2 weeks from now there will be another story saying it'll be out this summer...then 3 weeks later it'll be delayed until 2024. These people don't know shit!
Either the Mac Pro is dead, or so delayed that it’s become superfluous in their product lineup. Perhaps it’s become such a niche that any further investment simply is not a viability for Apple anymore. Either way, if Apple does not at least preview a Mac Pro option at WWDC, I think it’s dead Jim. Though $3000 ski goggles are considered the “next” thing - but I believe AR/VR is already past the public interest inflection point. WWDC is going to be a very interesting show.
Keep in mind how long Apple took to release the current Mac Pro.
And don't forget what the Mac Pro is. Apple's M chips as they currently exist are utterly inadequate to power a Mac Pro. They max out at 128GB RAM, and can't support any GPU option other than what's on-chip. The Mac Pro can use 1.5TB of RAM, and dual Radeon Pro GPUs. It's a FAR more powerful platform, and that's going to continue to be the case unless Apple actually develops a chip that can use off-chip RAM and non-Apple GPUs. Releasing something now and calling it a "Mac Pro" will just make Apple look silly, much like they did when they released that stupid trash can and called it a "Mac Pro" in 2013 - and at least that thing had expandable RAM (to 128GB, which was at least adequate 10 years ago) and dual GPUs that were theoretically upgradeable, and even though it turned out nobody released GPUs in that form factor you could at least add eGPUs.
Any "Mac Pro" Apple releases now with an Apple chip is going to be WORSE than the trash can. Their best bet if they have any intention of staying in the professional video market is to update the Intel Mac Pro, and that's going to be the case for the near and intermediate future.
I do not understand why Apple does not release the m3 MacBook Pro before the MacBook Air. It gives incentive to consumers to buy the very best.
What do you mean very best? What's the very best to you doesn't mean it's the very best for someone else just looking for an ultra-portable MacBook. Not everyone wants a more expensive heavier laptop with more ports as its overkill for their needs. Different people have different needs which is why MacBook Air exists in the first place.
I mean like the MacBook Pro 16 M3 Pro or the Max. Heck, I would love them to bring back the 17in MacBook Pro. However, I was reading the production yields have not been good with the 3nm process. You would think the would release something that sells less at a higher cost it would be a good starting point for production until they get better with 3nm process. The MacBook Air will definitely sell better no doubt as it is much cheaper. For the average consumer, it is a great laptop.
I do not understand why Apple does not release the m3 MacBook Pro before the MacBook Air. It gives incentive to consumers to buy the very best.
Economics and clever use of resources.
Also, the single core performance is the same, regardless of how many cores are used. Video editing rely on multicore processing to split several frames across the cores and that improves performance significantly. It's also same with Lightroom or similar products with exporting several images at once. Pros need to have many more cores in order to achieve faster completions.
I suspect they due process for pro type chips but given the higher rate of failure with more cores added, they just use the chips that meet the minimum requirements. It makes no sense to discard expensive chips just because one of the cores failed. And they carefully designed multicore chips for pro level chips.
So, it really makes sense to go with the entry level offering the latest architecture and let it scale up for pros.
While the single core performance will increase on entry level, they won't see the significant gains as they would hope for with previous generation of Pro chips on certain applications.
Either the Mac Pro is dead, or so delayed that it’s become superfluous in their product lineup. Perhaps it’s become such a niche that any further investment simply is not a viability for Apple anymore. Either way, if Apple does not at least preview a Mac Pro option at WWDC, I think it’s dead Jim. Though $3000 ski goggles are considered the “next” thing - but I believe AR/VR is already past the public interest inflection point. WWDC is going to be a very interesting show.
Keep in mind how long Apple took to release the current Mac Pro.
And don't forget what the Mac Pro is. Apple's M chips as they currently exist are utterly inadequate to power a Mac Pro. They max out at 128GB RAM, and can't support any GPU option other than what's on-chip. The Mac Pro can use 1.5TB of RAM, and dual Radeon Pro GPUs. It's a FAR more powerful platform, and that's going to continue to be the case unless Apple actually develops a chip that can use off-chip RAM and non-Apple GPUs. Releasing something now and calling it a "Mac Pro" will just make Apple look silly, much like they did when they released that stupid trash can and called it a "Mac Pro" in 2013 - and at least that thing had expandable RAM (to 128GB, which was at least adequate 10 years ago) and dual GPUs that were theoretically upgradeable, and even though it turned out nobody released GPUs in that form factor you could at least add eGPUs.
Any "Mac Pro" Apple releases now with an Apple chip is going to be WORSE than the trash can. Their best bet if they have any intention of staying in the professional video market is to update the Intel Mac Pro, and that's going to be the case for the near and intermediate future.
While the trash can was a disaster, it was more Apples non-commitment to upgrading the thing that was the failure. Honestly, the newest mac pro didn’t fare that much better. Has nothing to do with engineering or expandability. Had everything to do with management failure.
With Apple Silicon, it may SEEM like there is no way to expand, but that’s also a misnomer and using old tech thinking to look at future tech.
There are numerous ways to address compute expansion. First, there can be a combination of entire SOCs on one die, providing tons of cores and RAM at time of order - amd it would be insanely powerful - but this lacks true expandability and will still be somewhat limited in RAM.
