MP 7,1 Mac Pro 8.1 - WHAT IF...

[deconstruct60]

But hasn't it been said and confirmed, from multiple sources, that they WILL reuse the same case/outside for the 8,1?

'confirmed' is a huge stretch. It has been said. But also true that many of these rumor sources read each other's stuff. Being 'said' again is what? echo chamber or an independent source ?

For almost two years Gurman was saying 'half sized'. Him back pedaling out of 'half sized' doesn't say a lot for his 'sources' track record. Gurman has also mentioned though that there was thought of selling 7,1 and 8,1 side by side. So Apple still making 7,1 cases shouldn't be a total shocker. The longer they are going to keep selling the 7,1 the cheaper the 8,1 gets by reusing the exact same part. (e.g., Mini , MBP 13" now on 3rd internals iteration. )

So in this point, it wouldn't be old hardware...

It isn't an announced product so can't really be shown as a "M-series Mac product family" portrait.

The other problem though is that a tower would blow out the symmetry of the family portrait at the end. The mini and Studio and placed on the opposite end on purpose to create symmetry in that photo composition. The "smaller to bigger" screens are carefully crafted to do the same thing. go from smaller to biggest screen and then back to smaller on the other side.

If stuck a goofy more than 2x taller tower in the middle , then you would basically mess up the shot composition. For not 'good' reason because it is not a M-series Mac (yet) [ and they still aren't going to put the rackmount Mac Pro in that family picture. So the 'completely inclusive of every option' is unlikely even later. ] .

 

[prefuse07]

your essays crack me up, man

 

[MisterAndrew]

It makes sense to me that they keep the design the same. They spent a lot on 7,1 R&D and probably haven't yet made the return on that they are hoping for. That could also be why they sold the 6,1 for so long. It took them several years to reach their target figures. I also don't see any reason why it needs to be smaller. The rack version certainly doesn't need to be smaller.

It would be interesting though if the chassis and case are exactly the same so that the 7,1 could accommodate an 8,1 motherboard. It would be cool if Apple offered an upgrade service for existing owners.

 

[ZombiePhysicist]

Obviously they'll cherrypick benchmarks to make themselves look good, but what company hasn't? And what professional cares about benchmarks versus actual performance in the real world?

As for "faking" wins against a loaded Mac Pro, I doubt they're going to try and do that, and it doesn't make sense to do so, especially given what we know about its potential capabilities. Even among people buying Mac Pros, vanishingly few people are buying $40,000+ configurations. For the vast majority of potential customers, the new Mac Pro doesn't need to beat a 28-core Xeon with 1TB of RAM and dual W6800X Duos, because that's not a realistic matchup for most people.

You can see what they actually do on the Mac mini page, where they compare the new machines to the outgoing model as a baseline, along with the last Intel iMac with a 5500XT. That's not a random choice for comparison, because they're clearly aiming the M2 Mac mini at 27" iMac fans (because yeah, that product doesn't seem likely to come back) and also they pick what was probably one of the most popular GPU upgrades (5500XT), rather than the 5700 or 5700XT BTO options. Because that's what most people probably upgraded to, and it's a more useful example as well as making their numbers look larger. Likewise, what's the comparison on the Mac studio page to? A 16-core Mac Pro and a 5700X, because that's the person they're targeting with that—someone who had a $8K or 9K Mac Pro and can now get much more power (if not the expandability) from a $2K and up machine.

So you can bet since the 28-core Xeon processor is going to get smoked by the M2 Max, they'll have that as a bar, but I bet they're just going to compare the machine's GPU power to something like the 6800X or 6900X—where not only will they look more favorable, but it's much closer to the GPUs most people actually have in their Mac Pros and will be looking to upgrade from.

(Here the fact they haven't added MPX 7000-series GPUs also helps them, I suppose.)

If the thing ships with an M2 Max and Ultra, then clearly the story they're telling with the Mac Pro isn't going to be about its straight performance anyhow—it's going to be about what makes it different than a Studio (to be determined.)

They lied when they compared the studio to the Mac pro 7,1 with a 6900xt. They cherry picked some bulls*** process optimized on their gpu, and yet the 6800xt, forget the 6900xt, smokes the ultra studio in metal and most other real world activities and actual representative benchmarks.

