I think the iPod had somewhat of a halo effect however I do feel the iPhone had a more profound halo effect. Combine that with the Macs switch to Intel processors and the Mac became much more available to users. However, the iPhone is Apple's bread and butter these days. Without it I don't know where Apple would be. I don't believe they'd be nearly as big as they are today.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MP 7,1 Mac Pro 8.1 - WHAT IF...
- Thread starter maikerukun
- Start date
-
- Tags
- 7.1 mac pro next generation
- Sort by reaction score
I think the iPod had somewhat of a halo effect however I do feel the iPhone had a more profound halo effect. Combine that with the Macs switch to Intel processors and the Mac became much more available to users. However, the iPhone is Apple's bread and butter these days. Without it I don't know where Apple would be. I don't believe they'd be nearly as big as they are today.
I think the above is fair, and I'm glad you crucially pointed out there were different size halo effects for different categories of things. I'll add one. I'd push that OSX also had a bit of a halo effect, likely to a far lesser degree than the above noted ones.
But there is a bit of a domino game going on. OS X influencers influenced the iPhone team. It is the same OS with different libraries essentially. That was the case inside of apple. The Forstall team won the bakeoff and their opinions on unix objects etc won the day. They were the influencers of that halo. And one set of smaller halo influencers influenced the next and so on. Of course the effect isn't immediate; there is lag/delay between the halo spheres effecting one after the other.
After the influencers influenced the influencers that influenced the YouTube influencers that influenced the masses, each of these dominos got us to the mass halo effect. But there are smaller ones that were key dominos that had their own smaller halos.
But unless something blocks the path, or prevents the teeny tiny (seemingly) insignificant domino from toppling, the dominos will cascade and eventually fall and break.
Last edited:
So far, there's no doubt that Mac hardware has seen a significant boost due to the AS transition.
We may feel differently in the future though if Intel is resurgent and / or it turns out expandable, Mac Pro-style hardware just isn't practical with an SoC architecture.
We may feel differently in the future though if Intel is resurgent and / or it turns out expandable, Mac Pro-style hardware just isn't practical with an SoC architecture.
I am really hoping the latter becomes a reality, because having to buy a new machine everytime you want to upgrade is just stupid, both for consumers as well as for the environment (and yes, screw corporate profits).
I disagree. This presumes the world has to only be the one SoC design apple started with.
Sure they had to start with that and build that up, but now imagine they come out with a TRUE desktop SoC that has plenty of PCI lanes, allows eternal RAM, ECC, etc... What would that mean? And why do we STILl not have a 27/30" iMac/pro?
Well it would mean apple deploys a TRUE desktop chip for the Mac Pro, but guess what, now they can release a 27" iMac with a REAL desktop chip, and one that STILL allows discrete graphics cards. Oh, and they can put that in the next Studio update. Oh, and they maybe can even make a pared down one for the MacBook Pro to allow a discrete GPU option.
And so they extend the original iPhone A chip to a 'glorified' M chip to give them time to make their 2nd tier of chips, true desktop chips. And once that is there, they can back fill down the line.
Why must it be 'we make only one kind of chip for everything'? Seems reasonable for them to do that while they NEED to but then to supply a true desktop level chip that can handle expandability too.
Now maybe they wont do that, but if you think about using such a chip ONLY for the Mac Pro, doesn't make too much sense, but if you think about it being used first for a Mac Pro but then used to down fill their entire product line with a more expandable pro chip, it makes plenty of sense.
Sure. It’s just a bit worrying when 95% of Mac’s are able to use the beefed-up A Series. But anyway, I won’t repeat myself; the 8,1 will be fascinating, as only then will we know Apple’s future direction with the AS MP.
95%? You may be right. Let's see, the iMac 27", the Mac mini, and I suggest the top end Mac book pro 16" could benefit from that. What percent is that. 5%? Maybe. Could be as much as 30% though, and I would argue, 30% of they're much higher margin machines. If so, does that maybe add up to near 50% of margin?
