Apple Developing 'Mac Studio,' Described as a Mac Mini and Mac Pro Hybrid

[Apple Knowledge Navigator]

Plot twist

The low-end (M1) Mac mini is replaced with a the model from the Prosser renders, rebranded as Mac Studio, and given BTO options up to M1 Max Duo. No doubt a config already with Pro and Max, too.

This would make a lot more sense, given the 'mini' nature is less of a novel feature today. 'Studio' implies that the small device can fit into a range of creative setups, and actually has both the CPU and GPU power to complete those tasks, unlike the high-end Intel mini.

 

[JippaLippa]

Plot twist

The low-end (M1) Mac mini is replaced with a the model from the Prosser renders, rebranded as Mac Studio, and given BTO options up to M1 Max Duo. No doubt a config already with Pro and Max, too.

This would make a lot more sense, given the 'mini' nature is less of a novel feature today. 'Studio' implies that the small device can fit into a range of creative setups, and actually has both the CPU and GPU power to complete those tasks, unlike the high-end Intel mini.

A mini with an M1 Max Duo would be almost too good to be true.
I'd snag it in a heartbeat.
The M1 Max is adequate for my workflow, but a DUO would be even better for futureproofing.
An M1 Max Quadra would be way too much power for my needs.

 

[theluggage]

Considering the SOC-approach with limited memory, this "Mac Studio" isn't going to be able to meet "THE most top end user's needs". Not without significant changes to the system architecture.

Most of the "most top end users" have likely already voted with their feet after 10 years of Apple dithering and pursuing dead ends (trashcan, iMac Pro, falling out with NVIDIA, new Xeon-W + AMD Mac Pro just before they announce that the future is Apple Silicon and iGPUs)... and moved to PC workstations that can be made-to-measure for their specialist needs and can use whatever GPUs their software needs.

The 2019 Mac Pro was firmly aimed at customers who were locked-in to a MacOS-based work flow which would be so expensive to change that they'd pay a huge premium for compatible hardware. Ever notice that all the speed comparisons on Apple's Mac Pro web page are with iMac Pro (2017) and Trashcans (2013)? No comparisons with PC - because at the end of the day it is just another Xeon PC in an over-engineered case with a neater way of powering GPUs. They're selling to a shrinking pool which has now had another few years to shrink - and the brief advantage in being an early adopter of the new Xeon-W has passed.

The whole 2006-2021 Intel Mac era worked well for Apple in many ways - but did limit how innovative they could be while still depending on the same "standard" parts from Intel, Nvidia (initially) and AMD that their competitors were using. The Trashcan is an example of what happened when they tried to push the envelope. Dropping the XServe and Mac OS Server was kinda inevitable once Apple themselves chose generic PC Server kit for their own data centres (the old PPC-based XServe started off by offering something that generic x86 hardware and Windows couldn't).

So, partly Apple Silicon allows Apple to once more produce distinctive personal computer hardware - and iPhone-funded R&D that resulted in the A-series chips probably gave them a massive head start. But Apple Silicon's party trick is high performance at low power, and is at its most impressive in mid-range laptops and SFF/all-in-one desktops. Which, honestly, are Apple's bread and butter.

While lower power consumption (& hence heat output/noise) is still "nice to have" it's not quite the killer feature that it is in laptops/SFF (thin & light, good battery life) or in high-density computing/server applications (data centres with huge, directly attributable electricity bills and requiring massive air conditioning systems).

Apple Silicon's other advantage is having as much as possible either on-die (GPU & accelerators) in-package (RAM) or closely connected (SSD). Now, that's tricky to scale - even if LPDDR5X lets you get 1TB of RAM on a Mx Max x4 package, not everybody is gong to want that much (expensive) RAM, so that means Apple making SoC packages for (probably) 128GB, 256GB, 512GB and 1TB... the higher of which are only going to sell to a handful of customers.

Now, Apple could invest a ton of money and come out with an Apple Silicon based Xeon-W-killer chip, but developing custom silicon just for that machine would be hugely expensive... and adding conventionally expandable RAM and PCIe GPU support erases some of the performance advantages of Apple Silicon.

So it's not surprising that all the rumours so far have been about ways of using multiple M1 Max dies (...or maybe the mythical 12 core iMac-level M1 Max+ dies...) - and I don't doubt that you could make an impressive system that way but it could be highly dependent on software and workflows being optimised to take full advantage of the unusual architecture.

...which is a problem when most of your current Mac Pro customers are only there because they are highly conservative about changing their workflow and may rely on obscure and expensive software.

A "Mac Studio" based on M1 Pro/Max tech and primarily pitched as a FCPx/Logic Pro "appliance" for small workgroups and one-man-bands could be a much better showcase for Apple Silicon than a quad-M1 Max pretending, and failing, to be a Xeon-W in a direct 2019 Mac Pro replacement.