Since we have no real tech press, just moron lackeys, that let them get away with this bs last time, I suspect apple will do the same and bald face lie to us again with misrepresentative cherry picked stats again.

 

[prefuse07]

They lied when they compared the studio to the Mac pro 7,1 with a 6900xt. They cherry picked some bulls*** process optimized on their gpu, and yet the 6800xt, forget the 6900xt, smokes the ultra studio in metal and most other real world activities and actual representative benchmarks.

Since we have no real tech press, just moron lackeys, that let them get away with this bs last time, I suspect apple will do the same and bald face lie to us again with misrepresentative cherry picked stats again.

Yup, for sure they will try to pull this clown show again, and then there will be a bunch of idiots creating threads about it here to stir $h!t up, as always -- Frankly I'm surprised apple didn't try doing it with the unveiling of the new m2 mini the other day...

 

[Mac3Duser]

I must say that since the last mac m2, I am rather optimistic.
I'll just tell you why.
I think the 7.1 was created knowing that there would be the 8.1 and for the 8.1 to be better in every way (and the same price level).

Now a projection of what 8.1 might be: (and of course I don't know, it's just speculation, based on what seems possible to me)
- storage that can be increased via a proprietary card (with regular SSD M2 or not, that, we'll see)
- additional graphics card (an Apple graphics card, a bit like they did with afterburner, it seems more logical to me than an AMD) with very good results on Blender (they invested, if I'm not mistaken) and possibilities to put 2 or even 3 additional cards
- card for audio, for Logic, pro tools
- definitely some kind of extra ram module because they know that 192 gb may not be enough for some pros so something like an accelerator card with a lot of memory to avoid swapping

 

[avkills]

Yup, for sure they will try to pull this clown show again, and then there will be a bunch of idiots creating threads about it here to stir $h!t up, as always -- Frankly I'm surprised apple didn't try doing it with the unveiling of the new m2 mini the other day...

Lol. There is another thread where people are impressed with the Metal results. Even if you double the score (M2 Max) we are still falling short of even a W6800x. Granted there is a huge power envelope difference, but if any of the other 2019 Mac Pro users are like me, I don't care about power usage. I just want things to be done being rendered...lol.

I just wish more software (Hello Adobe) would utilize multiple GPUs. Geekbench 5 never even manages to get my GPU usage to 50%, and it only tests 1 GPU at a time; getting around 140,000 Metal. Seems like a useless benchmark to me, oh well.

 

[deconstruct60]

It makes sense to me that they keep the design the same. They spent a lot on 7,1 R&D and probably haven't yet made the return on that they are hoping for.

This is a metal case with no electronics. It costs a couple hundred dollars a pop. They have been selling it with a 25+ % mark up for over two years and they haven't made their money back? Probably not. Apple has made the money back on the case. No way Apple was/is selling that case at a loss and Apple's component mark ups are way higher than the industry norm for cases. They are charging a hefty sum for it to the end users.

They can throw the fans and power supply in on the reuse. Again talking a couple of hundred dollars with 25+ % mark up passed onto the customers for last 2+ years.

The main logic board is NOT being reused. It wouldn't be reused if Apple moved onto another Intel socket. Neither one of those would be the same.

The two I/O daughter boards (well three boards if include the rackmount) for the Thundergbolt and USB ports and wi-fi/bluetooth. Those won't be the same design either. The Thunderbolt controllers are in the SoC and Wi-FI/Bluetooh are both getting updates. The "slot 8" design would have be substantively different as just routing TB protocol signal in a more complete state (no edge discrete controllers needed ).



Folks have hand waved that the MPX connector was some humongous R&D cost sink. It probably is not. And the external case is completely decoupled from it. Reusing the exterior enclosure doesn't automatically get you a required MPX connector on the logic board.

That could also be why they sold the 6,1 for so long. It took them several years to reach their target figures.