We still have to wait 3 months and we will know.
there will be 8.1 slides with diagrams compared to 7.1 with the 8 core xeon and with the 580x, "200% better"
Of course it will be amazing
there will be 8.1 slides with diagrams compared to 7.1 with the 8 core xeon and with the 580x, "200% better"
Of course it will be amazing
I high performance M2 CPU cluster + a dGPU would basically have the TDP of an Ultra. The Studio is already a mini with an Ultra, and it needs a big ‘forehead’ filled with copper heat sink.
On an related note, Apple would have the option of putting an Ultra in a 27” iMac or MBP, but would likely rather prioritise thinness and power consumption / battery life / noise, as well as keep the price reasonable.
Youre not getting it. I dont care about the TDP. It would have very different capabilities and wouldnt need mass of GPU crap in current chips, because youd have a discrete and more powerful GPU.
Well, it would mean a very expensive system to pay for the development of a "desktop" SoC that was only ever sold in relatively tiny quantities in Mac Pros and hypothetical "pro" iMacs.
It would mean a system constrained by the same bottlenecks of external RAM speed, separate RAM and VRAM, PCIe speed and the capabilities of 3rd party graphics cards as competing Intel and AMD systems. (...and if you're going to have external RAM and GPUs how is it still a "system on a chip"?)
Supporting discrete (non-Apple) GPUs would remove any incentive for the makers of high-end software (the sort of specialised stuff that sells to potential Mac Pro customers) to optimise it for Metal/Mx Ultra.
It would probably give Intel and AMD a good run for their money - but it would be very hard to create an Xeon/Threadripper-killer: going for external RAM, PCIe graphics etc. would throw away the advantages of fast, unified RAM and GPUs optimised for Metal and the economies of a "true" SoC (no PCIe, RAM, M.2 slots etc.) that have helped Apple thrash Intel in the thin laptop/small-form-factor market.
Apple Silicon/ARMs biggest "killer feature" is still computing-power-per-Watt. That's a killer feature in laptops. It's important in SFF (if you want small and quiet). It's also important in server farms/"high density computing"/cloud applications - hence the likes of Amazon's Graviton - but Apple don't have a horse in that race. But on a "big box'o'slots" high-end personal workstation like the Mac Pro, alongside a quartet of power-guzzling discrete GPUs, it may be "nice to have" but it's hardly a killer feature.
Economies of scale are everything with electronics - and Apple have a very healthy slice of the global market for the sort of prosumer laptops and small-form-factor systems that can used what is essentially re-purposed iPhone tech (which Apple sells in even larger quantities). What they've achieved so far is to use the same building blocks to cater for everything up to - well, whatever people are already happily doing with M1 Max & Ultra. The chip you are describing would be just for the replacement for the 2019 Mac Pro which is not a big seller and only significant in the "people who are wedded to Mac-only software" market.
I'm not saying that Apple won't produce a workstation-class Apple Silicon chip, but to me it sounds like a vanity project, and I'm half expecting the "new Mac Pro" to be, essentially, a rackmount version of the Mac Studio (with Mx Extreme processor if they can join 4 Maxes together without mounting two of them in hyperspace).
In any case, the desktop market is being eaten away by more powerful laptops at one end, faster external peripherals via TB4/USB4 in the middle and on-demand cloud computing at the high end. It's not going away any time soon, but it's hardly a growth area to invest in.
...because Apple are once more selling a choice of external displays (possibly with more in the pipeline) and a much improved range of headless desktops. A Mac Studio Max/Studio Display combo costs about the same as a top-end i9 iMac with 32GB RAM - and has comparable performance - while a Studio Ultra/Display combo is a lot cheaper than a comparable 18 core iMac Pro was. For the sake of two extra cables you get the choice of two Apple displays (SD or XDR), a plethora of 3rd party options and the opportunity of replacing/upgrading the computer and display independently.
Plus, a lot of people who previously had a MBP for the road and an iMac on their desk for power are now going to realise that a Mx Max doesn't suffer much from being in a laptop and they don't need a separate desktop any more. Hence the Studio Display, which actually makes more sense as a MacBook docking station than as a display for a desktop Mac.
Maybe we'll see a new 'large' iMac if/when there's something better than a warmed-over version of the 2017 5k panel to put in it. I mean, its great that the 2017 panel is still so good 5 years later, but I don't see it staying top of the heap much longer and wouldn't want to have one welded into a new computer today.
