MP 7,1 Is the new Mac Pro too niche?

[hans1972]

Why would they eat into iMac / iMac Pro sales? Perhaps because there is a demand for such an offering?

Yes, there is some demand on the PC side, but I doubt the demand on the Mac side is very high for towers in the $2500-$4000 range taking into consideration the price difference to PCs, to support Apple's 30-50% margin.

Many of those people who wants to buy such a desktop Mac wants to give as little money to Apple as possible and then upgrade using 3rd party hardware.
They want to keep the machine as long as possible.
They would demand support for NVIDIA and CUDA.
They complain a lot and are never happy.
Therefore they are the worst kind of customers for Apple.

In addition, it would require more support and testing of 3rd party drivers since so much different hardware could be placed in such a machine.

I am pretty sure that the average yearly revenue and profit from customers buying high-end MacBook Pros and iMac (Pros) are much higher, and providing a good and cheap alternative would eat into the revenue and profit from the other models.

 

[Flaming Gerbil]

Filmaker's view on Mac Pro.
[doublepost=1560725997][/doublepost]
[doublepost=1560726188][/doublepost]Simple solution to the Apple / Nvidia feud:

Nvidia market cap = 88B
Apple market cap = 880B
​

Need I say more?

 

[ct2k7]

Aha!

I wonder, though, how long licensing models will work "per computer"?

Bear with me, as I outline a similar situation in database servers.

I use IBM Cloud (I'm a rebel, LOL) for backend for a mobile app. And I use PostgreSQL for relational database needs. I noticed an announcement of a new service on IBM Cloud running PostgreSQL on IBM LinuxOne Z hardware (rather than X86). (For those that might be interested, it's called HyperProtect DBAAS, and is currently beta testing at no cost).

Not being familiar with Z/LinuxOne server, (well, I have SOME familiarity, as I programmed in 360 assembler many decades ago - the Z-series is the latest in a line of upgrades traced back to the 360, and will actually run 360 machine code...) I downloaded a monograph explaining the advantages of the hardware platform. It was meant for companies considering purchase of mainframe hardware, but I found it useful to understand the advantages of running PostgreSQL on this hardware "in the cloud" as well.

As I expected, of course, it touts that the hardware was designed with security in mind. Built-in high performance encryption hardware, and of particular interest, largely not susceptible to the kinds of side-channel attacks that have been in the news almost constantly lately regarding Intel processors.

Now, that's a big plus for a database server. As well, relational database software doesn't scale "out" well. (That is, to multiple servers.) It best scales "up". (To faster processors, more cores on the same chip/in the same box, massive amounts of RAM etc.) It's a workload well-suited to "big iron" like this.

(Getting to the point, patience!)

Here's what I did not expect.... they cite lower total cost of ownership. OK, I think, I can see how using one big mainframe might be financially advantageous over using a boatload of commodity X86 servers, after considering maintenance cost (replacing failed servers for one - these mainframe servers have a MTBF of 50 years...), floor space, cost of power, etc.

And, indeed, they made a case for lower TCO on this basis. This was reassuring to me, as I've been speculating at the (unannounced, as of yet) pricing of the new cloud service. This makes me think it might not be much more costly than the current service I use that runs on X86 hardware. i.e. IBM - eating their own dogfood - would be expected to have a lower TCO running database servers on Z hardware vs X86.

However, there was an unexpected major component (actually THE major component) of cost savings...

Software licensing.

It turns out that the way database software licensing (I'm talking the likes of Oracle, here, obviously not PostgreSQL - which is open-source) works, there is a HUGE savings from running licensed database software on One Ginormous Box rather than on a bunch of "big" X86 boxes.

You've pointed out that a similar licensing situation exists with audio production software - and I presume - as well with video production software.

How long, though, before they change their licensing models? "Per computer" licensing boggles my mind... it's so simplistic, and I presume widely "gamed" this way. Is it that if they priced the license based on some metric of the processing power of the hardware, it would make IT executives and purchasing agents heads to explode?

Anyway, it sounds like by offering the closest thing to "big iron" that Apple can (currently) offer (I STILL want them to go to IBM Z chips - I know I am in a tiny .00001 minority, LOL) the savings in licensing make the cost of the hardware - which seems so shocking to outsiders - largely a non-issue.

As a good friend often likes to say: "it's a rounding error".

So, I think for the video/audio production industry...

"Come for the software licensing savings. Enjoy the shiny chrome for free!"

To answer this, I won’t name the company I was working for, but I believe they were trialling this... licensing by the active core usage / seat.

