Insider preview of the 2013 Mac Pro design principles.

[wonderspark]

It's a shame that the nMP was designed to be a disposable can of compromises. It proves that the designers were hog-tied and forced to only offer an inferior connect (Thunderbolt) over PCIe, no internal storage, proprietary GPUs and everything else that is a step backwards to the prior design.

I deplore devices that further the disposable attitude our society finds ourselves deeper entrenched in every year. If Apple truly cared about users, they'd have designed it large enough to include PCIe slots and internal storage. They don't care, and only want to force Thunderbolt on users, as well as get them to buy a new one more often. This is proven by the design.

 

[Rich.Cohen]

There were rumors for months that apple was debating abandoning the pro market because the investment vs payoff of a new redesign was a negative sum.

How was this handled or discussed internally?

Did they give any restrictions what so ever to the design teams?

What was the decision to do dual video instead of dual CPU?

Have you heard any rumblings about a new Cinema Display?

You've asked some excellent questions. I don't have any answers, but I would expect that Jony Ive would insist of a minimum of constraints and that Apple management would agree. That's just a conjecture based on my understanding of how top flight designers work.

 

[flat five]

The irony, like the flying wing from WWII which became the B2, the nMP was initially discussed in Germany and then brought to the USA afterwards. Hence the reason why the public information was published on this thread. History seems to be on a loop tape sometimes.

damn germans!.. always coming up with the good ideas

 

[Rich.Cohen]

Tim: Mac Pro sales are steadily dropping and the folks on MacRumors are starting to think the Mac Pro is dead... What should we do?

Phil: We need something to resurrect our reputation for innovation. Jony, what can we do with this "Truck"?

Jony: <English accent> I just bought a Dyson on the weekend, I think I have an idea...

ROFL

Wonderful!!! That's a lot more believable that the SR-71 argument.

 

[flat five]

Whether the nMP is a design triumph or not is a matter of opinion.

well, yeah.
especially in the context of an internet forum such as this.. i mean i really don't expect you to accept my opinion on the design as being an opinion which is of any importance to be relevant.. i don't expect you to respect my opinion at all and of course i acknowledge the skepticism you may have..

but when/if a multitude of world class designers/engineers/artists/etc begin to express their positive opinions on the design then there should at least be a bit more hesitancy in countering with your opinion of 'it's crap' (or whatever)

----------

I'm pretty sure it went something like this...



Tim: Mac Pro sales are steadily dropping and the folks on MacRumors are starting to think the Mac Pro is dead... What should we do?

Phil: We need something to resurrect our reputation for innovation. Jony, what can we do with this "Truck"?

Jony: <English accent> I just bought a Dyson on the weekend, I think I have an idea...

haha
#

i think you're right.

 

[Rich.Cohen]

the a12, to me, looks to be a purer design based off these two pictures (granted, it could be the difference in lens perspective which is causing me to say this)..

the sr71 looks to have attempted slightly more softend / less angular approach.. it appears as if the gist of the original idea (for lack of better words) has shown through more prominently in the a12 than the sr71

I think the visual difference between the A-12 and SR-71 is the result of lifting body design techniques that were not understood when the A-12 design was created. Essentially they widened the aircraft body to provide more lift and improved flight efficiency. There may well have been other considerations like stability and control, but I would guess those are still classified.

this picture tries to show that everything on an aircraft doesn't have to be so exact/perfect/precise from an engineering standpoint to the extent that a designer has zero freedom of how the plane actually looks or is shaped

Good engineering design with a margin of safety, thank God.

i'm not quite sure people who have spoken so negatively etc in the thread up to now will actually admit what it is that they like so much about the sr71.. or why it was the most awesome thing in the world to them when they were 8years old.. but i truly (yes, truly) believe it's because of how it looks 1st & foremost.. the color/shape/sleekness/etc..
the specs (altitude/speed/etc) comes in second and further adds to the appreciation of step1.. the actual engineering aspects come some time after and are a much less important factor in causing someone to say "wow, that's amazing"

I was about 24 when McNamara made the first announcement. At that time I was an Army officer involved with anti-aircraft missiles. I was a subscriber and reader of Aviation Week so I got a lot of details, not all correct. I was initially impressed by all the thinks you mentioned. Over the years I picked up more details and became even more impressed. Another reason is that I'm enamored by "inholdings" organizations that can do things impossible for their parent organization. The Lockheed Skunkworks is an outstanding example of an "inholding" and I'm a fan. DARPA is also an inholding and I'm a fan.

