Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
http://techresearch.intel.com/articles/None/1813.htm

http://www.engadget.com/2009/09/26/exclusive-apple-dictated-light-peak-creation-to-intel-could-be/


Apple specified constraints on the solution. That is not the same as developing it. Neither has Intel put this in front of a standards body either.
They are trying to stick some aspect of it in front of the USB body but this isn't necessarily the right standard.

Apple (and Intel) could ship something before reaching a significant consensus from the industry, but long term that may end up hurting as much as helping.

Remember, for this to be useful and widespread it has to be someone other than Apple and Intel that has working offerings. There were some vendors who said Firewire was cool too .... until they have been drifting away. ( maybe for this. )


Some of Light Peak seems half baked (e.g, USB like power over the same cable. ) Frankly have trouble with the Light Peak distance claims ( 100 meters) http://en.wikipedia.org/wiki/Light_Peak and providing power over those kinds of distances. Which either gets you short distance with power Light Peak and pure optical Light Peak or dropping of some of the "yeah and that too" claims.

Likewise all the of multi/demultiplexing of putting random stuff out of the lab on the same wire.


Mac Pro with USB 3 might be more likely. Or that Apple is going to by-pass it (like eSATA).

Apple could bypass USB3...
It's seems the Intel side of Light peak came out of the sidelined usb3 optical spec. Free of USB it could then take on board ideas it couldn't before.

Also thinking it adds to the work of UWB which also able to carry lots of data stream types sort of natively but with light as the carrier not radio.

The problem is it's all very short on details.
Yup. Lightpeak is at this point a technology demonstration. It's nowhere close to production if they're having to modify existing motherboards to show it off.

The Mac pro is intel's tech demo machine of choice.
 
For the most part, MOST users will get nothing but bragging rights. BUt then again, the people buying up these mac Pros generally are video/gfx pros - like myself - and therefore use apps that would utilize them all.

Prime example being 3D apps. Sure not the most widely used apps on the planet, but the app i use now, CInema 4D uses every single core to its 100% max to render frames from preview and for final output. Its takes your workspace, say 1280x720 pixels, and on a dual core machine, chops it in half, givin the workload for the top part of the image to one processor, and the bottom half of the image to the other. On a 4 core its split up in 4 parts and each part works simultaneously. On an 8 core, so forth Etc. On a 12 core machine, especially one with a pretty fast clock on a 32nm chip, it'd blaze through what takes my dual core macbook pro at 2.8GHz 2 minutes to render, in theoretically, about 45 seconds. THats some significant time savings. This grows exponentially when you're rendering out a 3d animation as you're saving crazy amounts of time for every frame, 24 or 30 times every second, for several seconds. THe more cores the merrier.

there's also the Compressor encoding app. Adobe after effects. Logic Pro. Various third party encoding apps and a slew of other 3D and motion graphics apps.

However, OpenCL should change this game up pretty significantly. I'm excited to see about that.

This system will make a compelling reason for ANSYS, Pro/Engineer and others to make a Mac native Port.
 
This announcement makes Apple's Mac Pro line sound top-notch, ergo I wouldn't expect any price drops at all. Personally, I think that its a justification for people who really enjoy having the most top-of-the-line system at all times get lured into Apple's product segment. Sure, they can OC, but at the same time.... ZERG-OH-MY-GAWD, SIX CORES.

I think that Intel has to benefit from this as well - hence the early release only on Intel firmware (EFI, if I am not mistaken). New Mac Pros might be Intel's version of a "public beta", where any defects that occur on large scales can be targeted and eliminated from the production line later on.

There might also be a firmware update to something completely different, maybe UEFI? Doubtful, but Apple is a part of the UEFI group, so not out-of-the-question. Which would mean incompatibility of all older cards with the new stuff Apple has to roll out.

And there could be a warrentee update, as well.... (Processor removal + sticker to check if that is the case).

But I'm also looking forward to its performance as well, despite my large doubts; maybe SATA won't be rarefied this time around. But it still remains that a "temporary exclusive" might not be the best thing in the world
 
I think that Intel has to benefit from this as well - hence the early release only on Intel firmware (EFI, if I am not mistaken). New Mac Pros might be Intel's version of a "public beta", where any defects that occur on large scales can be targeted and eliminated from the production line later on. IIRC, aren't these the first 32nm chips to be mass produced? I could be wrong.
Arrandale and Clarkdale are 32nm Westmere architecture as well.
 