The trial they had had good results, they could milk out 10k more from people in this model because their machines had so many more cores... yes: it didn’t matter how many cores the software could use, it looked at how many cores were available to be used, and used the theoretical maximum for the licensing unit.

 

[rkuo]

This computer isn't for everyone, nor is it supposed to be "reasonably priced" by most standards.

It was important to take back the thought leadership position. The biggest computer company in the world needs to be able to make the "best" and "most powerful" computer. Cost be damned. Apple had increasingly been looking like a dinosaur that couldn't pull its head out of the iPhone's ass.

It's very hard for people to believe you're a great computer company if you can't make a power user's computer. And for good reason. Thank god they finally hit the mark here ... or at least, they're close enough to the mark that people can argue about it.

 

[cram501]

It was important to take back the thought leadership position. The biggest computer company in the world needs to be able to make the "best" and "most powerful" computer. Cost be damned. Apple had increasingly been looking like a dinosaur that couldn't pull its head out of the iPhone's ass.

It's very hard for people to believe you're a great computer company if you can't make a power user's computer. And for good reason. Thank god they finally hit the mark here ... or at least, they're close enough to the mark that people can argue about it.

What is that supposed to mean "thought leadership position"? It sounds like a meaningless commercial sound bite. There is nothing special about the Mac Pro. It is just an enterprise product. Some in the video industry may be happy it was created because of Mac OS and the tools that become available, but I"m not seeing what is so special about it.

This is a tool for a niche market. We will see if Apple sticks with this or lets it languish for years again.

It isn't for everyone. It is not for me. But that doesn't mean it is ground breaking.

 

[mavericks7913]

Simple. There is a gap in Mac lineup that the mid-range desktop is missing!

https://appleinsider.com/articles/1...vjLSUwW9AESBC_Bfu5Kn3Ki71-jgYShUePP99OLnFjuRU

Mac Pro series served from prosumers to high-end users and yet this Mac pro-2019 is meant only for high-end. The fact is there are prosumers and low to mid-range Mac Pro users than high-end users. Apple believes that iMac Pro is a mid-range professional computer and this is a problem. AIO cannot replace the mid-range desktop. How many times do people have to suffer from AIO issues such as cooling performance and dust?

At this point, there is a huge gap in the middle of Mac desktop because Mac Pro 2019 is made for high-end production.

 

[flygbuss]

Filmaker's view on Mac Pro.
[doublepost=1560725997][/doublepost]
[doublepost=1560726188][/doublepost]Simple solution to the Apple / Nvidia feud:

Nvidia market cap = 88B
Apple market cap = 880B
​

Need I say more?

Finally! This!

He’s absolutely right, honest and backing up all his arguments.
Thanks for the link!

 

[rkuo]

What is that supposed to mean "thought leadership position"? It sounds like a meaningless commercial sound bite. There is nothing special about the Mac Pro. It is just an enterprise product. Some in the video industry may be happy it was created because of Mac OS and the tools that become available, but I"m not seeing what is so special about it.

This is a tool for a niche market. We will see if Apple sticks with this or lets it languish for years again.

It isn't for everyone. It is not for me. But that doesn't mean it is ground breaking.

It may be a foreign concept to you, but the halo effect from a top tier product is a real concept in product design and marketing. High end and concept cars, for example, are often not intended to turn a direct profit. Rather, they exist to bolster the credibility of the brand and the engineering team. They do this by demonstrating how their engineers and designers can push the limits of what a car is or what it does in some way. And frequently those innovations end up trickling their way down to consumer level vehicles.

I disagree this computer isn't special. It's the most powerful and expandable expression of a Mac desktop ever, and that alone means something. It has real innovations on the PCI interface, thermal design, and video editing fronts, so it's not just a PC knockoff in an Apple designed case.

The only thing that makes it niche is the pricing. In the context of being a halo product, the pricing means very little.

 

[cram501]

It may be a foreign concept to you, but the halo effect from a top tier product is a real concept in product design and marketing. High end and concept cars, for example, are often not intended to turn a direct profit. Rather, they exist to bolster the credibility of the brand and the engineering team. They do this by demonstrating how their engineers and designers can push the limits of what a car is or what it does in some way. And frequently those innovations end up trickling their way down to consumer level vehicles.

I disagree this computer isn't special. It's the most powerful and expandable expression of a Mac desktop ever, and that alone means something. It has real innovations on the PCI interface, thermal design, and video editing fronts, so it's not just a PC knockoff in an Apple designed case.