 

[flat five]

Did they give any restrictions what so ever to the design teams?

i'd be willing to guess one of the restrictions placed on the team was that no matter what, the final product upon being in stores is to cost no more than $3000

i say this mainly because of them selling 3 sticks of ram in the base config.. i can almost imagine them nearing the finalization of the project going "crikey!.. after the accountants ran all this stuff through their calculators, we're looking at $3100.. something has to make the website show $2999.99 on launch day.. upper mgmt refuses to eat that hundred"
then
"well, i guess we could take out a stick of ram?"

but i doubt (highly) that the actual design team had it in their mind to create four ram slots except leave a single one empty.

 

[Rich.Cohen]

It's a shame that the nMP was designed to be a disposable can of compromises. It proves that the designers were hog-tied and forced to only offer an inferior connect (Thunderbolt) over PCIe, no internal storage, proprietary GPUs and everything else that is a step backwards to the prior design.

You're welcome to your opinion about the nMP, but it's design proves nothing about the constraints imposed on the engineers. It may well be that the engineers sold this design to management as something "insanely great."

 

[handsome pete]

It's a shame that the nMP was designed to be a disposable can of compromises. It proves that the designers were hog-tied and forced to only offer an inferior connect (Thunderbolt) over PCIe, no internal storage, proprietary GPUs and everything else that is a step backwards to the prior design.

I deplore devices that further the disposable attitude our society finds ourselves deeper entrenched in every year. If Apple truly cared about users, they'd have designed it large enough to include PCIe slots and internal storage. They don't care, and only want to force Thunderbolt on users, as well as get them to buy a new one more often. This is proven by the design.

I think the notion of disposability regarding the new Mac Pro is mostly nonsense. Of all the Mac Pros that came before it, what percentage of them do you honestly think were given extended life through GPU upgrades? That's really the only upgrade that lends any validity to the disposable argument. And while it's purely anecdotal, in all of the shops/studios I've been through, I've rarely seen any of the older Mac Pros tinkered with beyond the occasional RAM upgrade. Those of us who do more to extend the life of the machine are just a small percentage.

I think the gripes people have with its limitations make more sense. And I agree with your stance on the consumerism aspect, but I don't think the new Mac Pro fits here.

 

[VirtualRain]

ROFL

Wonderful!!! That's a lot more believable that the SR-71 argument.

haha
��

i think you're right.

Consider the benefits of this accessory...
- Forced air cooling with a 15AMP motor
- HEPA Filter to eliminate dust build up
- Roll your Mac Pro anywhere with ease

If you're willing to put up with a bit of added noise, the cooling benefits could be enormous.

 

[goMac]

I think the notion of disposability regarding the new Mac Pro is mostly nonsense. Of all the Mac Pros that came before it, what percentage of them do you honestly think were given extended life through GPU upgrades? That's really the only upgrade that lends any validity to the disposable argument. And while it's purely anecdotal, in all of the shops/studios I've been through, I've rarely seen any of the older Mac Pros tinkered with beyond the occasional RAM upgrade. Those of us who do more to extend the life of the machine are just a small percentage.

Yep, I've seen that a lot too. Most Mac Pros I've seen have never been upgraded beyond RAM, maybe another disk.

Most people (likely outside of these forums) buy Mac Pros because they think they will last a long time, and they're towers.

There were rumors for months that apple was debating abandoning the pro market because the investment vs payoff of a new redesign was a negative sum.