I have mixed feelings about the move to AS. While AS is quick and low power its inability to natively run x64 software is, for me, a significant drawback.
Yup, and just think of all the legacy software that will never be ported over. Emulation can only get you so far...
Emulation gets you most of the way there, and it's significantly faster than the Intel it replaced so it's not that burdensome.
From a pro standpoint, the sticking points are things like plugins and the like that require native version, but having a legacy system around for X random piece of legacy software is I feel an age-old tradition. My old Red Giant plugins have finally broken and I ain't paying Maxon an extra $300 a month for the privilege of using them again
I agree it's hard to see them designing a very different product than 96% of their product offerings, but by the same token, they don't have to worry about software writers ignoring Metal and such, because by that same logic most of their audience isn't going to be on the high-end niche product. They're on their MBPs and Mac Studios like most others in the market.
A much bigger risk is developers just don't make their software for Mac at all, not that they'd only target a niche audience.
it could have say just 4 graphics cores. Enough for a remedial built in display. The high end versions may have say 24 cores, while a more laptop Friendly version 12.
Like a max in some way, different in others. A max will have say 24graohics cores, and have limited pci lanes, where this could have more pci lanes and less graphics cores but beefier different and beefier mix of cpu cores.
I think we were talking at cross purposes a bit. I agree that a 'pro' version of the SoC could be basically all-CPU, with its interface bandwidth used for PCIe rather than talking to GPU cores.
This would be fine for the MP, but I'm still confused about how the economies of scale could work, in terms of using it in other machines. One of these chips + a worthwhile dGPU (it would need to be better than a Max's) would surely add up to greater TDP than Apple would be comfortable with in a mini or MBP?
It could work in a 27" iMac, but only if Apple don't go as thin with that one as they did with the 24". Though it's likely moot anyway, given the mini / Studio + Studio Display appears to have replaced the 27" model.
I mentioned the machines it could work in. Mac Pro (tiny niche), Studio (say 3% of volume? With slightly scaled don chip), 27” iMac (good percent, say 5%? with more scaled down version of chip), then maybe an top end macbook pro with discrete graphics chip and even more scaled down version of the chip (say 5%?). The only wild card is if apple makes its own rack version of the Mac pro, they could then be their own largest customer if they eat their own dog food. Perhaps they can also throw in a top end version of the Mac mini that has a scaled-down version of the trip? Maybe that adds one or 2%?
So that adds up only till about 15% volume, which does support your point of view that it may not really achieve substantial economies of scale. It’s hard for me to know. but the margins on all of those products are so much higher, I bet, 15% of volume may make up 30 to 40% of all the margin. That makes it valuable.
And look at Apple Watch. What overall percentage does it make up with regard to iPhones iPads Mac? Or the custom chips in their headphones? Just because it doesn’t enjoy the same insane scale say the iPhone, doesn’t mean it’s beyond Apple to make.
as for the TDP of the Mac mini, I really don’t think they care. I could be wrong, and you may know something I don’t there. For the MacBook Pro, of course you are right, it matters. But they were able to make it work with Intel’s insane ships and discrete graphics cards, I think even a desktop version of apples chip will be more efficient than what Intel does. My 2019 maxed out MacBook Pro is insanely hot, and the scary thing is that it was cooler than some of my previous MacBook Pros!
Would this have more CPU than a Max? And how much more powerful would the dGPU be than the one integrated into the Max?
I suppose it comes down to how much clear air there is between the existing Max chip, and the scaled-down versions of the chip you're talking about. If scaled down too much, it would be hard to justify making a separate chip, especially with the Max enjoying close integration and the efficiencies that entails.
I suppose it comes down to how much clear air there is between the existing Max chip, and the scaled-down versions of the chip you're talking about. If scaled down too much, it would be hard to justify making a separate chip, especially with the Max enjoying close integration and the efficiencies that entails.
Hmmm, I'll be honest with you, I'd LOVE for the 8.1 to be a vanity project. Just Apple going balls to the wall showing off Everything that they're capable of...cuz I'll tell ya what, that's a machine I wouldn't be able to wait to buy!
and to be honest, that's kinda what I wish they would do with their Pro/Enthusiast stuff.... sort of like a Skunk Worx type thing going on inside the mothership... you know what I mean?