What do you think is coming up first after March 7th? In which order?

[Jst0rm]

intel may go on a 3 year schedule out past haswell somewhere. We are coming close to the end of moores law and need something different to keep moving forward at the same rates.

When else in the history of computing have you been able to hold onto a computer for so long?

 

[derbothaus]

intel may go on a 3 year schedule out past haswell somewhere. We are coming close to the end of moores law and need something different to keep moving forward at the same rates.

When else in the history of computing have you been able to hold onto a computer for so long?

When else have we had so little forward movement in terms of software? Most HW is still wasted thanks to code.

 

[Jst0rm]

When else have we had so little forward movement in terms of software? Most HW is still wasted thanks to code.

I think its been difficult to truly code things for multiple processors. It has certainly been a challenge. Hopefully things will get better in this regard but it is certainly lagging the hardware side.

The current conundrum is we are not able to really hit over 4ghz because of the exponential growth on power needed to hit that point and lots of software craves raw ghz.

 

[goMac]

I think its been difficult to truly code things for multiple processors. It has certainly been a challenge. Hopefully things will get better in this regard but it is certainly lagging the hardware side.

It's not hard. A better question is "Why?" If a single core is fast enough to run iChat well, why go back and rewrite iChat for multicore?

We're seeing apps that actually need the speed (FCPX, Premiere, Logic, After Effects) get rewritten for multicore. Apps that don't need multicore don't bother.

(The one exception is Photoshop, but Adobe couldn't code their way out of a cardboard box on that one. The iOS version of Photoshop is written in Flash. Enough said.)

 

[Jst0rm]

It's not hard. A better question is "Why?" If a single core is fast enough to run iChat well, why go back and rewrite iChat for multicore?

We're seeing apps that actually need the speed (FCPX, Premiere, Logic, After Effects) get rewritten for multicore. Apps that don't need multicore don't bother.

(The one exception is Photoshop, but Adobe couldn't code their way out of a cardboard box on that one. The iOS version of Photoshop is written in Flash. Enough said.)

There are a few exceptions and usually they are complex software systems with tons of legacy code.

 

[goMac]

There are a few exceptions and usually they are complex software systems with tons of legacy code.

Yeah, it depends how legacy. Heavy engineering and science apps have been running on multicore since the early nineties, so some legacy apps have actually been multicore this entire time.

UNIX mainframes with multicore have been pretty common the last few decades.

 

[derbothaus]

It's not hard. A better question is "Why?" If a single core is fast enough to run iChat well, why go back and rewrite iChat for multicore?

For the exact reasons given. iChat does not run in a vacuum. This IS what SW engineers are doing and why we have the current state of things. If, however, we have some progressive foresight, all would be multi and all would benefit because of it. You have thousands of individuals making those "OK for my app" calls and we are left with a hodgepodge environment where most apps are all clustered around the single thread execution point. Thereby wasting all the HW we currently have.

 

[jsolares]

intel may go on a 3 year schedule out past haswell somewhere. We are coming close to the end of moores law and need something different to keep moving forward at the same rates.

When else in the history of computing have you been able to hold onto a computer for so long?

I don't think it's entirely because of the end of moores law, i think it's mainly because the end of AMD as a real competitor.

 

[goMac]

For the exact reasons given. iChat does not run in a vacuum. This IS what SW engineers are doing and why we have the current state of things. If, however, we have some progressive foresight, all would be multi and all would benefit because of it. You have thousands of individuals making those "OK for my app" calls and we are left with a hodgepodge environment where most apps are all clustered around the single thread execution point. Thereby wasting all the HW we currently have.

Yeah, but again, if my app doesn't need the power of more than one core, why bother?

It's not a matter of writing the code with progressive foresight. It's a matter of "will my app ever need more than 400 mhz of computer power?" A lot of things I write run just fine on a 1.6 ghz core on a Macbook Air, or an 800 mhz core on an iPhone 4, and they will never need more power. For me, the time really isn't worth it.

