New Mac Pro at WWDC June 2016

[MacVidCards]

The reality distortion field is powerful...

Apple & AMD PR team have really been working overtime. And such an uphill battle, too.

The logical extension of their argument is that a couple of old 7970s (intro in 2011) and a 2 gen old Xeon in a 40oz Al extrusion is still worth $10K today.

The absurd "powerful workstationette that only uses 450 Watts" lead to the custom PCBs. The custom PCBs lead to the $10K bad joke Apple offers today. No way to decouple.

And yet the fervent excuses for a tech company that could design and fund a Mars mission if they wanted.

 

[MarkusL]

Apple & AMD PR team have really been working overtime. And such an uphill battle, too.

The logical extension of their argument is that a couple of old 7970s (intro in 2011) and a 2 gen old Xeon in a 40oz Al extrusion is still worth $10K today.

No, the thing that makes it worth $10K is the huge advantage of OSX. And that advantage gets greater as the case gets rounder... Or something.

 

[s-hatland]

are we married to "nnmp"? what are we going to call the 3rd gen? nnnmp? how many "n"s can we really add here? i feel like nmp2 is easier to understand. what's that? more important points of discussion? ahhh... well then, carry on!

 

[AidenShaw]

are we married to "nnmp"? what are we going to call the 3rd gen? nnnmp? how many "n"s can we really add here? i feel like nmp2 is easier to understand. what's that? more important points of discussion? ahhh... well then, carry on!

If it's ever updated, whatever Apple calls it will be one option. Of course, MP7,1 will also work if Apple holds to form for the unambiguous name.

 

[Larry-K]

That's one of my bigger questions since nMP was launched. Why do they have to stick with this small case and thermal envelope that forces internals to be downclocked ? Isn't the thermal envelope supposed to serve the internals and not the other way around ?

You're not applying Apple's brand of "Magical Thinking" to the problem.

There's no Louis Sullivan at Apple.

 

[flat five]

are we married to "nnmp"? what are we going to call the 3rd gen? nnnmp? how many "n"s can we really add here? i feel like nmp2 is easier to understand. what's that? more important points of discussion? ahhh... well then, carry on!

mac pro.. or mp.
mp6... mp7.. mp5.. etc

(edit) nvrmnd.. what Aiden said
[doublepost=1463442490][/doublepost]

And yet the fervent excuses for a tech company that could design and fund a Mars mission if they wanted.

some people just happen to like the computer.. is that so hard to understand? why if someone likes the thing are they then 'making excuses'?

to me, the nuts excuses come from the people who, day-in-day-out sit around here complaining about macs.. buy the HP & move on.. what's the excuse for not doing that? OS X? gimme a break.

why sit around here complaining all day long? why?
answer that one, then read the answer.. then see the absurdity in the excuse.

 

[Melodist]

mac pro.. or mp.
mp6... mp7.. mp5.. etc

(edit) nvrmnd.. what Aiden said
[doublepost=1463442490][/doublepost]

some people just happen to like the computer.. is that so hard to understand? why if someone likes the thing are they then 'making excuses'?

to me, the nuts excuses come from the people who, day-in-day-out sit around here complaining about macs.. buy the HP & move on.. what's the excuse for not doing that? OS X? gimme a break.

why sit around here complaining all day long? why?
answer that one, then read the answer.. then see the absurdity in the excuse.

Okay, so you didn't buy the Mac Pro for the OS, but for the hardware specs? Or was it the design? Tell me.

To me, people buying a Mac not for the OS is nuts. Because otherwise, why would you even buy one nowadays? Let's take the fashion aspect and decadence aside.

 

[flat five]

nowadays? the design.. typically the higher spec'd versions of a design i like.
so- design then specs.

i like os x too.
but..

aren't you on and on about adobe? or is it avid?

it doesn't matter if you have windows or os x when you're using the computer.. the apps are the same. they do the same stuff, your content is the same, you can switch between the same application on windows or os x with barely a hiccup.

the real stuff.. the important stuff.. the productivity stuff.. it doesn't matter at all osx vs windows.. the applications are what matter.. and your skill with those applications are what matter.

whether or not you like finder or start button or launchpad or stoplight graphics is all second tier importance..

definitely ---> definitely not of such importance that you should be buying hardware you dislike --or worse-- as the claims often made around here, that the hardware is too slow for production environments..