Second, Apple could do as you suggest and turn the Mac Pro into a basic PC with a CPU, some RAM off to the side, and Nvidia/AMD GPUs off to the side. But while powerful, this approach leaves too much on the table in terms of efficiency and performance and goes against the ethos apple established long ago with its own SOCS, which squeeze every last ounce of performance from every watt of electricity.
Thirdly, Apple could basically make a large portion of the logic board into a system level “fabric” that allows for multiple SOCs to be added as needed. Each SOC would contain CPU/GPU/RAM and it’s own cooling. You need more RAM? You just so happen to get tons more CPU and GPU also. Need more GPU? You’re going to make sure that there is no memory bottleneck. So you can buy an ultra fast Mac Pro off the bat and upgrade it significantly down the line.
Furthermore, if Apple doesn’t like the idea of its SOCs going around with potential for undesirable use cases such as hack style computers, they can set up a system where the part is shipped to an Apple Store and it’s installed/soldered, etc. for you as a “free” service.
Finally, Apple can develop a desktop class SOC separate (though based on) the m series and allow much more breathing room with wattage, thermals, and clock speeds as well as greater RAM, etc. and add more as needed as well.
When it comes to Apple having more hardware control, I think we need to expand our thinking in terms of what’s possible and likely. Apple isn’t constrained by PC standards any longer. The worlds most creative and successful company is free to blow our minds. It’s ok. The Mac Pro is the ultimate in computing performance. It’s no wonder it’s taken time to get a fledgling (yet still voracious) architecture up to speed beyond mobile and small form factor desktops.
I might be in the vast minority here, but I’d love to see a rack mount Mac Pro.
Across the world Apple can probably sell 1/2 to 1 million units per quarter running Apple OS on a such a device, last quarter Apple sold about 4 million Mac’s. But we will never know because they won’t make it, at least not until those ex-Apple engineers and Qualcomm release something to the market in tandem with Microsoft?
"Finally, Apple can develop a desktop class SOC separate (though based on) the m series and allow much more breathing room with wattage, thermals, and clock speeds as well as greater RAM, etc. and add more as needed as well."
BlueLightning said: Like many home users, I use more CPU than GPU cycles.
There are also hardware video coders and decoders, as well as the 16 neural/AI cores that are used for video functions. Apple is mostly a consumer electronics company. Compared to my former late 2013 13" i5 2-core mbp (with no dedicated gpu cores), the 2023 14" M2 Pro mbp (10/16/16 cores) kicks ass. 4k video playback caused full fan speed on the 2013 i5 (and lagging video). Fans aren't even activated on the 2023 M2 Pro (and I see short parts of video that were skipped on the Intel chip).
Oh, I totally agree they are ahead and it is a better solution for a lot of average consumers. But, how many of even those people are going to run into GPU-required things that aren't going to work well or be possible? Gaming would be a great example.
The interesting thing about Apple Silicon, though, is that even the base machines are now plenty powerful for those people. Maybe people should just start buying a MBA or mini in near-base configuration, and then just add a mid-level-gaming PC with some good remote-control software. If Apple doesn't fix the GPU situation with the M3, that's probably what I'll end up doing. There is little compelling reason anymore for people who need GPU-power to buy the upper 80% of Apple's product lineup.
Marvin said: They said there was a Mac Pro coming, it sells below 1% of their units volume so it's not that important. It could launch with M3 but I think they'll announce an M2 model at WWDC. If it's an M2 Ultra with a larger GPU chip (2x M2 Ultra GPU cores = 56TFLOPs), it will do just fine. M3 version would be close to 100TFLOPs.
They really surprised us there, eh? M2 Studio in a bigger box with some slots. (sigh)
darkvader said: ... Releasing something now and calling it a "Mac Pro" will just make Apple look silly ...
Any "Mac Pro" Apple releases now with an Apple chip is going to be WORSE than the trash can. Their best bet if they have any intention of staying in the professional video market is to update the Intel Mac Pro, and that's going to be the case for the near and intermediate future.
Apparently, they didn't mind looking silly... and as far as I can see, the Intel Mac Pros aren't available any longer. My hunch is upper management just said, 'screw it, we're going to complete the transition, they'll have to deal with it.' It might be a few years until there is a real Mac Pro again.
Comments
Also, the single core performance is the same, regardless of how many cores are used. Video editing rely on multicore processing to split several frames across the cores and that improves performance significantly. It's also same with Lightroom or similar products with exporting several images at once. Pros need to have many more cores in order to achieve faster completions.
I suspect they due process for pro type chips but given the higher rate of failure with more cores added, they just use the chips that meet the minimum requirements. It makes no sense to discard expensive chips just because one of the cores failed. And they carefully designed multicore chips for pro level chips.
So, it really makes sense to go with the entry level offering the latest architecture and let it scale up for pros.
While the single core performance will increase on entry level, they won't see the significant gains as they would hope for with previous generation of Pro chips on certain applications.
https://www.apple.com/shop/buy-mac/mac-pro/rack#The interesting thing about Apple Silicon, though, is that even the base machines are now plenty powerful for those people. Maybe people should just start buying a MBA or mini in near-base configuration, and then just add a mid-level-gaming PC with some good remote-control software. If Apple doesn't fix the GPU situation with the M3, that's probably what I'll end up doing. There is little compelling reason anymore for people who need GPU-power to buy the upper 80% of Apple's product lineup.
They really surprised us there, eh? M2 Studio in a bigger box with some slots. (sigh)