There is no decent quantitate data to support that. Apple's mark up is 25+ % above costs. They relented on price after 4 years in 2017 when the same 2013 value proposition was a huge joke. ( they rotated the 4 year old mid range processor down to the entry position and likely keep just as high , if not higher , margins. )

GPU wise there were no decent AMD updates during the first 2-3 years of the Mac Pro 2013 going stale. AMD had some newer stuff but none of the upper end 'newer models' improved on perf/watt so would still be in the same boat. The lower end stuff would incrementally help with the 'painted into the thermal corner' problem but be no performance improvement option. So Apple did mostly nothing in this area and spent most of they time actually growing the mac ecosystem substantively larger.

CPU wise there was a incremental performance bump, but also didn't provide any substantive relief on thermal corner.

$0.00 on new R&D was just cheaper. By 2016-2017 R&D funds on new Apple Slicon work was going to provide a much higher return on investment than any new " stuff a new board into the old 2009 case " would provide. Apple need a far more expensive box that they could slap 25+ % margins on to serve as a transition box for 2-3 years. They'd sell less but they get a quicker breakeven point also if priced with enough "low volume tax" applied to the base price. So they put a 100% increase on the entry model price.

I also don't see any reason why it needs to be smaller.

The case doesn't have to be smaller , but Apple also doesn't need to put as large of a main logic board in there either.

The rack version certainly doesn't need to be smaller.

CPU core count density wise that 5U rack version does suck. It is certainly less 'rack hostile' than the 2009 case, but it is chunky for a large number of use cases. Most macOS coloc and web services in cloud business is grounded on the Mini form factor. MacStadium bought several pallets of Mac Pro 2013 systems on the last day Apple was selling them having looked deeply at the 2019 rack case. It wasn't going to work there in most cases.


There were a steady stream of customers buying the MP 2013 all along. They may not have been the group 'protesting' and 'boycotting'. As long it was enough to pay the bills, limping along on the MP 2013 group that was paying for a creditable option for Apple.

It would be interesting though if the chassis and case are exactly the same so that the 7,1 could accommodate an 8,1 motherboard.

That won't really work that well if the daughter boards are different and Apple doesn't really general retail bare boards.

And one reported leak is that the slot count is down to 6 . So those currently with ≥ 7 cards won't be an upgrade (unless dump a few on the transition).

It would be cool if Apple offered an upgrade service for existing owners.

Did Apple offer a motherboard upgrade service to 2009 users? 2013 users? Nope. Not likely at all.
Apple knows there is a rogue market for Mac spare parts out there. After a while there will floating 'used' boards floating out there and the 'wouldn't be cool' crowd can bust out their trusty screen driver and thermal paste and tinker around inside the box. Apple really doesn't have to do much but make sure there is a huge run on spare parts that messes up the regular warranty (and out of warranty ) repair flow over the product service life time without ballooning their inventory costs too high.

 

[deconstruct60]

I must say that since the last mac m2, I am rather optimistic.
I'll just tell you why.
I think the 7.1 was created knowing that there would be the 8.1 and for the 8.1 to be better in every way (and the same price level).

The M2 actually demonstrates almost exactly the opposite. The M2 still has some of the same backslides that the M1 had. Still sub 32GB RAM capacity on the second generation. The number of display outputs went no where. The M2 is still on HDMI 2.0. Not sure how Apple's tech spec page is specifying compliance with Thunderbolt 4 qualifications where one video out is permanently stuck on HDMI. (if so , still technically just USB4).


So the notion that they are going to be "better in every way". Apple failed on both first and second generation with the plain Mx SoC. There are definitely areas of great improvement, but Apple commitment to some sweeping generalization of every single possible feature? Nope.


A M2 Ultra will get much closer to 'covering' a W6800X .

The fully BTO option version of the MBP 16" is up in the $6K range. (not that it makes sense in vast majority of situation to buy that much SSD capacity from Apple). It is up in the 7,1 Mac Pro entry price range. I think Apple pulling expectations into $6K isn't that bad zone was one of the basic points of the 7,1. That is probably the main precursor to the 8,1. It isn't exact duplicate features. It is price zone.

Now a projection of what 8.1 might be: (and of course I don't know, it's just speculation, based on what seems possible to me)
- storage that can be increased via a proprietary card (with regular SSD M2 or not, that, we'll see)

The primary boot drive? No. M2 clearly demonstrates that it is not 'regular M2". Apple hasn't done regular M2 since the T2 let alone the start of M-series SoCs.