Brand New Gulftown Mac Pro! Now with cutting edge Gulftown Intel processors and the fastest graphic card Nvidia has to offer - the 7300SE (Underlocked Edition)! Starting @ $3299.

Get your own now!
 
Hahaha, wow, on top of offering Xeon's exclusively, you can't even build a Mac Pro with a workstation graphics card? I just had to check to see if that was really true.

You can, NVIDIA do their Quaddro series of graphics cards for the Mac pro, ideal for 3D rendering, and video exporting etc.
 
http://techresearch.intel.com/articles/None/1813.htm

http://www.engadget.com/2009/09/26/exclusive-apple-dictated-light-peak-creation-to-intel-could-be/


Apple specified constraints on the solution. That is not the same as developing it. Neither has Intel put this in front of a standards body either.
They are trying to stick some aspect of it in front of the USB body but this isn't necessarily the right standard.
i thought they proposed the idea to Intel and they did what they did? i didnt mean develop, sorry - wrong word.

Apple (and Intel) could ship something before reaching a significant consensus from the industry, but long term that may end up hurting as much as helping.
i was under the impression that it it was all ready to go.. or did they recently just test it?

Remember, for this to be useful and widespread it has to be someone other than Apple and Intel that has working offerings. There were some vendors who said Firewire was cool too .... until they have been drifting away. ( maybe for this. )
why cant intel do it? not big enough? people wont see it as an "intel" thing to do because all the do is make CPUs?

Some of Light Peak seems half baked (e.g, USB like power over the same cable. ) Frankly have trouble with the Light Peak distance claims ( 100 meters) http://en.wikipedia.org/wiki/Light_Peak and providing power over those kinds of distances. Which either gets you short distance with power Light Peak and pure optical Light Peak or dropping of some of the "yeah and that too" claims.
i have not yet been educated on the workings of LightPeak, but the concept seems great. incorporating everything into the one.. seems ok to me.


Mac Pro with USB 3 might be more likely. Or that Apple is going to by-pass it (like eSATA).
but USB is so clunky... im not that keen on it its so unreliable.
 
Nothing new, just when is the release date.
Im not buying another MacPro for work (we have a dozen) until I figure why these things dont shut down properly without having to hold down the power button :p Plus the fact that the video card options are still at ProSumer level.
 
No, they're just charging more to the end-user who may not need an ECC memory workstation.

Users don't need mechanisms to protect them from errors? Really? Maybe the drive vendors should shave some costs by removing ECC off the hard drives too.

Memory isn't flaky. Gee...

http://news.cnet.com/8301-30685_3-10370026-264.html

With folks using many GB of memory, ECC should be a defacto feature for everyone.
 
Apple Roadmap

So by the summer of 2010 we will have:

  • 12-core/6-core 'Gulftown' Mac Pro
  • 4-core 'Clarksfield' iMac
  • 2-core 'Arrandale' mini
  • 4-core 'Clarksfield' MacBook Pro
  • 2-core 'Arrandale' MacBook
  • 2-core 'ARM-9 Cortex' Tablets???

Man ... I want the whole list ... well I guess I'll skip the MacBook.
 
I don't like this news, personally. I currently have an 8-core processor and no need for additional cores. If I buy a new computer, I'd have to decide between needlessly upgrading to 12 cores or shamefully downgrading to (a perfectly adequate) 6 cores. :mad:
 
I don't like this news, personally. I currently have an 8-core processor and no need for additional cores. If I buy a new computer, I'd have to decide between needlessly upgrading to 12 cores or shamefully downgrading to (a perfectly adequate) 6 cores. :mad:
Why would you buy a new computer then. Seriously your 8-core MP should be good for at least 3-5 years. By then you will be choosing between 16 or 32 cores.
 
A Mac Pro with 12 cores is awesome. Now, it needs the software to make it go bang.

I'd like to see Apple take a break from building 10.7 and focus on making their OSX core apps better. I'm talking both client and server side. It is all about the apps that make OSX nicer to deal with than Windows. As an example: make a Pro version of Mail and bundle it with iWorks X (calendar/address book/mail). On the OSX Server side solicit feature enhancements for the server apps and run with them. Intro some cloud computing services to challenge Amazon and Google. I'd like to see the recently acquired mapping software company's product in every aspect of OSX like the dictionary.