If I'm writing a complex algorithm that maxes out the hardware I'm working on, the first thing I go for is multicore optimizations. But if I'm writing Address Book which could run on a 300 mhz G3, what's the point? I could write address book with GPGPU optimizations too, but again, what's the point?

Multicore isn't "ZOMG multicore everything!" It's a tool for the right jobs, not all jobs.

The system automatically spreads single core apps out onto their own cores (and makes them share if necessary), so yeah, in OS X I can code my app like it is in a vacuum actually. Even if my app is single core, I'm not blocking another application.

 

[ActionableMango]

Not true.

I stand by my statement that there is a ton of crap rumors. What I can remember off the top of my head:

White iPhone 4 coming 1 month after black, okay a bit later, maybe later, ooh someone has white parts!, nope not here yet, maybe later... etc etc for a year.

iPhone 5 cases leaked! It has slanted/beveled edges! NOPE!

iPhone 4S or 5 coming summer 2011. NOPE!

iPhone 5 coming fall 2011. NOPE!

iPhone 4S and 5 will release together! NOPE!

iPhone 4S will have LTE! NOPE!

Larger screen... capacitive home button... a smaller-sized $99 version... NOPE! Did I miss anything? Probably!

 

[robbie12345]

i don't care which comes first but either way
http://www.fybertech.com/4get/13034936768621.jpg

 

[hector66]

I can see Apple updating their site with a new Mac Pro either on Tuesday the 6th or waiting two weeks afterward until after the iPad 3 (and hopefully TV) ruckus has died down.

Although they are two completely different markets I don't think Apple wants either message to be lost in the other.

 

[goMac]

I stand by my statement that there is a ton of crap rumors. What I can remember off the top of my head:

White iPhone 4 coming 1 month after black, okay a bit later, maybe later, ooh someone has white parts!, nope not here yet, maybe later... etc etc for a year.

iPhone 5 cases leaked! It has slanted/beveled edges! NOPE!

iPhone 4S or 5 coming summer 2011. NOPE!

iPhone 5 coming fall 2011. NOPE!

iPhone 4S and 5 will release together! NOPE!

iPhone 4S will have LTE! NOPE!

Larger screen... capacitive home button... a smaller-sized $99 version... NOPE! Did I miss anything? Probably!

Not that there aren't crap rumors, but a lot of the problem is Apple has a plan a and a plan b, and in the end they only pick one, but the rumors cover all possible directions.

The Apple phone and tablet rumors we had back in 2003 were only verified as true last year, when Apple admitted it did have tablets that far back. If the iPad existed for about 5-6 years before it was actually released, it's not unreasonable to think Apple had the iPhone 5 working last year, but didn't release it for one reason or another.

 

[LeahHaskell]

High school mall rats have a higher interest in iCrud. They are and have always been the marketed demographic. Just go to a an ATT store and check out who works there and the general attitude. Those are your new Apple faithful. So sad..."bro". "iPhone, baby!". Y'know, those types. The tasteless are on board. Now where is that Italian captain?

Uh uh, Apple has lost the grumpy old man demographic. Now they are really doomed!

 

[xgman]

Not that there aren't crap rumors, but

With the Mac Pro, we don't even get the "crap" rumors. Nothing, nada, zip! I'd settle for some false rumors at this point. The only rumor we get is that they are rethinking the line. I feel Mac Pro obscurity coming on regardless of this current update. Kind of sad really.

 

[goMac]

With the Mac Pro, we don't even get the "crap" rumors. Nothing, nada, zip! I'd settle for some false rumors at this point. The only rumor we get is that they are rethinking the line. I feel Mac Pro obscurity coming on regardless of this current update. Kind of sad really.

Well, it could be both. They could be thinking about the future of the line and working on a new model.

That said, the people who told us the Mac Pro was dead told us it would be discontinued in January (and that the Mini was dead.) Track record isn't looking so good.

Eventually Apple will kill the Mac Pro. But chances are looking better there will be an upgrade this year.

