Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Fingers crossed that it's merely a case of all the component ducks being lined up in a row. They'll want to launch with latest Xeon+latest AMD graphics, and not risk a repeat of the Power Mac G4 'Yikes' situation, or launching with one set of components, then doing an update two or three months later because the ones they wanted were finally available. After all this time, it's got to be right, or they might as well not have bothered.
 
I imagine we'll hear something about the processor front early April since Intel has its Data Innovation Day on the 2nd (at least if it's gonna use SP. Who the hell knows when -W news will really break.)
 
Well, the good thing is Apple have only got the obligatory iPhones and the new Mac Pro left on its hardware list for the year, and it’s only March.
So I expect endless iPhone leaks and rumours and possibly a Mac Pro leak or two?
Will it be launched this year?
 
Last edited:
  • Like
Reactions: filmak
Fingers crossed that it's merely a case of all the component ducks being lined up in a row. They'll want to launch with latest Xeon+latest AMD graphics, and not risk a repeat of the Power Mac G4 'Yikes' situation, or launching with one set of components, then doing an update two or three months later because the ones they wanted were finally available. After all this time, it's got to be right, or they might as well not have bothered.
Meanwhile, Apple's core pro audience is buying Z-series with the parts that are available now - parts that are far better than anything in the MP6,1 or the cMPs.

In tech - "you snooze, you lose". Apple's been snoozing for a long time. Waiting six months for some component that's marginally better is a losing game. Release now, and in six months do a silent update.
 
We’re only days away from the 2 year anniversary when Apple admitted “they were in a thermal corner” and were “working on a modular Mac Pro” what have you done in the last two years of your lives.
 
  • Like
Reactions: ssgbryan
Meanwhile, Apple's core pro audience is buying Z-series with the parts that are available now - parts that are far better than anything in the MP6,1 or the cMPs.

In tech - "you snooze, you lose". Apple's been snoozing for a long time. Waiting six months for some component that's marginally better is a losing game. Release now, and in six months do a silent update.
You're not wrong...but they're clearly not thinking like that, and never have....
 
Freudian lapsus? o_O
IMG_ACCF6E467FB4-1.jpeg
 
  • Like
Reactions: apolloa
You're not wrong...but they're clearly not thinking like that, and never have....

Macs were somewhat competitive during the GHz wars, stepped up their game with the G5s, and where there or thereabouts with the Intel MPs until 2010 or so .
Apart from available/usable GPUs, which were not quite as important back then .

When you look at Apple's history, neglecting Macs and ignoring user needs is a fairly recent thing .
 
  • Like
Reactions: ssgbryan
I imagine we'll hear something about the processor front early April since Intel has its Data Innovation Day on the 2nd (at least if it's gonna use SP. Who the hell knows when -W news will really break.)

NAB show is also in about a week. Last year that show started on the April 7th and get the "Pro update" techcrunch interview posted on April 5th. So shouldn't be some kind of freak out if April 2nd goes by and nothing.

The Cascade SP on April 2nd is more formal than breaking news. Intel was demoing stuff a week ago...

"... Intel, was showing off a super-dense quad-socket machine based on Cascade Lake as well as future machines with two, four, and eight sockets based on the follow-on “Cooper Lake” processors expected later this year. ...'
https://www.nextplatform.com/2019/03/26/a-first-peek-at-cascade-lake-xeons-ahead-of-launch/

I suspect that the slow drip , drip, info about the -W variants will start leaking out of Intel presentations after they have exhausted the Nov-April dripline on the SP versions. it should start more earnestly next week.
 
NAB show is also in about a week. Last year that show started on the April 7th and get the "Pro update" techcrunch interview posted on April 5th. So shouldn't be some kind of freak out if April 2nd goes by and nothing.

The Cascade SP on April 2nd is more formal than breaking news. Intel was demoing stuff a week ago...