The only thing that makes it niche is the pricing. In the context of being a halo product, the pricing means very little.

Apple made an enterprise product. The only thing really different about it (besides a proprietary PCIe interface for extra power delivery) is that it runs Mac Os.

Making flowery statements doesn't mean it is not using standard hardware that other enterprise systems use. Yeah, they designed their own motherboard, like everyone else in the space. Yeah, they spent development dollars making it expandable, like everyone else in the space. Yeah, they can use Xeons with a high core count, like everyone else in the space. There is nothing cutting edge about it just because Apple made it. I don't see Apple pushing any limits here.

I'm not saying it is a bad system. I"m saying it is an enterprise system, like everyone else in the space except it runs Mac Os. If it suits your needs, that is great but the high price doesn't make it special.

 

[Nosferax]

I wouldn't be surprised if Apple is considering an expandable desktop as noted here. They do it with iPhones, offering their high end and then following it up with a limited phone that contains some of the features in the high-end product.

Not going to happen.
[doublepost=1560875159][/doublepost]

In the past they did because Xeon was the only choice for SMP.

It would be impractical and expensive for Apple to make one Mac Pro production line for i9/i7 and another for Xeon. You have to make two different logic boards. Too much expense and risk.

Not a problem if the cpu and chipset is on a processor module/tray. The rest is just software (drivers).

 

[pl1984]

Yes, there is some demand on the PC side, but I doubt the demand on the Mac side is very high for towers in the $2500-$4000 range taking into consideration the price difference to PCs, to support Apple's 30-50% margin.

I think there's more than enough demand and I think they could make a nice profit on such systems.

In addition, it would require more support and testing of 3rd party drivers since so much different hardware could be placed in such a machine.

I don't believe Apple has done this is the past, why would do you feel they'd do it now?

 

[ct2k7]

Not a problem if the cpu and chipset is on a processor module/tray. The rest is just software (drivers).

The sockets are different, so it would need a different motherboard?

 

[AidenShaw]

The sockets are different, so it would need a different motherboard?

But motherboards are cheap to design - look at the number of models of ASUS, MSI, GigaByte,... motherboards available.

It is ludicrous to suggest that Apple can't afford two different motherboards. Many of the mid-range workstations (Dell/HP/Lenovo/Supermicro) have both Core and Xeon options using different sockets.

Apple could easily use the MP7,1 as a basis for an xMac. Ditch the big Xeon socket, go with i7 and i9 CPUs, and put a cheaper shell on the chassis. (Who needs a "work of art" case on a machine that's under the desk?) Ditch the proprietary MPX modules and just have some double-wide (or 2½ wide) and a reasonable assortment of 6-pin and 8-pin PCIe auxiliary power jacks.

Push through a revision to T-Bolt to make DisplayPort signals optional. (Have you ever wondered how popular and inexpensive T-Bolt would be today without the stupid decision to require putting video signals on a PCIe extension bus?)

 

[ekwipt]

Filmaker's view on Mac Pro.
[doublepost=1560725997][/doublepost]
[doublepost=1560726188][/doublepost]Simple solution to the Apple / Nvidia feud:

Nvidia market cap = 88B
Apple market cap = 880B
​

Need I say more?

Finally! This!

He’s absolutely right, honest and backing up all his arguments.
Thanks for the link!

The guy made my brain hurt, he does say it all why would Apple come out with a $6000 MacPro with a 8 Core machine, 580X GPU, 32 GB of RAM and 256GB SSD in 2019. Makes no sense at all. Also the Display is great, but where' s an Apple 5K display that doesn't cost a car. Yes the higher end MAcPros make sense, but only for the top end studios, who would only start at a 16 core machine, 512GB SSD and Radeon MPX module

 

[Flaming Gerbil]

Apple often makes the base model configurations in many lines a bad value, and they've outdone themselves with the 2019 Mac Pro. The sweet spots usually show up in the middle range of the lines.

Linus Tech Tips makes the argument that the only reason to buy the base Mac Pro is if you were planning to rip out the RAM and processor, install your own RAM + processor, and see what you can get on eBay for the stuff you took out (He argues the 8-core chip would probably be worthless as the motherboard it fits into is so expensive it doesn't make sense to stick an 8-core into it.)

The cheap graphics card isn't so much an issue since not every application needs fast graphics (like Audio).

The paltry-sized SSD is also not necessarily an issue since many buyers in this category will have fast network storage infrastructure available (but, still, it looks pretty cheap to only include 256). And I'm guessing it will be pretty simple to install your own PCIe/NVMe SSDs.