 

[deconstruct60]

Actually you have not provided a single source linking to Apples statement of its Mac Pro sales. If you are going to quote them perhaps you could? All you did was post a link to Gruber who has been wrong about things before so isn't a purely trustworthy source IMO, he certainly won't have seen Apples actual sales figures.

Gruber was actually at the meeting. So it would be quite hard for him to get it 'wrong'.


"Phil Schiller:
... Third on the list is Mac Pro. Now, Mac Pro is actually a small percentage of our CPUs — just a single digit percent. However, we don’t look at it that way.

The way we look at it is that there is an ecosystem here that is related. So there might be a single digit percentage of pros who use a Mac Pro; there’s that 15 percent base that use Pro software frequently, and 30 percent who use it casually, and that these are related. These are not distinct little silos. There’s a connection between all of this.

..."
https://forums.macrumors.com/thread...a-mac-mini-and-mac-pro-hybrid.2336611/page-12

Phil Schiller was chief of marketing for Apple. So the notion that he doesn't know what the sales numbers accurately are is pretty close to zero. That is one of the principle jobs of doing Marketing: figuring out the size and scope of target markets and how well your products match up to those groups.

Lots of folks hand wave and infer than Single digit percent has to be some average between 1 and 10. That's is highly likely not true. 0.5% rounded up to 1% is 'single digit percent. Which is pretty much closer to the truth. Even back in the peak days of 2008-2009 timeframe Mac Pros were south of 250,000/yr run rate. In the 2013 and on era probably lower than 100K. ( once the iMac transition from mobile to desktop processors, it dropped among other factors over the last decade or so. )

iMac unit shipments are an order of magnitude bigger than the Mac Pro. It is a hobby product for Apple. One of the main reasons why the iMac Pro was such a much priority product for Apple in 2016-2017 than the Mac Pro ( which took two years after this meeting to ship. )

Why around 1% isn't "hand waving" ? Because there are public reports about who and how big the marketshare was for workstations. For example. 2010.


"..., Jon Peddie Research senior analyst Alex Herrera reports the industry shipped 903,700 workstations, a 6.4% sequential gain from the previous quarter.
...
HP and Dell spent several previous quarters jostling for market leadership, each taking a turn with slim volume leads, with neither seizing a sustainable advantage over the other. HP, which just five years ago trailed Dell by nearly 12 percentage points, had steadily closed the gap and over the last two quarters again managed to edge out its rival.

But this time, HP’s surge appears more vigorous, with the company in Q4 taking four points more volume share than the former number one (41.3% and 37.2% , respectively).
... "

Workstation market sets record in 2010 Q4

Jon Peddie Research analyst Alex Herrera says HP has retaken the lead among workstation vendors. The workstation market continues to recover from recessionary lows, setting a record for units shipped during the fourth quarter of 2010. Completing analysis of the workstation and professional...

gfxspeak.com

Dell and HP had 78% of the market. Lenovo/IBM was the historical large third place finisher. Apple was firmly in the "others" category. Apple's overall PC market share of 5% applied to that 900K/quarter run rate is about 45K per quarter. So a 4 quarters of that is around 180K. It is doubtful that Apple even had 5% in this market subsegment, so that is a high unit estimate. ( more expensive than average GPU cards. relatively limited drivers , etc. Plus the huge skew of folks buying with explicit, active support contracts(i.e., why HP/Dell/Lenovo so much better than mainstream marketshare averages ) )


Get to the 2013 and 2019 era probably not cracking 100K because either not "container expansion" or "priced thousands higher than a average workstation". ( average workstation being something like a mid-range HP/Dell. Z4 / 5000 not Z8 / 7000 ) .


P.S. The Peddie website appears to have pulled down their old archive of press release summaries about the numbers so I pulled some quotes from articles covering their press releases. And no, I didn't cherry pick some low 2010 one either.

" ... Q1’11 is now in the books, with the quarter seeing 860.0 thousand workstations shipped worldwide, representing a 4.8% sequential decline and an 18.6% year-over-year gain. Both of those figures need to be weighed in context ..."
https://www.businesswire.com/news/h...rket-Slows-to-a-More-Sustainable-Pace-in-Q111

900K/quarter is a quite healthy historical quarter.

At this point I have to ask myself why on earth am I arguing with you and your 'opinion' on Apples sales figures of its devices.
The devices are selling. Probably a lot more then you think in the case if the Mini. A new higher end Mini Pro or Studio will fill a gap very nicely.

If go back to the era where the Mini has a super low end mobile processor the Mac Pro used to outsell it. But since around 2014 (and the mini getting a very healthy "mac web services" role ) the Mini sells at much higher rates than the Mac Pro does. 2019 pricing only skewed that even higher. Very decent chance that Mac Pro's share percentage is being rounded up to get to 'single digit'. (and that is one contributing reason why Apple has a "low volume" tax on it to make it interesting enough for them to pursue as a product. ).

