU variant is just the icing on the cake.
It'll be generations behind AMD in performance and won't gain much now that AMD will own Xilinx. Both will be introducing FPGAs and more DSPs to their CPU designs. AMD has the most advanced CPU designs in the world.
Prices will be the same but their profit margins will increase due to being the CPU designer.
I see you’re still butt-hurt that your favourite chip house didn’t figure into Apple’s plan.
You haven’t seen ASi, and you haven’t seen AMD’s new whatever, so you don’t know which is faster and which runs cooler.
You also don’t how much it cost to develop the new chips, or do you think Apple’s chip designers work for free?
Right. There is a huge penalty to keeping the x86 architecture. Before factoring in energy savings, ASi outperforms Intel and AMD by 4-5x on some operations because they are not stuck with legacy decisions. ASi has a better memory model for concurrency which can boost multi-threaded code. The ASi memory model is also much more secure than Intel/AMD which is obviously important to Apple. Same on the GPU side where Apple is using TBDR and Intel / AMD / NVIDIA are still using pseudo immediate mode rendering for compatibility reasons. Apple has much more control over how their developers build software to help make this kind of transition.
There will also be savings from getting rid of 3rd party GPUs!
Unless they don’t get rid of discrete GPU in the machines that have them already.
Rumors point to Apple releasing their own discrete GPU. There are probably some savings from making their own GPU, but Apple will probably price at a premium. Of course they will have savings from omitting the GPU on lower end machines that will still have discrete level of performance comparing to existing discrete chips. The integrated GPU is rumored to be competitive with a Radeon 5500 based on iPad performance. I'm guessing Apple's discrete chip will be at least twice as fast as a Radeon 5500, but we'll see. I assume ASi chips with more cores will require the discrete GPU.
I doubt this will happen. Sure the cost of build might godown... but apple’s pricing? Nah. At the least they will maintain prevailing price points for the time being.
The performance + battery improvements alone are enough to fundamentally change Apple's unit market share in the PC industry. This analyst seems to be missing that fundamental point. A lower priced SKU variant is just the icing on the cake.
It'll be generations behind AMD in performance and won't gain much now that AMD will own Xilinx. Both will be introducing FPGAs and more DSPs to their CPU designs. AMD has the most advanced CPU designs in the world.
Prices will be the same but their profit margins will increase due to being the CPU designer.
The performance + battery improvements alone are enough to fundamentally change Apple's unit market share in the PC industry. This analyst seems to be missing that fundamental point. A lower priced SKU variant is just the icing on the cake.
It'll be generations behind AMD in performance and won't gain much now that AMD will own Xilinx. Both will be introducing FPGAs and more DSPs to their CPU designs. AMD has the most advanced CPU designs in the world.
Prices will be the same but their profit margins will increase due to being the CPU designer.
I see you’re still butt-hurt that your favourite chip house didn’t figure into Apple’s plan.
You haven’t seen ASi, and you haven’t seen AMD’s new whatever, so you don’t know which is faster and which runs cooler.
You also don’t how much it cost to develop the new chips, or do you think Apple’s chip designers work for free?
AMD's newest CPU is the Ryzen 5000 series based on their newest Zen 3 core, so technically yes we have. Other sites have tested and reviewed it so we know how well and how cool it performs. And it's pretty darn good. But you are right, we haven't Apple Silicon Mac SoC's, so there's no basis for comparison yet.
In the US, school districts buy mostly Chromebooks (60% share) with bad screens and keyboards that fall apart in a year. Apple has a tiny, tiny slice of that education market now. This might be an opportunity for Apple to release a bare-bones iBook with a cheaper screen, older A12Z, and 1 each USB-A and USB-C ports, sufficient for Zooming and thumb drives. priced around $500. I wouldn't imagine such a device cannibalizing Apple's existing laptop lines.
The performance + battery improvements alone are enough to fundamentally change Apple's unit market share in the PC industry. This analyst seems to be missing that fundamental point. A lower priced SKU variant is just the icing on the cake.