"... Intel, was showing off a super-dense quad-socket machine based on Cascade Lake as well as future machines with two, four, and eight sockets based on the follow-on “Cooper Lake” processors expected later this year. ...'
https://www.nextplatform.com/2019/03/26/a-first-peek-at-cascade-lake-xeons-ahead-of-launch/

I suspect that the slow drip , drip, info about the -W variants will start leaking out of Intel presentations after they have exhausted the Nov-April dripline on the SP versions. it should start more earnestly next week.
Intel needs quad socket to keep up with AMD??
 
Intel needs quad socket to keep up with AMD??

Xeon SP had quad socket support all along. Xeon SP subsumed both the E5 2600 and E7 line up.
"Glueless" 4 sockets is supported directly by the associated chipsets. Glueless 8 sockets I think needs a bit of firmware magic dusk and some OS support.

Two years ago, 2017....
"... There is support for up to 8 sockets, and most of the Platinum SKUs have 24 to 28 cores, and is clearly using the XCC silicon. ..."

https://www.anandtech.com/show/11544/intel-skylake-ep-vs-amd-epyc-7000-cpu-battle-of-the-decade/7

[ Which isn't 'new' for SP processors. E7 did 8 ( with a little bit more tap dancing )
"
  • Max CPU Configuration 8
"
https://ark.intel.com/content/www/u...-processor-e7-8891-v4-60m-cache-2-80-ghz.html

]
What the OpenSystems boards are doing ( like the article notes that several of the IBM power boards have been doing for years ) allows for the two sockets board to be scaled up by lashing them together in increasing NUMA effect. With four sockets one socket can talk to all other three directly ( one hop ). Once get to 8 sockets some ( some of the others a 2+ hops away). For system vendors who wanted to do their own "glue" that could be scaled further. ( most don't anymore because it doesn't make economic sense given likely number of systems to be sold. ).

But Intel doesn't particularly need quad sockets. If it is "core count" in a single package thats what their AP variants are going to do. They won't exactly match core count, but will probably be close enough for most folks. They'll put two dies in a single package. Then just two of that mega sockets will pragmatically give four dies worth. That isn't what these demos were doing. Eight of those sockets won't scale (unless that is one of the major changes coming with Cooper Lake).

[ Earlier material looked like Cooper Lake would be a socket transition in prep for Ice Lake but that doesn't seem to be the case. Going to be interesting what else is in Cooper Lake besides the bfloat thing. ]

Apple using some of these OpenSystems design reference boards in their own data centers would be surprising. A small hardware team designing custom "lego" modules for their data centers. [ most other large cloud service vendors are doing the same thing... that is what that conference was about. ] Doesn't mean it is an retail end user product though. ( or that it even boots macOS. Probably won't. )
 
In tech - "you snooze, you lose". Apple's been snoozing for a long time. Waiting six months for some component that's marginally better is a losing game. Release now, and in six months do a silent update.

Pros switched to Windows long ago, at this point Apple needs to release a product that will convince pros to come back.
 
  • Like
Reactions: Ursadorable
Pros switched to Windows long ago, at this point Apple needs to release a product that will convince pros to come back.
A nice upgrade that's a "one trick pony" won't bring the pros back. Updating every five to seven years, while throwing out real expandability, is a losing proposition.

The MP7,1 is unlikely to bring the pros back. If there's an MP8,1 in 2020, and an MP9,1 in 2021 - there might be some migration back to Apple.

If these are systems based on A-series ARM SOCs, forget about it.
 
A nice upgrade that's a "one trick pony" won't bring the pros back. Updating every five to seven years, while throwing out real expandability, is a losing proposition.

The MP7,1 is unlikely to bring the pros back. If there's an MP8,1 in 2020, and an MP9,1 in 2021 - there might be some migration back to Apple.

If these are systems based on A-series ARM SOCs, forget about it.

Aiden NOBODY ever said or implied the MacPro was gonna get Arm processors.
 
Aiden NOBODY ever said or implied the MacPro was gonna get Arm processors.
Heh, with how much wild speculation and what-ifs have been thrown out here, I wouldn't be too sure.