The M1 Mini having the same performance as the M1 iMac 24" has probably helped Mini sales. Apple's high tolerance for fratricide between the iMacs and the "headless" desktops appears to be higher since either way it is more M-series sales of the exact same silicon die (and RAM and NAND storage ) and it is a relatively cheaper display component at this point. ( not so sure they will have same tolerance for. mini-LED 27" iMac panels where they need to drive the unit sales up to get the costs down. )

 

[PinkyMacGodess]

In case it hasn't been said yet: Apple needs to come up with something between the Mac Pro and the iMac for people that are 'creatives', but can't afford the Mac Pro, but could use the power and expandability that model offers.

 

[Wxchaser]

Apple developing Mac N Cheese after Mac Studio

 

[cmaier]

In case it hasn't been said yet: Apple needs to come up with something between the Mac Pro and the iMac for people that are 'creatives', but can't afford the Mac Pro, but could use the power and expandability that model offers.

So you want the power and expandability of a Mac Pro, but you don’t want to pay for a Mac Pro. And if Apple was to introducer this hypothetical new computer, why would anyone buy a Mac Pro, given that it would be neither more powerful nor more expandable?

 

[theluggage]

Lots of folks hand wave and infer than Single digit percent has to be some average between 1 and 10. That's is highly likely not true. 0.5% rounded up to 1% is 'single digit percent.

Except describing 0.5% as "a single digit percent" on that justification would be deeply disingenuous. Anybody speaking honestly would describe that as "less than 1%" especially if the whole point they were making was that the sales were small, so there would be no point in exaggerating them.

Also, note the wording "our CPU sales". Not HP or Lenovo's CPU sales which we all know are far higher than Apple's across the board. If you look at total desktop/laptop sales, Apple are usually around 4th or 5th place but that's still only a fraction of Dell/HP sales - which include a metric shedload of basic corporate PCs for a market segment that Apple just doesn't participate in. So when you calculate "Number of pro workstations we sell/Number of computers we sell" for Apple about the only thing you know for sure is that the denominator is much smaller for Apple than for the wider PC market.

I'm not saying that you're necessarily wrong but I don't think there is any freely available evidence, short of Apple's own confidential sales figures, one way or another. All we have is what Schiller said - and either he meant that the sales were at least a few % or he was being dishonest, in which case his words mean nothing.

Anyway - when that statement was made, the only Mac Pro that had been available in the last 5 years was the Trashcan, and a major point of that meeting was Apple coming clean the trashcan didn't meet many Pro users needs... Which would have been an odd admission if it had been selling hand over fist... The other way you'll have low sales is if you charge more for the product - the trashcan was already about $500 more than the pre-2012 entry Mac Pro and Apple's solution to it was to double the price when they introduced the 2019 Mac Pro starting at $6000 (for a system spec that made zero sense until you added $10k of expansion). So yeah, if your only product in a market segment is a dud and/or too expensive then you're going to have low sales. Duh!

The problem is that the higher-end 2019 Mac Pro's claim to fame is the raw power of the high-end Xeon-W CPU and multiple AMD GPU options - it is ultimately just another Xeon-W PC workstation blessed to run MacOS (plus a bit of neat design which eliminates some ugly but perfectly functional power and interconnect cables). If Apple want to replace this with an Apple Silicon equivalent, they'll have to build the equivalent of a Xeon-W and top-end AMD GPU from the ground up.

 

[Realityck]

So you want the power and expandability of a Mac Pro, but you don’t want to pay for a Mac Pro. And if Apple was to introducer this hypothetical new computer, why would anyone buy a Mac Pro, given that it would be neither more powerful nor more expandable?

The current expandable Mac Pro would remain Intel based and not change for 2022 year, while this smaller M1 Max model would offer increased connectivity for someone looking for something better then Mac mini. The question of something like a AS based tower hasn’t been defined yet on what’s can be user expandable.

 

[deconstruct60]

Except describing 0.5% as "a single digit percent" on that justification would be deeply disingenuous. Anybody speaking honestly would describe that as "less than 1%" especially if the whole point they were making was that the sales were small, so there would be no point in exaggerating them.

Three things.

First, Apple isn't trying to give the specific number. It is against the corporate rules. So it is kind of ridiculous to jump to some hyper accurate reading here when they aren't even trying to give a specific number.

Second, Apple execs do weekly benchmarks on sales of the product line up. I think it is reported to be every Monday morning they look. Most likely that kind of reporting done to nearest whole number because just easier to review the dashboard more quickly. Pretty good chance that anything over zero is rounded up to 1%. Because if something registers to 0% then it quickly (or automatically) gets canned as a product. "Let's stop making these because 'nobody is buying them'. " ( roughly the commentary that Jobs used when the XServe got "Steve'd" / axed. If it is zero ... kill it is pretty likely the internal Apple "rules". It is rounded up because otherwise the convention is that it automatically gets killed. )

[ and likely why the Mac Pro struggles to get resources regularly assigned to it because that 1-2% is really in the zone of often being on the active discussion zone for "should we kill it ... are the trend lines going to hit zero in the future. " products. ]

Everything about the speed of product updates and level of effort that Apple puts into the Mac Pro substantially indicates that it is this low. That it doesn't show up on whole market share report cards only adds to that.