It'll be generations behind AMD in performance and won't gain much now that AMD will own Xilinx. Both will be introducing FPGAs and more DSPs to their CPU designs. AMD has the most advanced CPU designs in the world.
Prices will be the same but their profit margins will increase due to being the CPU designer.
What advantage will AMD get by owning Xilinx?
I'm not going to go down the entire list of Xilinx's product line, but suffice it to say anything Apple has in development like Afterburner [FPGA], or their custom AI Chips [FPGA/DSPs] and Neural Architecture [Machine Learning FPGAs], etc., Xilinx has in production in Data Centers now.
In the US, school districts buy mostly Chromebooks (60% share) with bad screens and keyboards that fall apart in a year. Apple has a tiny, tiny slice of that education market now. This might be an opportunity for Apple to release a bare-bones iBook with a cheaper screen, older A12Z, and 1 each USB-A and USB-C ports, sufficient for Zooming and thumb drives. priced around $500. I wouldn't imagine such a device cannibalizing Apple's existing laptop lines.
They could also release it as an education only model. There is a lot of precedent for this among Macs.
Price: I do think we'll see a lower base price point, but the specs of the system will be such that it will be very common, almost necessary, to pay more to bump them up. Something like a MacBook Air with 8GB RAM, 128GB SSD starting at $899 or similar Mac mini for $599.
Cost: The cost to design and develop these new Mac SoCs is not as much as some people seem to think. These chips will share IP blocks (CPU, GPU, ISP, ANE, DSP, etc) with their A-series SoCs. Any cost will be in designing "custom" blocks specifically for Macs, like say bigger GPU cores.
Discreet GPU: Not going to happen. The GPUs on any MacBook SoC will be just as fast as anything offered by a mobile AMD GPU. In place of a discreet GPU in "pro" laptops, Apple could choose to include an "afterburner" (FPGA) type option.
When talking about higher end desktops, they could continue to offer discreet AMD options. Think of Apple's custom GPUs as a replacement for the built-in Intel GPUs, but much, much faster.
I doubt they will lower the entry-level pricing, but the entry-level models will provide significantly more processing power. Thus, they are "cheaper" in that to get a competitive Intel machine, you would have to spend hundreds of dollars more on the processor, and possibly get a much larger and heavier system. For example, let's say a new Mac chip computes as fast as a 45W i7 processor that goes in the 16" MBP, but with a TDP which can be managed by a MacBook Air's cooler. The MacBook Air would still start at the same price, but would be cheaper than a 16" MBP with that i7.
With this interpretation, everybody is right. Apple keeps their premium market position, they keep their margins where they like them, and a given level of performance gets a lot cheaper. They don't "lower prices" in that you still have to pay $1k to get a MacBook Air at all, but they do "lower prices" in that it competes with the $2400 MacBook Pro.
This is just throwing out numbers. It's not hugely likely the base config of the new MacBook Air will actually compete with the base config of the existing 16" MacBook Pro. A new Air will almost certainly beat the existing 13" Pro, and a new 13" Pro will almost certainly beat the existing 16" Pro.
Of course Apple could come out with cheaper Macs. The Apple TV could be reconfigured to run Big Sur. $200 is expensive end of TV boxes, though I think it is worth the money. $200 is cheap for a Mac, yet they can have the same hardware. Change the software and use iCloud storage or external drives. Early Apple TV’s were hacked to enable them to run Mac OS GUI. Current Apple TV’s are much more powerful than early versions. For a good deal of people this would be good enough.
I’d like to see a fan- and headless Mac the size of an AppleTV. Physical ports: power, RJ45, HDMI and a couple of Thunderbolts. Wireless:Wifi, IR and Bluetooth.
I’d stick it in my hifi rack as Roon server and for various video streaming services (with or without VPN). 8 GB RAM, 256 GB SSD and USD 400 or less.
Price: I do think we'll see a lower base price point, but the specs of the system will be such that it will be very common, almost necessary, to pay more to bump them up. Something like a MacBook Air with 8GB RAM, 128GB SSD starting at $899 or similar Mac mini for $599.