Apple's SSD modules (which are not SSDs ) isn't really speculation. If Mac Studio and Mac Pro 2019 have them ... why would that change?


A plain PCI-e slot to put a M2. SDD carrier card into. Those already work with TB Enclosures for M1 and M2 so not much of 'prediction' or 'speculation that would work or be pretty much essential requirement for a Mac Pro case of the same size.

- additional graphics card (an Apple graphics card, a bit like they did with afterburner, it seems more logical to me than an AMD) with very good results on Blender (they invested, if I'm not mistaken) and possibilities to put 2 or even 3 additional cards
- card for audio, for Logic, pro tools

Again there are over 50 PCI-e cards that already work with M1 and M2 systems. That a Mac Pro would leverage those is not some huge speculative leap. If it didn't leverage those then that would be more speculative. It doesn't make any rational sense for the Mac Pro to skip over parts that already are working and for sale.

- definitely some kind of extra ram module because they know that 192 gb may not be enough for some pros so something like an accelerator card with a lot of memory to avoid swapping

The M2 Pro / Max roll out doesn't put much credence to that at all. The M2 Pro is the exact same 32GB limit as the M1 Pro. Apple isn't in some giant panic state about ramping up RAM sizes.

Apple once again on the M2 Pro and Max introductions got up and preached yet another sermon about maximum pref/watt is their primary focus.

It if far , far , far more critical that Apple adds some decent x16 PCI-e v4 (or two ) backhaul to the Mac Pro SoC than any memory slot that will basically bust Apple's "Uniform , Unified" Memory model. No PCI-e backhaul and all that pointing at discrete GPUs cards is basically a bust. Even 8K HDR video capture cards. Or 4 x M.2 PCI-e v4 carrier cards.

The current M1 and M2 models are way further behind on PCI-e backhaul than they are on memory capacity.

This affordable $120 CPU from Intel has more readily available PCI-e backhaul than the M2 Pro/Max.

New 2023 N-Series Gracemont CPUs Are a HUGE Upgrade

Intel has nine new Alder Lake-N SKUs as part of its N and X series of efficient core processors with new features and up to 2x the core count

www.servethehome.com

 

[Joe The Dragon]

Apple's "Uniform , Unified" Memory model
does use the SSD as swap and an pool of ram as an ram disk (for swap and tmp files can be an boost over ssd sawp)
with less ware

 

[avkills]

Apple's "Uniform , Unified" Memory model
does use the SSD as swap and an pool of ram as an ram disk (for swap and tmp files can be an boost over ssd sawp)
with less ware

I do not understand what point you are trying to make here. Every single version of Mac OS X has used the OS boot drive for swap. The fact the all new systems are on SSD drives just makes swap faster.

 

[Joe The Dragon]

I do not understand what point you are trying to make here. Every single version of Mac OS X has used the OS boot drive for swap. The fact the all new systems are on SSD drives just makes swap faster.

and the new mac pro can use ram sticks for that

 

[innerproduct]

This is the kind of cpu/gpu one should expect from a 10000$+ mac pro.
If nvidia can do it, why not apple? https://wccftech.com/nvidia-grace-c...-5x-efficiency-gain-over-amd-epyc-milan-cpus/

 

[Mac3Duser]

This is the kind of cpu/gpu one should expect from a 10000$+ mac pro.
If nvidia can do it, why not apple? https://wccftech.com/nvidia-grace-c...-5x-efficiency-gain-over-amd-epyc-milan-cpus/

I think this is what Apple will do for mac pro 8.1
(Metal, Machine Learning, AI, etc)

 

[avkills]

and the new mac pro can use ram sticks for that

Meh; any applications that need that much RAM pretty much let you adjust settings and such.

 

[AndreeOnline]

Since we have no real tech press, just moron lackeys...

I've been thinking a bit about picking up the torch after Barefeats. But it's easier said than done.

[Hijack]
The actual work part, and presenting the information in an interesting way, is something I would consider.