 

[xgman]

That said, the people who told us the Mac Pro was dead told us it would be discontinued in January (and that the Mini was dead.) Track record isn't looking so good.

Exactly. One rumor I will be happy to have as false.

 

[ActionableMango]

With the Mac Pro, we don't even get the "crap" rumors. Nothing, nada, zip! I'd settle for some false rumors at this point. The only rumor we get is that they are rethinking the line. I feel Mac Pro obscurity coming on regardless of this current update. Kind of sad really.

We get rumors.

OS X drivers hint at desktop card support for 6000 and 7000 series cards, including cards that cannot be used in any other model Mac.

There was the story about a new case that could be used as desktop or rackmount.

Rumors of release date in fall 2011, remember the fiasco with the new "mac pro" part numbers showing up? Rumors of release date in 1Q2012. Rumors of MP outright demise.

 

[derbothaus]

Yeah, but again, if my app doesn't need the power of more than one core, why bother?

Because it is holding the entire "computing environment" back. Devs need to take some responsibility for polluting their own landscape because of shortsightedness. But truly it needs to happen from on high. Apple or Intel needs to offer better compiling tools to make this happen. I see the future of computing as being way more organic (not just green) We will take models of successful organisms and copy and make fit for tech. You watch.

 

[wallysb01]

Because it is holding the entire "computing environment" back. Devs need to take some responsibility for polluting their own landscape because of shortsightedness. But truly it needs to happen from on high. Apple or Intel needs to offer better compiling tools to make this happen. I see the future of computing as being way more organic (not just green) We will take models of successful organisms and copy and make fit for tech. You watch.

So, are we talking neurons instead of silicon? You lost me in that last bit.

 

[derbothaus]

So, are we talking neurons instead of silicon? You lost me in that last bit.

Maybe, why not since making stuff up? I was thinking more like nano fibers and design aesthetics. Use a small battery to start your laptop, the insides are sprayed with carbon nano fiber battery glop to hold and use all the charge that the case takes in from sunlight and diffused lite. You never need to recharge your laptop. It works independent of power. And kind of like a car battery/alternator setup.

----------

Uh uh, Apple has lost the grumpy old man demographic. Now they are really doomed!

Has nothing to do with age. It is about "taste". When the best thing you can think of is a Vegas vacation, you are not even close to being there.

 

[goMac]

Because it is holding the entire "computing environment" back.

Again, how? If my app doesn't need multithreading and doesn't even use one full core, I'm not holding squat back. I'm just wasting time and introducing possible bugs by multithreading for 0 performance gain.

Devs need to take some responsibility for polluting their own landscape because of shortsightedness.

What short sightedness exactly?

But truly it needs to happen from on high. Apple or Intel needs to offer better compiling tools to make this happen. I see the future of computing as being way more organic (not just green) We will take models of successful organisms and copy and make fit for tech. You watch.

I don't think this makes any sense. Programming is not "USE ALL THE TOOLZ". It's using the right tool for the right job i.e. using multithreading where it actually makes sense and actually helps.

This idea that everything including the kitchen sink needs to be multithreaded is really silly, and it isn't going to make anything actually run any better or faster, so I'm not even sure why it's being suggested other than "MULTITHREADED SHINY".

I mean really, please explain to me how multithreading every app, including ones that don't even use a full single core, helps anything.

 

[derbothaus]

I mean really, please explain to me how multithreading every app, including ones that don't even use a full single core, helps anything.

Sorry Go. I just wrote some huge response and then forums signed me out and I lost everything. Maybe we take this up later. Not feeling like re-typing.

 

[goMac]

Sorry Go. I just wrote some huge response and then forums signed me out and I lost everything. Maybe we take this up later. Not feeling like re-typing.

No problem. I'll be around

I do find the topic interesting. In college I did a lot of my research on multicore and GPGPU development, and it is one of my specialties, so I'm not trying to come across as the lazy programmer. I just finished multicore optimizing a project, actually. I'm just saying, I do multicore within reason.