I'll bite...

From everything I heard, I don't think this was ever really a risk. I think it was talked about, but probably not for very long.

What was the decision to do dual video instead of dual CPU?

Dual video is the future, dual CPU is not. Dual video cards with OpenCL are much more suitable and faster for pro tasks than a fast CPU. You can get more milage out of FCPX with dual video cards than you can with dual CPUs. That's an easy one.

Yes, some apps are still CPU bound, but Apple cares about their own apps more than they care about third party apps. And Apple's apps are all OpenCL and OpenGL based.

Apple has also found that running OpenCL and OpenGL off of the same card hurts performance, so if you have two cards, you can run OpenCL off of one, and OpenGL off of the other, and bypass that bottleneck.

 

[Derpage]

All ye consumed by rational thought: Do not tread amongst these thread dwellers. Tis a trap.

 

[ybz90]

All ye consumed by rational thought: Do not tread amongst these thread dwellers. Tis a trap.

+1. You can't teach those who don't wish to listen.

 

[flat five]

All ye consumed by rational thought: Do not tread amongst these thread dwellers. Tis a trap.

rational thought is an important step.. but there are other steps which can lead to a more beneficial exchange of info/ideas between all involved.

i feel a flippant remark like that is possibly a breakdown of step1 or more likely step2 .. it would be helpful if you could repeat back to the OP how you interpret what he's saying.. just to make sure you two are actually talking about the same thing..

if that never occurs, then rational thought/evaluation has never truly occurred.
(see step6)

if it's more of a stance of "i really don't care what these looneys are talking about" then that's cool too.. it would just be helpful to others if you would clarify that as your stance. (or just ignore the thread in the first place)
you see?

Step 1: Knowledge
In terms of critical thinking, the basic level of acquisition of knowledge requires that you be able to identify what is being said: the topic, the issue, the thesis, and the main points.

Step 2: Comprehension
Comprehension means understanding the material read, heard or seen. In comprehending, you make the new knowledge that you have acquired your own by relating it to what you already know. The better you are involved with the information, the better you will comprehend it. As always, the primary test of whether you have comprehended something is whether you can put what you have read or heard into your own words. Remember that comprehending something implies that you can go beyond merely parroting the material back but instead that you can give the material your own significance.

Step 3: Application
Application requires that you know what you have read, heard, or seen, that you comprehend it, and that you carry out some task to apply what you comprehend to an actual situation.

Step 4: Analysis
Analysis involves breaking what you read or hear into its component parts, in order to make clear how the ideas are ordered, related, or connected to other ideas. Analysis deals with both form and content.

Step 5: Synthesis
Synthesis involves the ability to put together the parts you analyzed with other information to create something original.

Step 6: Evaluation
Evaluation occurs once we have understood and analyzed what is said or written and the reasons offered to support it. Then we can appraise this information in order to decide whether you can give or withhold belief, and whether or not to take a particular action. Never put evaluation ahead of the other steps in critical thinking steps; otherwise, you will be guilty of a "rush to judgement." When emotion substitutes for reasons, evaluation incorrectly precedes analysis.

 

[Rich.Cohen]

Most people (likely outside of these forums) buy Mac Pros because they think they will last a long time, and they're towers.

HEY, I'm one of those people! I was thinking of replacing my Mac Pro 1,1 two years ago, but I didn't like what I saw in the summer of 2012 so I've been hanging on waiting for this round. I have the ability to do an upgrade, but I other things I'd prefer to do in my free time. All I've ever done to my Mac is add memory and additional Hard Disks.

Apple has also found that running OpenCL and OpenGL off of the same card hurts performance, so if you have two cards, you can run OpenCL off of one, and OpenGL off of the other, and bypass that bottleneck.

That's interesting to know.

 

[Rich.Cohen]

rational thought is an important step..

Re: Knowledge, comprehension, application, etc.

That sounds a lot like what I was taught in graduate school studying system engineering.