But I don't know how to get hold of the hardware (the retailer-tester culture is most likely different in Switzerland than in the US, for example). Then again, I haven't reached out to anyone (local Apple, or Premium Reseller) either yet.

The same character and ethics that is suitable for presenting valid, non-BS data, would prevent someone from "taking advantage" of free return within 2-4 weeks and whatnot. Is that what all the YouTube box openers are doing? "Buying" stuff, ripping it open, and then returning it with their fatty fingerprints all over within the free return period? If that is the case: lock them up and throw away the key....

But who knows... step one would be making contact with the right dealer. Semi-pun intended.

Today's money/commercial interest lives in 'the world of average'. That's where you find all your potential customers. YouTube thrives on it. The "stars" there are average people, on the cute side of the spectrum, that people relate to. Film yourself with your iPhone, with glossy lipstick as you are "reviewing" make-up in the mall and you'll get 3.4M hits.
Produce a perfectly edited, straight to the point video that shows how to fix glossy car reflections under subdivision in Blender and you'll get 1834 views. Keep the good work up releasing new videos weekly and your channel is sure you reach 16408 subscribers within the year.
[/Hijack]

 

[StuAff]

I've been thinking a bit about picking up the torch after Barefeats. But it's easier said than done.

[Hijack]
The actual work part, and presenting the information in an interesting way, is something I would consider.

But I don't know how to get hold of the hardware (the retailer-tester culture is most likely different in Switzerland than in the US, for example). Then again, I haven't reached out to anyone (local Apple, or Premium Reseller) either yet.

The same character and ethics that is suitable for presenting valid, non-BS data, would prevent someone from "taking advantage" of free return within 2-4 weeks and whatnot. Is that what all the YouTube box openers are doing? "Buying" stuff, ripping it open, and then returning it with their fatty fingerprints all over within the free return period? If that is the case: lock them up and throw away the key....

But who knows... step one would be making contact with the right dealer. Semi-pun intended.

Today's money/commercial interest lives in 'the world of average'. That's where you find all your potential customers. YouTube thrives on it. The "stars" there are average people, on the cute side of the spectrum, that people relate to. Film yourself with your iPhone, with glossy lipstick as you are "reviewing" make-up in the mall and you'll get 3.4M hits.
Produce a perfectly edited, straight to the point video that shows how to fix glossy car reflections under subdivision in Blender and you'll get 1834 views. Keep the good work up releasing new videos weekly and your channel is sure you reach 16408 subscribers within the year.
[/Hijack]

Rob Morgan's missed. Nice guy who did a lot of really useful, practical real-world testing.

 

[avkills]

Rob Morgan's missed. Nice guy who did a lot of really useful, practical real-world testing.

Yes I was sad to hear of his passing. Very valuable data that he presented. AnandTech seems almost dead now. Slowly going downhill after Ian C left.

 

[chfilm]

I've been thinking a bit about picking up the torch after Barefeats. But it's easier said than done.

[Hijack]
The actual work part, and presenting the information in an interesting way, is something I would consider.

But I don't know how to get hold of the hardware (the retailer-tester culture is most likely different in Switzerland than in the US, for example). Then again, I haven't reached out to anyone (local Apple, or Premium Reseller) either yet.

The same character and ethics that is suitable for presenting valid, non-BS data, would prevent someone from "taking advantage" of free return within 2-4 weeks and whatnot. Is that what all the YouTube box openers are doing? "Buying" stuff, ripping it open, and then returning it with their fatty fingerprints all over within the free return period? If that is the case: lock them up and throw away the key....

But who knows... step one would be making contact with the right dealer. Semi-pun intended.

Today's money/commercial interest lives in 'the world of average'. That's where you find all your potential customers. YouTube thrives on it. The "stars" there are average people, on the cute side of the spectrum, that people relate to. Film yourself with your iPhone, with glossy lipstick as you are "reviewing" make-up in the mall and you'll get 3.4M hits.
Produce a perfectly edited, straight to the point video that shows how to fix glossy car reflections under subdivision in Blender and you'll get 1834 views. Keep the good work up releasing new videos weekly and your channel is sure you reach 16408 subscribers within the year.
[/Hijack]

Would be much appreciated if someone would continue the work of barefeats. I loved that site, it was so valuable!!