So if Apple says "less than one percent" that is effectively saying it also likely in the context of whther it is a 'coin toss' whether the product gets canned. There is no way Apple wants to convey tittering on the edge of killing it when they decided to take another shot at it. ( there is already a faction inside that saying kill it anyway. All they need to is chase away more folks and it becomes self fulfilling. )


Third, it is extremely unlikely that this is some constant single digit. In the initial demand bubble where the Mac Pro 2019 was new along side the MBP 16" 2019 their percentages over that initial quarter or two are likely higher than when the products got to steady state. The 2012 toward the end was probably quite low. The 2019 is likely somewhat lower now. If average over the long gaps in product refresh that can drop below 1% but Apple is happy enough with the spikes above 1% to 'count' that as the percentage. (there is no capture timeline given in Apple's reported percentage either. In 2017 at the tail end of a 4 year run for the Mac Pro it was a "mid-high" percentage ... you're smoking something. 4 year old products don't sell like "gang busters". Not even Apple ones. )

Also, note the wording "our CPU sales". Not HP or Lenovo's CPU sales which we all know are far higher than Apple's across the board. If you look at total desktop/laptop sales, Apple are usually around 4th or 5th place but that's still only a fraction of Dell/HP sales - which include a metric shedload of basic corporate PCs for a market segment that Apple just doesn't participate in.

That is because in the general market HP/Dell/Lenovo are in the 30-15% of the market range. No where near the high 30's and low 40's they are in the workstation market that they dominate even more. It is about 10% higher there. Which leaves an even smaller gap for Apple to carve out a subset segment on.

Apple's workstation line up is narrow. Even back the 2006-2010 days of the Mac Pro forums here there was a constrant drumbeat that the Mac Pro was. too big / too small and should have another model in the line up. ( wasn't as big as a Z8 or not as affordable and/or turbo, high clock-able as a Z4-like xMac ).

Apple has far more breath of offerings in the general market than they do in the workstation one.

So when you calculate "Number of pro workstations we sell/Number of computers we sell" for Apple about the only thing you know for sure is that the denominator is much smaller for Apple than for the wider PC market.

I'm not saying that you're necessarily wrong but I don't think there is any freely available evidence, short of Apple's own confidential sales figures, one way or another. All we have is what Schiller said - and either he meant that the sales were at least a few % or he was being dishonest, in which case his words mean nothing.

If going to be pedantic about what he said, then he did not say a few %. He said single digits; which is just one digit. You folks are trying to make each of the single digits equiprobable. They aren't. Even more so when weave in what other folks measure about the whole workstation market; move outside the pure vacuum of only Apple knows . Intel definitely knows to a very high degree how many CPUs Apple sold. AMD effectively knows over the last decade too because the GPU count lines up with the system count ( only GPU type on the box). Major high end mac wholesalers/resellers know how many Mac Pro's they sell relative to other non workstation Macs. It would not be too hard for a dedicated market analyst to sort all of this out.

Anyway - when that statement was made, the only Mac Pro that had been available in the last 5 years was the Trashcan, and a major point of that meeting was Apple coming clean the trashcan didn't meet many Pro users needs... Which would have been an odd admission if it had been selling hand over fist... The other way you'll have low sales is if you charge more for the product - the trashcan was already about $500 more than the pre-2012 entry Mac Pro and Apple's solution to it was to double the price when they introduced the 2019 Mac Pro starting at $6000 (for a system spec that made zero sense until you added $10k of expansion). So yeah, if your only product in a market segment is a dud and/or too expensive then you're going to have low sales. Duh!

Apple was still not on the Peddie Workstation top 3 list even back in the 2007-2011 era. Pointing at the MP 2013 and 2019 isn't the primary issue why Apple's was low in workstation market.


As iMac and upper end MBP became more capable there was a general shift from Mac Pro to those other systems. Schiller mentions this also in the same meeting. Apple saw folks moving off of Mac Pros over time. ( not really surprising as same general overall industry forces involved in moving folks from desktops to more laptops over time).

Apple exacerbated it with the doubling of the Mac Pro entry price in 2019, but the general trend was already there in 2017 (and earlier).

 

[theluggage]

If going to be pedantic about what he said, then he did not say a few %. He said single digits; which is just one digit.

So it is kind of ridiculous to jump to some hyper accurate reading here when they aren't even trying to give a specific number.

Seriously? You're trying to insist that "single digit" means "less than one" because 0.5 rounded up is "1" which has a single digit... and you're accusing others of being pedantic?

"Single digit" could plausibly mean "less than 10" or "a few" or "about 5". It doesn't mean "less than one" unless the speaker is deliberately obtuse. Meanwhile, there's simply no freely published information to show how many Mac Pros are actually sold.