"i buy hardware that i don't like, that's too old and slow.. because i like how the icons animate in os x".. is absurd.

you should buy the hardware you like and you should buy the applications you like and you should buy hardware that's capable of running those applications.

the equation here is incredibly simply.. there are only two factors at play.
hardware and applications.

if those two items work well together and you like both of them then congratulations, you have a good basis for enjoyable productivity.

if not, then you're doing something wrong.. try something else.. i'm approx 100% positive the right combo is out there.
for every single person on the planet.

Let's take the fashion aspect and decadence aside.

design encompasses more than fashion or decadence.

good design generally takes on those traits.. but it doesn't work the other way.. just because something is fashionable or decadent does not mean it's well designed.

 

[tuxon86]

nowadays? the design.. typically the higher spec'd versions of a design i like.
so- design then specs.

i like os x too.
but..

aren't you on and on about adobe? or is it avid?

it doesn't matter if you have windows or os x when you're using the computer.. the apps are the same. they do the same stuff, your content is the same, you can switch between the same application on windows or os x with barely a hiccup.

the real stuff.. the important stuff.. the productivity stuff.. it doesn't matter at all osx vs windows.. the applications are what matter.. and your skill with those applications are what matter.

whether or not you like finder or start button or launchpad or stoplight graphics is all second tier importance..

definitely ---> definitely not of such importance that you should be buying hardware you dislike --or worse-- as the claims often made around here, that the hardware is too slow for production environments..

"i buy hardware that i don't like, that's too old and slow.. because i like how the icons animate in os x".. is absurd.

you should buy the hardware you like and you should buy the applications you like and you should buy hardware that's capable of running those applications.

the equation here is incredibly simply.. there are only two factors at play.
hardware and applications.

if those two items work well together and you like both of them then congratulations, you have a good basis for enjoyable productivity.

if not, then you're doing something wrong.. try something else.. i'm approx 100% positive the right combo is out there.
for every single person on the planet.



design encompasses more than fashion or decadence.

good design generally takes on those traits.. but it doesn't work the other way.. just because something is fashionable or decadent does not mean it's well designed.

And yet, in the end you didn't buy one...

 

[flat five]

And yet, in the end you didn't buy one...

and?

I probably won't anymore. cloud computing puts owning a dedicated rendering station at a disadvantage.

 

[lowendlinux]

Maybe they'll make is square??

Nothing on my desk is tube shaped maybe we'll get a genuine cube replacement

 

[Santabean2000]

That would be interesting, but I cannot imagine that happening if we see new GPUs and CPUs.

It was a jab at their record with the mini.

 

[tuxon86]

and?

I probably won't anymore. cloud computing puts owning a dedicated rendering station at a disadvantage.

Then the design wasn't enough to get you to buy one...

 

[flat five]

Then the design wasn't enough to get you to buy one...

huh?
I like a lot of different designs.. doesn't mean I should buy all of them.

computer wise, I have an iMac and mbp.
the laptop is well built.. I know people like to say apple stuff is just shiny etc but mine is scratched and dented from being used in a shop environment. it's well designed.

the iMac is sweet looking. if it were just a display and I owned a nmp, I'd probably buy this iMac looking display. so now, I have a display I like and no other computer type module on my desk? I'll take it.. again it's designed well and to my liking.

this isn't really to do with my earlier point though.

 

[beaker7]

and?

cloud computing puts owning a dedicated rendering station at a disadvantage.

Thats why you have both.

 

[flat five]

Thats why you have both.

maybe.. maybe not.

idk, i'm pretty sure a lot of people don't quite grasp the increases available via cloud.

example.. today: like a few hours ago..
client wanted to see different proportions on a fixture.. i did the readjustments in rhino then sent to a360..

