The problem with ARM architecture for applications other than servers, netbooks and phones / tables, is their IPC rate still si far behind not current x86 (Zen 3, 10th gen core), is behind AMD bulldozer, so it's single Thread performance in common applications it's like an underpowered Celeron, not competitive for applications impossible to rewrite for SMMP logic (as gcal geometric library), so Apple may launch today an ARM Mac but likely an entry level product not a mainstream/pro cuz is a huge step back.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MP 7,1 Waiting for Mac Pro 7,1 A1991 (no more)
- Thread starter Mago
- Start date
-
- Tags
- macpro 2016
- Sort by reaction score
Well, I'll be out doing Show and Tell with a couple classes of fourth graders (giving them hope that they too can make a career as a hack graphic designer / illustrator) so I'll miss the real time excitement / despair on this thread.
I'll check in later. Keep it classy, everyone, and set your comments to "Stunned"!
I'll check in later. Keep it classy, everyone, and set your comments to "Stunned"!
https://www.anandtech.com/show/14384/arm-announces-cortexa77-cpu-ip
This curve is not flattening for Intel actually Intel had gain near 50% since
as long as we're throwing out last minute fantasy wishlists...
The Mac Pro will be a system built with ESXi-ish hypervising as the native way it operates - it's the first ARM Mac, but all the ARM system (an iOS-looking derivative) does is manage / abstract the hardware, in an extension of the T(x)'s role managing storage in the iMac Pro, and present virtual computers to the guest operating systems that all actual work by users on the machine is done in. There are no "Native" apps for it.
It's like an Apple-TV for Hypervising.
It's no longer a standard Intel or AMD "PC" - you can't boot Windows, macOS, or Linux on it directly, any more than you can on an iPad, but it uses standard parts (storage, RAM, Processors, GPUs etc) connected together on a proprietary bus / interconnect - think of it like a modern BeBox. It still needs a Xeon (or Threadripper) to supply to the VMs to run guest operating systems, but you can run multiple instances so app sets are fully compartmentalised / manageable, you can switch between instances, or have other operating systems (plug in an Nvidia GPU, pipe it to your Windows guest OS, plug in a Radeon, pipe it to the macOS guest). CPUs & ram are on daughter cards, so processor makers changing sockets don't matter any more, you just plug in the daughter card with the cpu in its socket.
*hooray, staying up till 3am waiting for the keynote to start...*
The Mac Pro will be a system built with ESXi-ish hypervising as the native way it operates - it's the first ARM Mac, but all the ARM system (an iOS-looking derivative) does is manage / abstract the hardware, in an extension of the T(x)'s role managing storage in the iMac Pro, and present virtual computers to the guest operating systems that all actual work by users on the machine is done in. There are no "Native" apps for it.
It's like an Apple-TV for Hypervising.
It's no longer a standard Intel or AMD "PC" - you can't boot Windows, macOS, or Linux on it directly, any more than you can on an iPad, but it uses standard parts (storage, RAM, Processors, GPUs etc) connected together on a proprietary bus / interconnect - think of it like a modern BeBox. It still needs a Xeon (or Threadripper) to supply to the VMs to run guest operating systems, but you can run multiple instances so app sets are fully compartmentalised / manageable, you can switch between instances, or have other operating systems (plug in an Nvidia GPU, pipe it to your Windows guest OS, plug in a Radeon, pipe it to the macOS guest). CPUs & ram are on daughter cards, so processor makers changing sockets don't matter any more, you just plug in the daughter card with the cpu in its socket.
*hooray, staying up till 3am waiting for the keynote to start...*
It's possible. My post above was mainly aimed at the technical aspects of a theoretical Intel to ARM transition. It's simply not technically feasible for Apple to just develop "a Rosetta solution" to the problems mentioned. If they do choose to migrate to ARM it will come at the expense of Boot Camp and performant Virtualization. That's just the cold reality of the situation. And, of course, Apple could make that decision despite of the downsides.
Looking at the business strategy and political side of things, I'm not sure your thinking is aligned with the realities of Apple's current precarious reputation in the pro marketplace. They may choose to drop support for Boot Camp. They may decide that supporting virtualization in macOS (in the form of VMware, Parallels, Docker for Mac, etc) is no longer important to the platform. Those are choices they can make.
All of those regressions put some amount of friction and market resistance on customers embracing the new modular Mac Pro in a landscape where suspicions and ill sentiment are already about as high as one can imagine. It would be a very risky choice for them to make. The eyes of the industry are on the new Mac Pro. In many ways it's a bellweather for the future viability of macOS in general. There aren't many safe paths forward for Apple, and there are many possible missteps that could cause even more damage to macOS and the Apple ecosystem. Especially when considering the Mac Pro or iMac Pro.
Any discussion of an ARM migration has to account for this reality.
Is Apple's position strong enough right now that a loss of Boot Camp and virtualization would "motivate customers to request proper ports" or is it weak enough that a loss of Boot Camp and virtualization would motivate customers to migrate to Windows or Linux? That's the crux of the issue.