Cost: The cost to design and develop these new Mac SoCs is not as much as some people seem to think. These chips will share IP blocks (CPU, GPU, ISP, ANE, DSP, etc) with their A-series SoCs. Any cost will be in designing "custom" blocks specifically for Macs, like say bigger GPU cores.
Discreet GPU: Not going to happen. The GPUs on any MacBook SoC will be just as fast as anything offered by a mobile AMD GPU. In place of a discreet GPU in "pro" laptops, Apple could choose to include an "afterburner" (FPGA) type option.
When talking about higher end desktops, they could continue to offer discreet AMD options. Think of Apple's custom GPUs as a replacement for the built-in Intel GPUs, but much, much faster.
There are credible rumors in the supply chain that Apple is in the final stages of producing a discreet GPU. It is unclear if it will look like a traditional discreet GPU though. I assume this will be to offload silicon to a companion chip for devices with more cores. No rumors on if the discreet chip will have a different memory architecture. Maybe Apple will tell us more today.
Right on Intel. The Intel GPU design is memory bandwidth starved. Embedded GPUs (or any GPU that needs to reduce power use) would work better with a TBDR design, but it seems Intel is unwilling to look past the short term. TBDR can be up to 4 times more efficient with memory bandwidth. With Microsoft continuing to improve support for TBDR through Qualcomm chips, this is probably going to look really short sighted soon as all ultra-portables shift to ARM.
Comments
The cost of creating a wholly new hardware system is exorbitant. The cost of the transition is exorbitant. The cost of enhancing and supporting two different platforms is exorbitant.
There must be roadmap to more features down the line is exorbitant -- and trying not to leave too far behind Intel platforms with its millions of users.
Licensing fees and manufacturing costs mentioned elsewhere are just rounding errors.
The last time I was actively planning a MBP purchase was for the late 2016 refresh.
I was ultimately put off by many factors but price killed it for me.
Since then I got by magnificently on older but upgraded machines but lower priced options are a good place to start if costs can be reduced.
Whether Apple will pass any possible gains onto the consumer though is a different story.
Just visit https://www.xilinx.com and research their solutions and IP.
I do think we'll see a lower base price point, but the specs of the system will be such that it will be very common, almost necessary, to pay more to bump them up. Something like a MacBook Air with 8GB RAM, 128GB SSD starting at $899 or similar Mac mini for $599.
Cost:
The cost to design and develop these new Mac SoCs is not as much as some people seem to think. These chips will share IP blocks (CPU, GPU, ISP, ANE, DSP, etc) with their A-series SoCs. Any cost will be in designing "custom" blocks specifically for Macs, like say bigger GPU cores.
Discreet GPU:
Not going to happen. The GPUs on any MacBook SoC will be just as fast as anything offered by a mobile AMD GPU. In place of a discreet GPU in "pro" laptops, Apple could choose to include an "afterburner" (FPGA) type option.
When talking about higher end desktops, they could continue to offer discreet AMD options. Think of Apple's custom GPUs as a replacement for the built-in Intel GPUs, but much, much faster.
With this interpretation, everybody is right. Apple keeps their premium market position, they keep their margins where they like them, and a given level of performance gets a lot cheaper. They don't "lower prices" in that you still have to pay $1k to get a MacBook Air at all, but they do "lower prices" in that it competes with the $2400 MacBook Pro.
This is just throwing out numbers. It's not hugely likely the base config of the new MacBook Air will actually compete with the base config of the existing 16" MacBook Pro. A new Air will almost certainly beat the existing 13" Pro, and a new 13" Pro will almost certainly beat the existing 16" Pro.
Wireless:Wifi, IR and Bluetooth.
8 GB RAM, 256 GB SSD and USD 400 or less.
Right on Intel. The Intel GPU design is memory bandwidth starved. Embedded GPUs (or any GPU that needs to reduce power use) would work better with a TBDR design, but it seems Intel is unwilling to look past the short term. TBDR can be up to 4 times more efficient with memory bandwidth. With Microsoft continuing to improve support for TBDR through Qualcomm chips, this is probably going to look really short sighted soon as all ultra-portables shift to ARM.