"Render Time: a few seconds" .. or the larger one "a minute"..
as in, i'm sending sets of rendered design changes in minutes.. granted, these are mainly 1000px wide images for emailing but they are raytraced at high/final quality.

compare to a few years ago.. in this type of situation, client is getting screenshots from a 3D modeling application as opposed to photoreal.. if i was to be doing any renderings for them, at best, it would be "ok, i'll send you some tomorrow" since i'd need them to cook overnight..

some of these renders are completed faster than i can set up the next view.. they're not eating into my computer's resources so i can continue to work smoothly on the next view.. or- i'm just not waiting on renders to complete any more.
it's a complete night&day difference from my workflow a couple of years ago.
[edit] or 6 months ago even [/edit]


---
also - the computer i was using was my laptop..
also - internet was via tethered phone.

also- these renders (16 of them today) didn't cost me anything

---
idk, you're going to need some very convincing arguments to make me believe i'll benefit by having a dedicated rendering station.. i mean, i could be sitting at a 24core computer and it'd still make sense for me to push the cloud render button instead of the local render button since it's going to be faster and ready for collaboration faster than local.. not to mention the amount of renders i do while away from that particular workstation..

(and for clarity-- i'm talking about me and my own experiences.. i'm not wanting to imply what's right for me is right for everybody)

 

[flat five]

meh.
too far off topic.
*post deleted.
: )

 

[JesterJJZ]

Cloud rendering might be fine for 3d applications where you basically send digital blueprints that get rendered on a server farm, but for things like compositing, VFX and color correction the bandwidth isn't there. An average project I work on is in the neighborhood of 750GB-2TB worth of 4-6k footage, graphic assets and image and sound files. I need these files accessed locally on a fast array and be working with a powerful online system.

I also don't like being at the mercy of another party to get my work done. For guys like us, the only thing we use the cloud for is sharing files and sending dailies and wip edits.

 

[tuxon86]

Cloud rendering might be fine for 3d applications where you basically send digital blueprints that get rendered on a server farm, but for things like compositing, VFX and color correction the bandwidth isn't there. An average project I work on is in the neighborhood of 750GB-2TB worth of 4-6k footage, graphic assets and image and sound files. I need these files accessed locally on a fast array and be working with a powerful online system.

I also don't like being at the mercy of another party to get my work done. For guys like us, the only thing we use the cloud for is sharing files and sending dailies and wip edits.

For security and inetellectual property protection reason we can't send any file to an outside cloud where I work.
But we do manage a network that rival the biggest telco in the country internally. Even then, it is way more practical to set up local workstation and storage in the remote location for manipulating big dataset.

 

[beaker7]

maybe.. maybe not.

idk, i'm pretty sure a lot of people don't quite grasp the increases available via cloud.

example.. today: like a few hours ago..
client wanted to see different proportions on a fixture.. i did the readjustments in rhino then sent to a360..

View attachment 631792

View attachment 631793

View attachment 631794

"Render Time: a few seconds" .. or the larger one "a minute"..
as in, i'm sending sets of rendered design changes in minutes.. granted, these are mainly 1000px wide images for emailing but they are raytraced at high/final quality.

compare to a few years ago.. in this type of situation, client is getting screenshots from a 3D modeling application as opposed to photoreal.. if i was to be doing any renderings for them, at best, it would be "ok, i'll send you some tomorrow" since i'd need them to cook overnight..

some of these renders are completed faster than i can set up the next view.. they're not eating into my computer's resources so i can continue to work smoothly on the next view.. or- i'm just not waiting on renders to complete any more.
it's a complete night&day difference from my workflow a couple of years ago.
[edit] or 6 months ago even [/edit]


---
also - the computer i was using was my laptop..
also - internet was via tethered phone.

also- these renders (16 of them today) didn't cost me anything

---
idk, you're going to need some very convincing arguments to make me believe i'll benefit by having a dedicated rendering station.. i mean, i could be sitting at a 24core computer and it'd still make sense for me to push the cloud render button instead of the local render button since it's going to be faster and ready for collaboration faster than local.. not to mention the amount of renders i do while away from that particular workstation..

(and for clarity-- i'm talking about me and my own experiences.. i'm not wanting to imply what's right for me is right for everybody)

Who gives you cloud rendering for free?