I imagine the value will start to show up in the $10k+ configs with 12+ cores and custom configs to suit specific needs (like MPX graphics for video).

Presently it seems this machine serves the very specific niche of FCP/Logic users, and others wedded to Mac OS who require a lot of horsepower.

The things I personally like about this Pro design is the easy-access case, and the integrated cooling solution. I appreciate the ergonomics of easy access and low noise. And yes, I actually like the looks


Looking forward

Apple Insider makes the point that there's very large price gap in the 'headless' Mac lineup between the Mini and the Mac Pro, and there's a real need to fill the gap.

But, there is some hope that future model upgrades will help fill the gap now that Apple has the Mac Pro out of the way.

The iMac case design is very long-in-the-tooth and way overdue for redesign. Hopefully, this will be the next big upgrade we get, and the new iMac will solve the thermal issues plaguing both the current iMac and iMac Pro.

The Mini case design is also old. Maybe a Mini redesign is where we'll really get the gap filled between current Mini and Pro, with a larger more upgradable mini with it's thermal issues also solved.

And the BIG ONE looming over the entire Mac product line: the switch to Arm, which may finally start to distinguish Apple hardware from Intel/AMD chips the same way PowerPC did 25 years ago (my God, has it been that long: My 1st Mac was a 6100 , upgrading from my Atari MegaSTe with Mac emulation...the OG hackintosh ).

 

[AidenShaw]

(He argues the 8-core chip would probably be worthless as the motherboard it fits into is so expensive it doesn't make sense to stick an 8-core into it.)

$369.99 https://www.bhphotovideo.com/c/product/1381122-REG/asus_z11pa_u12_lga3647_intel_atx.html/?m=Y&c3api=0980,{creative}&msclkid=5c7ac3d3f6b318cfc99d9eda9d510587

Another LTT fail. You could almost buy three motherboards (qty 1, on the street) for the price of a monitor stand.

 

[ekwipt]

If nobody would buy the 8 Core Mac Pro then who the hell is it for?

 

[AidenShaw]

If nobody would buy the 8 Core Mac Pro then who the hell is it for?

It's to avoid completely scaring off potential customers by pricing the MP7,1 at "starting at $7999". Offer a $5999 model that almost nobody would buy.

Classic upsell technique - whether for computers, cars, dishwashers, toilet brushes....

Depending on your company procurement practices, a "low-ball" intro price can help with the approval process. Tell your manager that it's $5999. After she buys into that, show her the quote that you need options that bring it up to $7999. She approves, because she's already endorsed a $6K purchase, and you've only asked for $2K more.

(I'm an expert at using this trick, for example to turn 512GiB of RAM at the first approval into a purchase order for 1 TiB of RAM .)

$0 to $100K is a big step. $100K to $120K is noise.

 

[ekwipt]

It's to avoid completely scaring off potential customers by pricing the MP7,1 at "starting at $7999". Offer a $5999 model that almost nobody would buy.

Classic upsell technique - whether for computers, cars, dishwashers, toilet brushes....

Depending on your company procurement practices, a "low-ball" intro price can help with the approval process. Tell your manager that it's $5999. After she buys into that, show her the quote that you need options that bring it up to $7999. She approves, because she's already endorsed a $6K purchase, and you've only asked for $2K more.

(I'm an expert at using this trick, for example to turn 512GiB of RAM at the first approval into a purchase order for 1 TiB of RAM .)

$0 to $100K is a big step. $100K to $120K is noise.

I get it, I use a similar technique with my wife, told her about the new Apple Monitor and she just said "ridiculous" and then I said there are other ultrawides that are just $1000. So looks like I'm getting anew monitor before tax time in Australia

 

[Fried Chicken]

I know there's a lot of discussion going on regarding the price and I don't want to debate that here

You cant separate one from the other.

The price is what makes it too niche. If it came in at $2k or even $3k it wouldn’t be a niche computer, but an awesome computer. At $6k, it’s a niche as far as computers go, and that’s a bad thing.

 

[AidenShaw]

I get it, I use a similar technique with my wife, told her about the new Apple Monitor and she just said "ridiculous" and then I said there are other ultrawides that are just $1000. So looks like I'm getting anew monitor before tax time in Australia

Yes, good play. On the other hand you're talking about spending your money - my example was spending company money. But, really great way to make her believe that a $1000 monitor is cheap.