Apple's hardware is cool but make the software even better and don't rush to build new major version of OSX. Slow down as Snow Leopard has a good foundation. Those 12 cores will make good use of grand central with the right powerful apps. :)
 
His Keyword was SLI.
True, but he mentioned SLI because for some reason he seemed to think SLI was required for OpenCL to utilize multiple GPUs, which it isn't. People with current generation Mac Pros with 4 graphics cards, and even MacBook Pros with 9600M GT and 9400M can already use all devices simultaneously for OpenCL.
 
LMAO you do realize that 2002 was almost EIGHT years ago, yes? You see there's this thing called "inflation". Ah, nevermind, you're probably too young to understand.

Yeah, there is -- and given the inflation rate since then, $1700 in 2002 dollars is equivalent to about $2020 in 2008 dollars. (Don't have data for this year, yet, but trust me, the inflation rate has been pretty low so far.) That's still well shy of $2500.
 
LMAO you do realize that 2002 was almost EIGHT years ago, yes? You see there's this thing called "inflation". Ah, nevermind, you're probably too young to understand.

Well, you see, 12 months ago, we started a new thing. Called "deflation." It's why houses are 50% off, jobs are harder to find, pay isn't rising, cars are cheaper, every state in the nation has record drops in tax revenue, food is cheaper, and Colorado is lowering their minimum wage next year.
 
i wonder if Apple will update the EFI on the pros so you can shut off Hyperthreading. Some applications, especially 3d apps which take full advantage of cores up to their license limit actually suffer performance wise with it enabled. Especially MentalRay. Sadly, when i called Apple when i was ready to buy a MacPro, i asked about disabling Hyperthreading, they looked in to it and recommended me to not buy a MacPro because there is no way to turn it off.
Though 12 cores with Renderman would rock!
 
Yeah, there is -- and given the inflation rate since then, $1700 in 2002 dollars is equivalent to about $2020 in 2008 dollars. (Don't have data for this year, yet, but trust me, the inflation rate has been pretty low so far.) That's still well shy of $2500.

But are you also well short of the same functionality also?
In 2002 you could put 4 disk drives with provided sleds? digital audio?
The bottom end PowerMac lacked a L3 cache.
http://support.apple.com/kb/SP78
Fewer populated memory slots in default config. (either just one or two in 2002 versus 3 now. )
Bigger power supply now .
Three fewer USB sockets.
Three more CPU cores.



I think all that aluminum of the newer model probably costs a few more dollars.

Is that $500... probably not but does close the gap a bit more.

In the move from the 2009 Pro single quads from the 2008 single quads Apple slid in a more afforable processor 35xx and didn't really drop the price from the price year. It is a juicy target if pointing at them probably juicing the margins a bit.
 
Well, you see, 12 months ago, we started a new thing. Called "deflation." It's why houses are 50% off, jobs are harder to find, pay isn't rising, cars are cheaper, every state in the nation has record drops in tax revenue, food is cheaper, and Colorado is lowering their minimum wage next year.
To be fair, pegging computer prices since 2002 to inflation or deflation relative to other indicators is hard to do since there is so much variability that people can cherry pick specifications to make whatever point they want to.
 
Its getting old. There is nothing wrong with a refresh now and again.

I wouldn't count on it, we need a big case and the metal design is perfect. The pro models are stackable so a redesign would make them un-stackable, smaller would limit the possibility's and moving to plastic is a big step back. I believe this is the perfect design for workstations, we don't need colors and a design-statement beyond the current case that is to this day practically unmatched on the PC-side.
 
Why the heck do you want a different case design? :confused: The current one looks nice and works great. Change simply for the sake of change is idiotic. Why exactly do you want them to change it?

There's absolutely nothing wrong with the Mac Pro case. It's unique and everybody knows it's a Mac. It's only the internals that need revising every time an update is required.

It’s funny because both of you act like Apple couldn’t come up with something better.

I for one would like their next case to be lighter, shorter in height (the current case won’t fit under your average size desk), insulate sound better, move the power supply back to the bottom so you don’t have the power cord dangling in with all your other cords, SD card slot in the front, another row of USB 2.0 connectors, etc.

I also think the handles needing protection while be transported is somewhat of a design flaw, and they should reinforce the handles on the next case (maybe it’ll be built using a unibody process, wouldn’t that be wild — of course that’s not really feasible).

But whatever, this thread is about Gulftown.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.