And why aren't you doing this on an Apple Watch? That's what all the cool kids are doing.

At my company we've investigated cloud extensively and found it to be, at best, equal cost to having our own farm. Not that it matters as for security's sake we won't send design data to someone else's servers.

 

[Hank Carter]

For security and inetellectual property protection reason we can't send any file to an outside cloud where I work.
But we do manage a network that rival the biggest telco in the country internally. Even then, it is way more practical to set up local workstation and storage in the remote location for manipulating big dataset.

Same here. Rendering or storing data in the cloud is not an option due to confidentiality reasons. So it's a mute point from the beginning.

Also the sheer data size we deal with at work makes it impossible. We would have to upload hundreds of gigabytes or several terabytes of data every day over the Internet and that's not going to happen anytime soon.

And then there is cost. Over the course of a big project we could buy the hardware and with a second one the infrastructure needed to support it. Cloud computing is not free or cheap.

 

[flat five]

Who gives you cloud rendering for free?

And why aren't you doing this on an Apple Watch? That's what all the cool kids are doing.

At my company we've investigated cloud extensively and found it to be, at best, equal cost to having our own farm. Not that it matters as for security's sake we won't send design data to someone else's servers.

the free bit is misleading. it's only free for now though Autodesk does offer a generous trial period of 1 year. after that, cloud rendering remains free for hobbyist/casual user or $1-2 per render commercial. (though previews and smaller sized images will remain free of charge)

 

[rockyromero]

And then there is cost. Over the course of a big project we could buy the hardware and with a second one the infrastructure needed to support it. Cloud computing is not free or cheap.

I'm starting to understand cloud rendering as an alternate solution.

It's not clear or obvious to me as possible for my current application.

Cost is a factor, inevitably.

My sunk cost in existing hardware, software, and time invested keeps me on the direction of adding more computing power internally.

The level of scaling my computing abilities is proportional to the ability of obtaining more clients.

From a business standpoint, the greater my computing then the lower the cost of client acquisition.

With each additional nMP 12 core, conceivably, my client base would triple immediately.

There are other factors involved and there would be continuous dimishing returns.

Gauging resource requirements for business is an interesting exercise.

 

[flat five]

Gauging resource requirements for business is an interesting exercise.

(in my experience)

cloud rendering has made it to the point (in ease of use, availability, security, cost, reliability, speed) to where it's pretty much a no-brainer for the freelancer, sole-proprietor, or small firm..

it's virtually impossible (and certainly impractical) for a small business to acquire local resources which will match or even come anywhere close to what's available on cloud.

that said, with improvements in GPUs and gpgpu (which are now becoming available), an individual can see incredible speedups on modest personal computers while keeping everything local.. maybe another couple of years to where this, or similar options, is more widespread. but you can pioneer this avenue with software that's available today.


idk- i guess my point is multiple cpus and/or high core count cpus is nearly a dead end street for small business looking to improve their render processing power.. it's just far too expensive for the relatively minute boosts in speed compared to other options.. (like- doubling the cost of a computer in order to double the speed of rendering is a bad buy when comparing to spending a few hundred dollars for 20x + speed gains)

 

[AidenShaw]

(in my experience)

cloud rendering has made it to the point (in ease of use, availability, security, cost, reliability, speed) to where it's pretty much a no-brainer for the freelancer, sole-proprietor, or small firm..

it's virtually impossible (and certainly impractical) for a small business to acquire local resources which will match or even come anywhere close to what's available on cloud.

that said, with improvements in GPUs and gpgpu (which are now becoming available), an individual can see incredible speedups on modest personal computers while keeping everything local.. maybe another couple of years to where this, or similar options, is more widespread. but you can pioneer this avenue with software that's available today.


idk- i guess my point is multiple cpus and/or high core count cpus is nearly a dead end street for small business looking to improve their render processing power.. it's just far too expensive for the relatively minute boosts in speed compared to other options.. (like- doubling the cost of a computer in order to double the speed of rendering is a bad buy when comparing to spending a few hundred dollars for 20x + speed gains)

What cloud stuff do you do other than the builtin cloud support in Autodesk?