Of course, our company has a motto to "spend company money as if it were your own". I look for real value, but qustion that I would spend my own money to buy a 72C/144T server with 2 TiB of RAM and 7 TB of NVMe flash. But I spent $150K each for three of those systems for work - because we needed the power.

Actually, I would spend $500K of my own money for those systems if I thought that they would enhance my wealth by more than $500K. But I hold enhancing my portfolio to a higher ROI standard than the server purchases for work.

 

[Horselover Fat]

I agree with a lot of his arguments. He basically says ‚It’s not built for you‘. I knew it before I saw the price tag. Again (like the 6,1) it is targeted at filmmakers and production houses (you had to buy certain non-upgradeable GPUs with the 6,1). The earlier generations up to the 5,1 were expandable, suited and priced for non high-end use cases. The reluctance of the market to accept the 6,1, the years of mourning from disappointed users and Apple’s mea culpa led to expectations that the 7,1 would fit that hole again. It doesn’t. Now let’s look elsewhere.

 

[theluggage]

I agree with a lot of his arguments. He basically says ‚It’s not built for you‘. (snip) it is targeted at filmmakers and production houses

Not quite - it is targeted at those filmmakers and production houses who are already locked in to FCPX/ProRes and/or Logic Pro - which is the only solid argument he gives for sticking with Mac. Show us a filmmaker currently using PC/Windows hardware saying (without being paid) "wow! I'm switching (back) to Mac" and it would be more convincing. After years - and a couple of leasing/tax write-off cycles - of not having a credible Mac Pro, that pool has already been whittled down to a hard core who either can't or won't change their workflow. Unless the new MP can attract new blood, that pool is going to be ever-diminishing.

There are two separate issues with the new MP:

1. The lack of a credible entry-level, expandable, desktop in the $3k-$4k price range for customers that aren't editing and scoring Hollywood movies but were served by the "classic" Mac Pro (and would have been served by the trashcan if that hadn't been a dead end). Many of those customers were already paying a premium for a Xeon/ECC system that exceeded their requirements just to get an expandable, headless desktop Mac, so its not even worth getting into a HP X4 vs Mac Pro "Top trumps" argument.

2. For the "true pro" market that the new machine is "meant for" - unless you're locked in to FCPx or Logic, what exactly is the advantage of a MP over a Windows system running much the same applications and with NVIDIA cards that are better supported by pro software (even the guy in the video laments this)? We don't know what sort of value-for-money the higher-specced Mac Pros will offer yet - but it would be rather optimistic to assume that T2-compatible SSD upgrades and MPX modules won't carry a premium over their generic equivalents. As far as I can see, the only real innovation in MPX is a somewhat neater way of adding quad GPUs that avoids flying power leads.

One problem with the Trashcan was that it was basically a FCPx/Logic appliance - relying on applications being optimised for its peculiar architecture. The new MP is also, really, a FCPx/Logic appliance, because the cost just doesn't make sense unless you're already locked into that ecosystem.

I'll leave the display out of this. Either the new "pro" display is a viable replacement for high-end reference displays costing 3x as much or it isn't. Other displays are available (other desktop towers that can legally run MacOS aren't) Currently, we have Apple's word that it is, but the reality is dependent on industry approval/acceptance once these get into the hand of critical reviewers. The infamous stand is technically irrelevant, but does give a glimpse of what Apple's pricing policy looks like when not shrouded in techno-snake-oil (...and also distracts people from the $1000 anti-glare option and $200 VESA adapter which are probably more usurious 'hidden costs').

 

[flygbuss]

The guy made my brain hurt, he does say it all why would Apple come out with a $6000 MacPro with a 8 Core machine, 580X GPU, 32 GB of RAM and 256GB SSD in 2019. Makes no sense at all. Also the Display is great, but where' s an Apple 5K display that doesn't cost a car. Yes the higher end MAcPros make sense, but only for the top end studios, who would only start at a 16 core machine, 512GB SSD and Radeon MPX module

As he mentioned as well, most studios (myself included) use big and fast network storage solutions. I also don’t need a powerful gpu since I work with audio. Why should I pay for a powerful gpu if I have no need for it?
Also there are no benchmarks and real life tests how this configuration actually performs, so let’s wait until we see them.
Just because the specs on paper don’t make sense to you doesn’t necessarily mean they actually make no sense to others.
There are plenty of 5k displays out there. I don’t mind a dell display if it does the job. I need a licensed macOS on a reliable machine, for everything else I just buy what fits my needs.

 

[Nosferax]

The sockets are different, so it would need a different motherboard?

And said motherboard can be a module pluged into a slot on a backplane, like the oMP.