1. Collect information (Knowledge)
2. Build a model or models to describe what you've collected (Comprehension)
3. Use the models to better understand the problem (Analysis)
4. Develop a set of alternative solutions
5. Evaluate those alternatives using the models (Evaluation)
6. Synthesize a better solution from all the alternatives (TRADE-OFFS) (Synthesis)
7. Iterate steps 4-6 until you feel you've got a good solution. (Too many leave this step out.)
8. Develop a plan to implement your optimal solution.

Of course in the real world these "steps" don't happen sequentially and one doesn't stop because the next has started. You're continually getting new knowledge that may affect everything else. At any time you may have an insight for a better more useful model. At any time you may think of an alternative you overlooked.

Has anyone here read the wonderful paper by Dave Parnas titled A Rational Design Process: How and why to fake it? I highly recommend it.

 

[flat five]

Re: Knowledge, comprehension, application, etc.

That sounds a lot like what I was taught in graduate school studying system engineering.

1. Collect information (Knowledge)
2. Build a model or models to describe what you've collected (Comprehension)
3. Use the models to better understand the problem (Analysis)
4. Develop a set of alternative solutions
5. Evaluate those alternatives using the models (Evaluation)
6. Synthesize a better solution from all the alternatives (TRADE-OFFS) (Synthesis)
7. Iterate steps 4-6 until you feel you've got a good solution. (Too many leave this step out.)
8. Develop a plan to implement your optimal solution.

Of course in the real world these "steps" don't happen sequentially and one doesn't stop because the next has started. You're continually getting new knowledge that may affect everything else. At any time you may have an insight for a better more useful model. At any time you may think of an alternative you overlooked.

Has anyone here read the wonderful paper by Dave Parnas titled A Rational Design Process: How and why to fake it? I highly recommend it.

heh, i was going to pick up another one you recommended a few pages back..
exploring requirement- quality before design

sell me on one or the other as it's more likely going to be one that i read instead of two.

[edit] oh wait.. the new recommendation is a paper.. maybe i can do both(?)

 

[Rich.Cohen]

heh, i was going to pick up another one you recommended a few pages back..
exploring requirement- quality before design

I won't try to sell you on anything. I will say that I've read that book more than once and I've taught courses and lead study groups that used it with excellent results.

In fairness, I must say that I'm a groupie of Gerry Weinberg, one of the authors. I've read and enjoyed nearly every book he's written and I've met him on a number of occasions.

oh wait.. the new recommendation is a paper.. maybe i can do both(?)

It's free on the internet. Only about 20 pages as I recall. Really worth reading. I'd love to learn what you think of it.

 

[MiJuConcept]

You're welcome to your opinion about the nMP, but it's design proves nothing about the constraints imposed on the engineers. It may well be that the engineers sold this design to management as something "insanely great."

That is precisely what document one was pitching at
- difficult
- expensive
- radical
- iconic

I would rather Apple took to heart a great challenge and bother the elite user than scrap the entire family.

 

[MiJuConcept]

Ok so I'm being sucked into this one: I have a few questions about the project itself and Apples philosophy and approach to the "traditional" desktop space.

We all know Apple has shifted focus to their baby, the iphone, and iOS. They are now expanding with peripherals all based off this new direction with and iPad and this new "wearable" or watch probably coming out next fall.

There were rumors for months that apple was debating abandoning the pro market because the investment vs payoff of a new redesign was a negative sum.

How was this handled or discussed internally?

Did they give any restrictions what so ever to the design teams?

What was the decision to do dual video instead of dual CPU?

Have you heard any rumblings about a new Cinema Display?

The initial document referenced family history, titanium laptops and iconic forms of speed and power. Apart from the square central core becoming triangular, not much was changed. The prediction of cost and difficulty was entirely accurate and it would seem that throwing down a huge gauntlet gets people thinking.