 

[richinaus]

Three things.

First, Apple isn't trying to give the specific number. It is against the corporate rules. So it is kind of ridiculous to jump to some hyper accurate reading here when they aren't even trying to give a specific number.

Second, Apple execs do weekly benchmarks on sales of the product line up. I think it is reported to be every Monday morning they look. Most likely that kind of reporting done to nearest whole number because just easier to review the dashboard more quickly. Pretty good chance that anything over zero is rounded up to 1%. Because if something registers to 0% then it quickly (or automatically) gets canned as a product. "Let's stop making these because 'nobody is buying them'. " ( roughly the commentary that Jobs used when the XServe got "Steve'd" / axed. If it is zero ... kill it is pretty likely the internal Apple "rules". It is rounded up because otherwise the convention is that it automatically gets killed. )

[ and likely why the Mac Pro struggles to get resources regularly assigned to it because that 1-2% is really in the zone of often being on the active discussion zone for "should we kill it ... are the trend lines going to hit zero in the future. " products. ]

Everything about the speed of product updates and level of effort that Apple puts into the Mac Pro substantially indicates that it is this low. That it doesn't show up on whole market share report cards only adds to that.

So if Apple says "less than one percent" that is effectively saying it also likely in the context of whther it is a 'coin toss' whether the product gets canned. There is no way Apple wants to convey tittering on the edge of killing it when they decided to take another shot at it. ( there is already a faction inside that saying kill it anyway. All they need to is chase away more folks and it becomes self fulfilling. )


Third, it is extremely unlikely that this is some constant single digit. In the initial demand bubble where the Mac Pro 2019 was new along side the MBP 16" 2019 their percentages over that initial quarter or two are likely higher than when the products got to steady state. The 2012 toward the end was probably quite low. The 2019 is likely somewhat lower now. If average over the long gaps in product refresh that can drop below 1% but Apple is happy enough with the spikes above 1% to 'count' that as the percentage. (there is no capture timeline given in Apple's reported percentage either. In 2017 at the tail end of a 4 year run for the Mac Pro it was a "mid-high" percentage ... you're smoking something. 4 year old products don't sell like "gang busters". Not even Apple ones. )




That is because in the general market HP/Dell/Lenovo are in the 30-15% of the market range. No where near the high 30's and low 40's they are in the workstation market that they dominate even more. It is about 10% higher there. Which leaves an even smaller gap for Apple to carve out a subset segment on.

Apple's workstation line up is narrow. Even back the 2006-2010 days of the Mac Pro forums here there was a constrant drumbeat that the Mac Pro was. too big / too small and should have another model in the line up. ( wasn't as big as a Z8 or not as affordable and/or turbo, high clock-able as a Z4-like xMac ).

Apple has far more breath of offerings in the general market than they do in the workstation one.




If going to be pedantic about what he said, then he did not say a few %. He said single digits; which is just one digit. You folks are trying to make each of the single digits equiprobable. They aren't. Even more so when weave in what other folks measure about the whole workstation market; move outside the pure vacuum of only Apple knows . Intel definitely knows to a very high degree how many CPUs Apple sold. AMD effectively knows over the last decade too because the GPU count lines up with the system count ( only GPU type on the box). Major high end mac wholesalers/resellers know how many Mac Pro's they sell relative to other non workstation Macs. It would not be too hard for a dedicated market analyst to sort all of this out.





Apple was still not on the Peddie Workstation top 3 list even back in the 2007-2011 era. Pointing at the MP 2013 and 2019 isn't the primary issue why Apple's was low in workstation market.


As iMac and upper end MBP became more capable there was a general shift from Mac Pro to those other systems. Schiller mentions this also in the same meeting. Apple saw folks moving off of Mac Pros over time. ( not really surprising as same general overall industry forces involved in moving folks from desktops to more laptops over time).

Apple exacerbated it with the doubling of the Mac Pro entry price in 2019, but the general trend was already there in 2017 (and earlier).

Honestly I dont know why they bother except for vanity. I always felt the Mac Pro was a computer to say, look how good we can make a PC desktop. Unfortunately for studios like mine, it is too good and doesn't make financial sense [we are not burdened to Mac only apps].

Anecdotally, I was in the Apple Store discussing the Mac Pro with a genius, and was telling him about my needs and the PC specs for the studio. He agreed that the PC was better for half the price, and the Mac Pro was only really good for Final Cut Pro users.

This Mac Studio, if delivered correctly, that is fast, responsive, quiet and cool, will be great for so many people, and pricing should be in a range of affordability for a smaller professional creative studio [unlike the Mac Pro].

 

[jjcs]

If you look at Apple’s line up, it is about price categories, and luring you into a higher categorie than you initially want. If Apple knows one thing, it’s pricing. Capabilities are in function of pricing, not the other way around. If you analyze it, the overlaps are deliberate.

Good point. Especially considering the whole soldered RAM and now SSDs.