 

[AndreeOnline]

Rob Morgan's missed. Nice guy who did a lot of really useful, practical real-world testing.

Yes I was sad to hear of his passing. Very valuable data that he presented.

I loved that site, it was so valuable!!

Yes, it was great. Specifically since it was so focused and straight to the point using both standard components and upgrades for the Mac Pro, as well as sometimes including more experimental stuff out of curiosity.

That said, I think even mid range Macs now, with Apple Silicon, are so great at all the "basic" computing stuff (word processing, publishing, graphic design, normal photo work, and so on....) that most bases are more or less covered out of the box. I still think there is some interest in knowing if those memory upgrades really translate to real world performance and so on, but now with hardware video encoding chips across all Macs, the playing field is more level than it was.

But it would still be cool to have it black on white what performance increases can be gained (and where the true benefits are), by adding more and more GPUs—single and dual.

The tests are especially valuable since theory doesn't always (ever?) turn into reality. At the end of the day, we want to know EXACTLY how software X or Y reacts to component A and B. Or AA and BB.

 

[ZombiePhysicist]

We could start a collective and of pro/enthusiasts to take the baton over from bare feats. We have enough folks here to arm wrestle an avatar render farm.

 

[StuAff]

We could start a collective and of pro/enthusiasts to take the baton over from bare feats. We have enough folks here to arm wrestle an avatar render farm.

That would work. 'Remote mad scientists' contributed to a lot of Rob's testing. I did myself back in 2006, when he wanted stats for 1,1s with an X1900 XT.

 

[chfilm]

Yes, it was great. Specifically since it was so focused and straight to the point using both standard components and upgrades for the Mac Pro, as well as sometimes including more experimental stuff out of curiosity.

That said, I think even mid range Macs now, with Apple Silicon, are so great at all the "basic" computing stuff (word processing, publishing, graphic design, normal photo work, and so on....) that most bases are more or less covered out of the box. I still think there is some interest in knowing if those memory upgrades really translate to real world performance and so on, but now with hardware video encoding chips across all Macs, the playing field is more level than it was.

But it would still be cool to have it black on white what performance increases can be gained (and where the true benefits are), by adding more and more GPUs—single and dual.

The tests are especially valuable since theory doesn't always (ever?) turn into reality. At the end of the day, we want to know EXACTLY how software X or Y reacts to component A and B. Or AA and BB.

About the financing, didn’t he collect donations to order a full specked Mac Pro back in the day? 🤓

And about the YouTubers, I mean there are various flavors I guess, like how I understand maxtech for example orders and (probably) sends stuff back- while my colleague iknowreview from Germany for example gets stuff sent by apple in a lot of cases and has to return it in 6 months. But they only do that for big players with glossy videos and stuff, so..

 

[fuchsdh]

That would work. 'Remote mad scientists' contributed to a lot of Rob's testing. I did myself back in 2006, when he wanted stats for 1,1s with an X1900 XT.

Realistically that’s the only way to make it work since most of the testing requires continued access to expensive hardware. (Barefeats crowdsourced the 7,1.)

 

[maikerukun]

I must say that since the last mac m2, I am rather optimistic.
I'll just tell you why.
I think the 7.1 was created knowing that there would be the 8.1 and for the 8.1 to be better in every way (and the same price level).

Now a projection of what 8.1 might be: (and of course I don't know, it's just speculation, based on what seems possible to me)
- storage that can be increased via a proprietary card (with regular SSD M2 or not, that, we'll see)
- additional graphics card (an Apple graphics card, a bit like they did with afterburner, it seems more logical to me than an AMD) with very good results on Blender (they invested, if I'm not mistaken) and possibilities to put 2 or even 3 additional cards
- card for audio, for Logic, pro tools
- definitely some kind of extra ram module because they know that 192 gb may not be enough for some pros so something like an accelerator card with a lot of memory to avoid swapping

EXACTLY!!!! That's kind of what I was dreaming up in my OP to start this thread...That's really all I want...I DON'T CARE WHO MAKES THE GPU, it just needs to be THE FASTEST GPU AVAIALBLE...APPLE OR PC.