MacPro 8 Core Teaser!

[Frodo]

Welcome to Geekland!

Some of this is beyond me, but here ya go:

http://www.macdevcenter.com/pub/a/mac/2 ... essor.html

http://www.pcquest.com/content/Supercom ... 051004.asp

http://en.wikipedia.org/wiki/Multiprocessing

64-bit apps, etc:

http://developer.apple.com/macosx/64bit.html

64-bit Transition Guide-- facts/fiction/article links:

http://developer.apple.com/documentatio ... index.html

[newrigel]

It's all meaningless until software developers recode apps to make use of the extra resources.

The problem is, developers won't recode apps that make use of the extra resources until consumers buy enough of the machines with those resources to make it worthwhile. Classic chicken and egg.

Well... TIME is $$ in the pro industries so I don't see why the developers wouldn't get off thier ass because if I can make more $$ by more efficient time management then these developers have to be out of thier minds NOT to support it!

[pdube]

Thanks for the articles, Frodo. I understand it this way: If the app doesn't know there are multiple processors, it won't put out multiple threads, so it can only run on one processor even though the OS is aware of multiple processors. It also seems like the compiler rather than a programmer is actually responsible for most of the work required to make an app SMP aware.

Am I missing anything?

Thanks

[Frodo]

Thanks for the articles, Frodo. I understand it this way: If the app doesn't know there are multiple processors, it won't put out multiple threads, so it can only run on one processor even though the OS is aware of multiple processors. It also seems like the compiler rather than a programmer is actually responsible for most of the work required to make an app SMP aware.

Am I missing anything?

Thanks

pdube, your conclusions are as good as mine. I'm sure there is a little more to it, but in a nutshell-- when the "right hand" doesn't know what the "left hand" is doing, there will always be compromises and limits until all cooperating elements are optimized to work together at full tilt.

(how's that for a layman's explanation? )

[monkey man]

Incredibly, I understood it.
A rare glimpse of light for the monkey here in the Geekzone.

A lay-monkey's explanation, indeed.
Good to have a buddy who speaks Geek.
Oh, and monkeynese.
MM

[pdube]

These somewhat technical issues are important to understand if you want to upgrade your system in the most effective and cost efficient way possible. They can lead to analysis paralysis, however. Should a person buy an 8 core system if a key, resource hungry program like a convolution reverb isn't smp aware? You all know the dilemma that follows that: There are rumors that smp is in development for that reverb and if you wait for it you can get an even faster 8 core system cheaper then. By the time the smp convolution reverb appears, however, there are rumors of a 16 core Mac, so should you...AAARRRGGGHHH!

I'm sure this has been asked before, but has anyone heard of what might be called an upgrade protocol or algorithm to help resolve these problems effectively and under various scenarios?

[HCMarkus]

As long as your host DAW is multi processor aware, like DP is now, it should be able to work with 2, 4, 8, 16 or more processor cores. No current plug ins require more horsepower per instantiation than a fraction of that found within a single core. It is the host DAW's job to distribute available computing resources to the plugs.

For the foreseeable future, multi cores (and 64-bit computing) are where we will be finding increased system performance. I figure 16 cores to be the magic number for my move from PPC to Intel. Two years? One Year? Six Months?

[Frodo]

I've gotta say that I'm more optimistic about where things are going than I am about where things currently appear to be.

[Frodo]

CNET.com has posted benchmarks for the Clovertown and Kentsfield quad-core processors. They use both terms "unofficial" and "official"-- "unofficial", I think, refers to the potential 8-core machines to come while "official" refers to the notion that the benchmarks indeed came from Intel.

http://reviews.cnet.com/4531-10921_7-66 ... l?tag=blog

Just to underscore much of what we've already discussed, here is an important observation from the article:

Even though dual-core processors have been around for a while now, you'd still be hard-pressed to find many mainstream applications that can efficiently take advantage of both processing cores at the same time (typically referred to as a multithreaded-application). Double that number to four processing cores, and the list of supported multithreaded applications gets even shorter. Double it again to eight...and you get the idea.

Some professional multimedia and scientific applications, however, are designed to take advantage of as many processors as are present--and performance will scale accordingly, based on the number of processors available.

The word at macrumors this week is that ..Apple had been rumored to be introducing 8-Core Mac Pros as early as this month.

The wording is curious-- "had been"... perhaps just a symptom of the rumor mill.

http://www.macrumors.com/pages/2006/11/ ... 5356.shtml

[HCMarkus]

I've gotta say that I'm more optimistic about where things are going than I am about where things currently appear to be.

But Frodo, things could be a lot worse than they are now, couldn't they? Like you, I have found the past few weeks/months once again to be very productive using 5.1. I am making the best recordings of my life with BFD, Stylus, EW Silver and Ivory all running at once at 128 buffer, with Altiverbs as creamy frosting and plenty of audio tracks to boot. I gotta' say that, despite my earlier statement that 16 processors may push me into Intel land, I would not seriously consider going there for at least another year. As selfish as this may sound, I prefer to let others spill their blood on the (b)leading edge... I'll write some sad songs from the safety of my PPC, sing the pioneers' praises, and then collect the royallty checks!