 

[jjcs]

Most of the "most top end users" have likely already voted with their feet after 10 years of Apple dithering and pursuing dead ends (trashcan, iMac Pro, falling out with NVIDIA, new Xeon-W + AMD Mac Pro just before they announce that the future is Apple Silicon and iGPUs)... and moved to PC workstations that can be made-to-measure for their specialist needs and can use whatever GPUs their software needs.

Indeed. That trend isn't likely to turn around, either. They're already almost entirely consumer/prosumer more than anything else. It's a definite switch from the marketing in the early OS X era.

...which is a problem when most of your current Mac Pro customers are only there because they are highly conservative about changing their workflow and may rely on obscure and expensive software.

Yep. That's a dwindling market, too.

A "Mac Studio" based on M1 Pro/Max tech and primarily pitched as a FCPx/Logic Pro "appliance" for small workgroups and one-man-bands could be a much better showcase for Apple Silicon than a quad-M1 Max pretending, and failing, to be a Xeon-W in a direct 2019 Mac Pro replacement.

You may be on to the fact that they're going to drop the concept of a "Mac Pro" entirely without saying so.

 

[Manzanito]

A Mac Studio + Studio 7k display could be a winning combination. Maybe $5k for the pair…

How would you price both items? I’m guessing the computer will be around 2.000$, maybe even less (base config, assuming it will be a mbp sans the screen). Three grand for the monitor is a little bit too much for my liking. Still, this is apple.

 

[MichaelDroste]

Just a thought wouldn’t it be cool if they kept the Intel machines and then made cards with M1 max pro on them? I don’t even know if that would work but I hate to junk my three year old intel tower….

 

[Boil]

Indeed. That trend isn't likely to turn around, either. They're already almost entirely consumer/prosumer more than anything else. It's a definite switch from the marketing in the early OS X era.

Pro market is a tough one, especially with the whole Nvidia/CUDA/OPTIX thing; maybe Apple is taking a more prosumer stance, looking towards satisfying the FCP/Logic/Photoshop folks more than the ultra high-end AI/ML/etc. folks...?

You may be on to the fact that they're going to drop the concept of a "Mac Pro" entirely without saying so.

Apple may change the narrative on what is "Pro", catering more towards those who need ready-to-go desktop-friendly closed SFF systems for the one-man-band/small boutique studio end users...?

The 2019 Intel Mac Pro can be configured to an astounding US$55k price; just the tower, no displays...

RAM = $25k, GPUs = $12k, & CPU = $8k...!!!

That gives one a 28-core CPU, dual W6900X MPX GPUs, 1.5TB RAM, 8TB SSD, & an Afterburner card...

Obviously even a quad M1 Max SoC configuration (even with LPDDR5X RAM) cannot match that, overall performance-wise, but it can beat it on pricing...

Maximum spec (using LPDDR5X RAM) quad M1 Max SoC should be about US$25k; again, just the machine, no displays...

That would give one a 40-core CPU (32P/8E), 128-core GPU, 64-core Neural Engine, 1TB LPDDR5X RAM, & 8TB SSD...

 

[jjcs]

Apple wmay change the narrative on what is "Pro", catering more towards those who need ready-to-go desktop-friendly closed SFF systems for the one-man-band / small boutique studio end users...?

They already more or less do that in their marketing regarding what they consider a "pro" user. It's just like you describe. Small scale image, video, and audio editing. More "prosumer" than anything.

 

[Pezimak]

Gruber was actually at the meeting. So it would be quite hard for him to get it 'wrong'.


"Phil Schiller:
... Third on the list is Mac Pro. Now, Mac Pro is actually a small percentage of our CPUs — just a single digit percent. However, we don’t look at it that way.

The way we look at it is that there is an ecosystem here that is related. So there might be a single digit percentage of pros who use a Mac Pro; there’s that 15 percent base that use Pro software frequently, and 30 percent who use it casually, and that these are related. These are not distinct little silos. There’s a connection between all of this.

..."
https://forums.macrumors.com/thread...a-mac-mini-and-mac-pro-hybrid.2336611/page-12

Phil Schiller was chief of marketing for Apple. So the notion that he doesn't know what the sales numbers accurately are is pretty close to zero. That is one of the principle jobs of doing Marketing: figuring out the size and scope of target markets and how well your products match up to those groups.

Lots of folks hand wave and infer than Single digit percent has to be some average between 1 and 10. That's is highly likely not true. 0.5% rounded up to 1% is 'single digit percent. Which is pretty much closer to the truth. Even back in the peak days of 2008-2009 timeframe Mac Pros were south of 250,000/yr run rate. In the 2013 and on era probably lower than 100K. ( once the iMac transition from mobile to desktop processors, it dropped among other factors over the last decade or so. )

iMac unit shipments are an order of magnitude bigger than the Mac Pro. It is a hobby product for Apple. One of the main reasons why the iMac Pro was such a much priority product for Apple in 2016-2017 than the Mac Pro ( which took two years after this meeting to ship. )

Why around 1% isn't "hand waving" ? Because there are public reports about who and how big the marketshare was for workstations. For example. 2010.