Macs were somewhat competitive during the GHz wars, stepped up their game with the G5s, and where there or thereabouts with the Intel MPs until 2010 or so .
Apart from available/usable GPUs, which were not quite as important back then .

When you look at Apple's history, neglecting Macs and ignoring user needs is a fairly recent thing .

Eh.... The trend of Apple's pro computers has been pretty uniform since the G3 era. You had cutting-edge processors that started falling badly behind Intel as time went on (the dual processor G4s were absolutely smoked by Intel's offerings until the G5 came out and then we had the failure to reach 3GHz...) The Mac Pros were so amazing because they were substantially more powerful than the G5s they replaced and far cooler, quieter, and energy-efficient, not to mention cost-competitive. The sheer length of time languishing with a self-admitted dead end is the only novel thing about the current situation.
 
  • Like
Reactions: barmann
You can find posters in this very thread making exactly that claim.
If the rumors are true that Apple is switching to Arm Macs in 2020 then they are going to have to move the Mac Pro to Arm at some point. The Mac Pro coming out this year will run on Intel Xeons but it's not out of the realm of possibility that Apple may be secretly working on a supped-up Arm chip intended to be used in a future Mac Pro. With Intel on the third tock of their former tick-tock strategy and still no 7nm process in sight the time may come that Apple's Arm chip becomes fast enough to replace Intel. Why wouldn't Pros switch to Arm? The software will move over and the people married to x86 will simply switch to Windows. Mac users will make the switch just as they did from 68k to PPC and PPC to x86.
 
If the rumors are true that Apple is switching to Arm Macs in 2020 then they are going to have to move the Mac Pro to Arm at some point. The Mac Pro coming out this year will run on Intel Xeons but it's not out of the realm of possibility that Apple may be secretly working on a supped-up Arm chip intended to be used in a future Mac Pro. With Intel on the third tock of their former tick-tock strategy and still no 7nm process in sight the time may come that Apple's Arm chip becomes fast enough to replace Intel. Why wouldn't Pros switch to Arm? The software will move over and the people married to x86 will simply switch to Windows. Mac users will make the switch just as they did from 68k to PPC and PPC to x86.

Users including Pros switched to x86 because they were able to have a choice of OS on the same machine and run VM’s of Windows whilst in OSX. And in my mind it meant more programmes as it must be easier to port a programmes from one OS to another if the hardware is the same?

But at this point I see moving to ARM as taking all that away, plus I’ve read comments about how ARM is not as good at certain tasks as x86 is. But from day one or even two would you be able to run OSX and Windows on the same machine, or in a VM in the other OS?

And for consumers it’ll be a question of here’s a Mac with the same processor as an iPad, and with the new iPad Pro you can plug it to a monitor, so then it becomes a decision if you need a mouse and OSX? This will o think confuse decisions there.

We will see but I’m not convinced currently that ARM will work half as well as X86 does in a ‘pro’ multi thousand dollar machine.
 
  • Like
Reactions: barmann
Users including Pros switched to x86 because they were able to have a choice of OS on the same machine and run VM’s of Windows whilst in OSX. And in my mind it meant more programmes as it must be easier to port a programmes from one OS to another if the hardware is the same?

But at this point I see moving to ARM as taking all that away, plus I’ve read comments about how ARM is not as good at certain tasks as x86 is. But from day one or even two would you be able to run OSX and Windows on the same machine, or in a VM in the other OS?

And for consumers it’ll be a question of here’s a Mac with the same processor as an iPad, and with the new iPad Pro you can plug it to a monitor, so then it becomes a decision if you need a mouse and OSX? This will o think confuse decisions there.

We will see but I’m not convinced currently that ARM will work half as well as X86 does in a ‘pro’ multi thousand dollar machine.

I think that most of the consumers do not care or have an interest about the iPad's CPU, so I don't think that there will be any confusion.
Anyway, the desktop ARM should be way more powerful than the iPad's one, so there will be a clear distance between them, necessary for marketing reasons too.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.