Last edited:
Because the market for this product aren’t going to be interested in jumping on the product with no native software out of the gate, and probably have software that will behave more temperamentally/the non-native performance is unacceptable.
Personally I’m excited about the transition but no way I’d want to jump on the first gen product especially for my business.
I got to play with BeOS a little and really liked it. To bad it didn't make it. But if you go to https://www.haiku-os.org/ you can find what BeOS has become.That's a blast from the past...!
Won't be "hilarious", or won't be "real"?
conservative base clock isn't really a IPC constraint. From article referenced above.
"...
and Arm is proclaiming a similar 3GHz peak target frequency as its predecessor. Naturally since frequency isn’t projected to change much, this means that the core’s targeted +20% performance boost can be solely attributed to the IP’s microarchitectural changes.
To achieve the IPC (Instructions per clock) gains, Arm has reworked the microarchitecture and introduced clever new features, generally beefing up the CPU IP to what results in a wider and more performant design. ..."
There are a number of things that the x86 implementors are already doing that the ARM folks are just getting to with expanded transistor budgets. Apple is ahead of where A77 is getting to. However, a major reason for that is that they are not side-tracked with chasing other implementations of ARM aimed at completely different areas than iOS devices.
Can Apple manage to walk and chew gum at the same time is more operative question then whether general scope of all ARM architecture is somehow hamstrung or not.
[doublepost=1559575079][/doublepost]
Why would people care about something they don't have ( and aren't buying)? The number of folks who own an iOS device who do not have a Mac is far, far, far bigger than those who own both.
The iPod didn't completely take-off until it was decoupled from being essentially "mac only". The iPhone never was hard coupled that way. Neither was the iPad.
Where the users overlap Apple does put some focus on making the synergy better. But "Macs first and everything is has to be second place" that horse left the barn a long time ago.
The core problem is not whether iPods and now iOS got bigger than Macs. The real core issue is Apple's limited ability to walk and chew gum at the same time. The zero-sum game that they spread too wide over products ( even within just the Mac product line up. ). They don't have to do a 180 turn and shoot for "too many" products, but the scoping is off on what is "as simple as possible, but no simpler".
Last edited:
Apple’s ARM architecture is so different I’d be very careful comparing against reference designs. Benchmarks I’ve seen show the A series absolutely demolishing the reference chips.
But I don’t think we’ll see an ARM Mac today.
Apple is more time shifted. It is also which benchmarks. Anandtech had to modify Spec to fit within the user app memory constraints that Apple cuddles the A-series under.
Directly shipping? No. If Apple 'blows up' stuff that will be harder to port well, they may show something "between the lines" .
Time to overanalyse stuff again for the sake of nothing: Haven't there been alot of silence on Apple specific sites for a while now regarding the Mac Pro? Both Macrumors and 9to5mac for example usually runs several Mac Pro rumors and gossip everytime a Mac Pro is speculated to make a wild appearance at WWDC or any other event, for a while now i've only seen the Mac Pro mentioned in "what to expect at WWDC"-articles, not standalone Mac Pro ones.
Has tech blogs already gotten their hands on the nMP or gotten a preview thus signed an NDA/embargo.
Or, the more reasonable explanation: they simply don't care anymore
Has tech blogs already gotten their hands on the nMP or gotten a preview thus signed an NDA/embargo.
Or, the more reasonable explanation: they simply don't care anymore
Alright. This is almost it.
With mainstream tech media reporting quite a bit about a likely Mac Pro reveal, and Apple not controlling expectations by leaking a denial... we'll hear something about it today. Show and tell. Likely release this fall?
Release isn't that important to me, but I have to know what Apple is thinking. What their way forward is.
Reason for lack of rumors is American manufacture—with a different parts and manufacturing chain.
With mainstream tech media reporting quite a bit about a likely Mac Pro reveal, and Apple not controlling expectations by leaking a denial... we'll hear something about it today. Show and tell. Likely release this fall?
Release isn't that important to me, but I have to know what Apple is thinking. What their way forward is.
Reason for lack of rumors is American manufacture—with a different parts and manufacturing chain.
Let's hope that the vwMP title won't come true.
I do no think that presenting a normal Mac system, at least what it used to be called normal for some decades, should take so long. The knowledge is here and available, there are not any crucial technologies missing.
Anyway, the most irritating part is that most of people are asking for a usual kind of a workstation, an easier task, not a fancy ultra custom design, a lot more difficult task.
I do no think that presenting a normal Mac system, at least what it used to be called normal for some decades, should take so long. The knowledge is here and available, there are not any crucial technologies missing.
Anyway, the most irritating part is that most of people are asking for a usual kind of a workstation, an easier task, not a fancy ultra custom design, a lot more difficult task.
Well, I don't doubt it will be able to sustain their performance without thermal and power limit. ARM's problem is the lack of native applications, not its performance.