"..., Jon Peddie Research senior analyst Alex Herrera reports the industry shipped 903,700 workstations, a 6.4% sequential gain from the previous quarter.
...
HP and Dell spent several previous quarters jostling for market leadership, each taking a turn with slim volume leads, with neither seizing a sustainable advantage over the other. HP, which just five years ago trailed Dell by nearly 12 percentage points, had steadily closed the gap and over the last two quarters again managed to edge out its rival.

But this time, HP’s surge appears more vigorous, with the company in Q4 taking four points more volume share than the former number one (41.3% and 37.2% , respectively).
... "

Workstation market sets record in 2010 Q4

Jon Peddie Research analyst Alex Herrera says HP has retaken the lead among workstation vendors. The workstation market continues to recover from recessionary lows, setting a record for units shipped during the fourth quarter of 2010. Completing analysis of the workstation and professional...

gfxspeak.com

Dell and HP had 78% of the market. Lenovo/IBM was the historical large third place finisher. Apple was firmly in the "others" category. Apple's overall PC market share of 5% applied to that 900K/quarter run rate is about 45K per quarter. So a 4 quarters of that is around 180K. It is doubtful that Apple even had 5% in this market subsegment, so that is a high unit estimate. ( more expensive than average GPU cards. relatively limited drivers , etc. Plus the huge skew of folks buying with explicit, active support contracts(i.e., why HP/Dell/Lenovo so much better than mainstream marketshare averages ) )


Get to the 2013 and 2019 era probably not cracking 100K because either not "container expansion" or "priced thousands higher than a average workstation". ( average workstation being something like a mid-range HP/Dell. Z4 / 5000 not Z8 / 7000 ) .


P.S. The Peddie website appears to have pulled down their old archive of press release summaries about the numbers so I pulled some quotes from articles covering their press releases. And no, I didn't cherry pick some low 2010 one either.

" ... Q1’11 is now in the books, with the quarter seeing 860.0 thousand workstations shipped worldwide, representing a 4.8% sequential decline and an 18.6% year-over-year gain. Both of those figures need to be weighed in context ..."
https://www.businesswire.com/news/h...rket-Slows-to-a-More-Sustainable-Pace-in-Q111

900K/quarter is a quite healthy historical quarter.




If go back to the era where the Mini has a super low end mobile processor the Mac Pro used to outsell it. But since around 2014 (and the mini getting a very healthy "mac web services" role ) the Mini sells at much higher rates than the Mac Pro does. 2019 pricing only skewed that even higher. Very decent chance that Mac Pro's share percentage is being rounded up to get to 'single digit'. (and that is one contributing reason why Apple has a "low volume" tax on it to make it interesting enough for them to pursue as a product. ).

The M1 Mini having the same performance as the M1 iMac 24" has probably helped Mini sales. Apple's high tolerance for fratricide between the iMacs and the "headless" desktops appears to be higher since either way it is more M-series sales of the exact same silicon die (and RAM and NAND storage ) and it is a relatively cheaper display component at this point. ( not so sure they will have same tolerance for. mini-LED 27" iMac panels where they need to drive the unit sales up to get the costs down. )

Yoir first sentence is flawed, Phil Schiller stated single digits, you then wrote lots of text to seemingly try and proclaim that means 0.5% rounded up to 1%? That’s complete conjecture and personal opinion no matter how much text you type.

Either way I’m looking forward to what they announce on Tuesday with a Mac Studio or Mac Mini with Pro and Max chip options.

 

[Boil]

...looking forward to what they announce on Tuesday with a Mac Studio or Mac Mini with Pro and Max chip options.

Right there with you...!

A M1 Max Mac mini might allow more GPU cores & RAM than an equally-priced Mac Studio; but if the Mac Studio is a Cube 2.0, then I am all over that...! ;^p

 

[richinaus]

Right there with you...!

A M1 Max Mac mini might allow more GPU cores & RAM than an equally-priced Mac Studio; but if the Mac Studio is a Cube 2.0, then I am all over that...! ;^p

M1 Max duo in a cube

I am expecting a mini though but taller, which is fine by me. As long as we get something 50% more powerful than the MBP and quiet for the desk I will be super happy.

 

[kenaustus]

When I think of "Studio" I think of working all day on music or video in a classic "studio". Might be in a basement or own a real studio environment but the new Mac will probably attract the professionals that Apple wants to attract.

My guess is starting in the mid $2,000 range (without feet) with the more powerful version starting higher

 

[richinaus]

When I think of "Studio" I think of working all day on music or video in a classic "studio". Might be in a basement or own a real studio environment but the new Mac will probably attract the professionals that Apple wants to attract.

My guess is starting in the mid $2,000 range (without feet) with the more powerful version starting higher

We are a design studio . it is a creative space where we work in 3D space and objects.