Having looked at the video once more of Sir John speaking about his desire to break with the past (Nov 2012) I think he linked styling and dimension to past. Not necessarily ignoring the cube or the will to push a computer to the edge of being commercially unviable. As Apple has made that push many times before and lost a few times.

When I think of the team behind the nMP I can see they were bold and fearless in a flat global economy, at a time when so many are cutting costs and aiming for survival.

 

[ybz90]

When I think of the team behind the nMP I can see they were bold and fearless in a flat global economy, at a time when so many are cutting costs and aiming for survival.

Like most of your other hyperboles, I think you're being way too dramatic here.

In many ways, this isn't the first time Apple has tried this -- you could argue its spiritual predecessor is the PowerMac G4 Cube. Apple has been know to be ahead of its time, and when those products fail, they compartmentalize the good ideas and iterate on them to improve their shortcomings in the future. Another example might be the Newton.

 

[MiJuConcept]

Re: Knowledge, comprehension, application, etc.

That sounds a lot like what I was taught in graduate school studying system engineering.

1. Collect information (Knowledge)
2. Build a model or models to describe what you've collected (Comprehension)
3. Use the models to better understand the problem (Analysis)
4. Develop a set of alternative solutions
5. Evaluate those alternatives using the models (Evaluation)
6. Synthesize a better solution from all the alternatives (TRADE-OFFS) (Synthesis)
7. Iterate steps 4-6 until you feel you've got a good solution. (Too many leave this step out.)
8. Develop a plan to implement your optimal solution.

Of course in the real world these "steps" don't happen sequentially and one doesn't stop because the next has started. You're continually getting new knowledge that may affect everything else. At any time you may have an insight for a better more useful model. At any time you may think of an alternative you overlooked.

Has anyone here read the wonderful paper by Dave Parnas titled A Rational Design Process: How and why to fake it? I highly recommend it.

This is robust and useful ... but where is the emotional nonsense.

 

[Nugget]

i'd be willing to guess one of the restrictions placed on the team was that no matter what, the final product upon being in stores is to cost no more than $3000. i say this mainly because of them selling 3 sticks of ram in the base config..

I've given you a lot of grief on the forums lately, and I just have to say that I think you're spot-on 100% right about this.

 

[MiJuConcept]

I've given you a lot of grief on the forums lately, and I just have to say that I think you're spot-on 100% right about this.

We can safely assume that was one of the restrictions placed on the suppliers.

I read a very fine article interviewing the outgoing CEO of Intel and he regretted not dropping the price of the Intel CPU so that Apple would use it in the iPhone. Not only did Intel get their sums wrong by a volume factor of 1000 but so did apple by a factor of 100.

People are pretty reluctant to make those kinds of mistakes in 2013. Especially when Apple is pulling the big numbers. I think that is my way of saying that Apple knows how to price components.

 

[Rich.Cohen]

We can safely assume that was one of the restrictions placed on the suppliers.

I read a very fine article interviewing the outgoing CEO of Intel and he regretted not dropping the price of the Intel CPU so that Apple would use it in the iPhone. Not only did Intel get their sums wrong by a volume factor of 1000 but so did apple by a factor of 100.

People are pretty reluctant to make those kinds of mistakes in 2013. Especially when Apple is pulling the big numbers. I think that is my way of saying that Apple knows how to price components.

Years ago at the beginning of color television there were two competing approaches, one from RCA and one from CBS. The RCA approach was backward compatible with existing black and white sets, but the CBS approach was technically superior. The FCC told CBS they would go with the CBS approach if CBS would buy up all the existing black and white sets. CBS said no way and the FCC went with the RCA approach. Years later someone at CBS did a study and found they would have made a huge profit if they had accepted the FCC offer. Large corporations tend to be very risk adverse when faced with an unknown market.

As I recall the FedEx idea started as a term paper. The concept couldn't start piecemeal. It needed a sorting hub, a small fleet of Falcon jets and destination sites and all major cities. Getting the start-up funds was very difficult because it was an unknown market and investors were very reluctant to put up money.