 

[deconstruct60]

Honestly I dont know why they bother except for vanity. I always felt the Mac Pro was a computer to say, look how good we can make a PC desktop. Unfortunately for studios like mine, it is too good and doesn't make financial sense [we are not burdened to Mac only apps].

Between Mac Pro and MPX module pricing I would be utterly shocked to find out that Apple was loosing money on these products. It isn't 'move the needle' money to substantively change overall yearly Mac revenues levels. But it likely enough to smooth out the revenue hiccups with varying rollout demand bubbles for the rest of the line up that updates more often.

If the pocket a decent amount of profit during the idle years then that can help pay for a 18-24 month burst of activity on the product every 3-6 years without even really digging a big hole is resource cost outlay. ( a Xmas layaway savings fund for the once in a blue moon hobby project).

Apple uses them (that isn't solely enough to create the product, but does show they have utility). If you look at the M1 series introduction videos in the "secret apple lab" where they have workbenches you can see about 40 rack mount models on the wall in the background.

For example: about 16:10-16:20 in this video:

[ maybe running some of that Windows/Linux only chip design software that doesn't run on macOS. ]


So it isn't just vanity (unless those are completely non representative 'movie' sets) . A bit of "eat your own dog food" there. However, more profits than vanity driven. Profits is part of reason take long breaks also ( milk like a cash cow)

Anecdotally, I was in the Apple Store discussing the Mac Pro with a genius, and was telling him about my needs and the PC specs for the studio. He agreed that the PC was better for half the price, and the Mac Pro was only really good for Final Cut Pro users.

Apple also. The rack density of those Mac Pro units is way, way ,way less dense than a Mini. If the "Studio" is a three high Mini ( around 4") then would need some modified shelves brackets could dense pack probably get a 3x density improvement if tilted up on side like Mini's a commonly racked. ( ~17-18" w MP rack versus 5" for Studio (with some spacing between to decouple the case heating). )

It is a little sad though that Apple themselves is drifting deeper into that " we are only good at Final Cut , Logic , and some backward looking Adobe apps ... so let us double down on just those. " mindset. Somewhat likely a path to eventually painting themselves into another corner via groupthink.

This Mac Studio, if delivered correctly, that is fast, responsive, quiet and cool, will be great for so many people, and pricing should be in a range of affordability for a smaller professional creative studio [unlike the Mac Pro].

Folks who have higher than average working set RAM footprints are going to have problems. Apple's RAM is way off what can do with open market , commodity RAM. For folks with "same purchase order" constraints and have to buy from system vendor it is not so far off , but shops which do their own subassembly/subconfiguration are still going to grumble.

However, yes for folks look for a box to do work with, far more tinder with ... it probably will sell quite well. Enough so that Apple probably going to cost control what they "speed bump upgrades" they do to the Intel Mac Pro put a cap on how much they have to spend to keep it moving forward a while longer.

 

[Boil]

Apple uses them (that isn't solely enough to create the product, but does show they have utility). If you look at the M1 series introduction videos in the "secret apple lab" where they have workbenches you can see about 40 rack mount models on the wall in the background.

That is actually where they store the consciousness of Steve Jobs...!

Assuming max config for each unit:

• 1120 CPU cores
• 6400 GPU cores
• 40 Afterburner cards
• 60TB RAM
• 320TB SSD
• 80 10Gb Ethernet ports

It is a little sad though that Apple themselves is drifting deeper into that " we are only good at Final Cut , Logic , and some backward looking Adobe apps ... so let us double down on just those. " mindset. Somewhat likely a path to eventually painting themselves into another corner via groupthink.

I want to see Apple put together their own 3D/DCC software suite, expand the circle of software designed to work closely with Apple silicon hardware (looking at you, Mac Studio)...?

Maybe Apple even wraps all their DCC software (FCP, Logic, Phenomenon DCC Suite, etc.) into a bundle available with these new Mac Studio machines at a significantly discounted rate, this would (as another poster said previously somewhere on the forums here) serve to draw more folks to try Apple hardware & DCC software...?

Folks who have higher than average working set RAM footprints are going to have problems. Apple's RAM is way off what can do with open market , commodity RAM. For folks with "same purchase order" constraints and have to buy from system vendor it is not so far off , but shops which do their own subassembly/subconfiguration are still going to grumble.

Reducing the "Apple Tax" on RAM (as well as SSDs) would help folks buy into higher configured machines, allowing folks to see more of what these tightly integrated machines can do, maybe drive more sales...?

However, yes for folks look for a box to do work with, far more tinder with ... it probably will sell quite well. Enough so that Apple probably going to cost control what they "speed bump upgrades" they do to the Intel Mac Pro put a cap on how much they have to spend to keep it moving forward a while longer.

Tinkering with my 3900X build was fun at first, but then I just wanted a stable out-of-the-box experience, not to spend a bunch of time coaxing the last little bit of performance out of various overclocks (CPU/GPU/RAM)...