[Moogs]

I'm not real clear on this but it sounds like 65nm was a sort of barrier for increasing the performance of chips that are in use today. This article

http://news.bbc.co.uk/2/hi/technology/6299147.stm

seems to indicate that a successful transition to 45nm signals a new start. Is this simply going to mean less heat and more cores around 3GHz, or does it mean once again the race for higher clock speeds is on, and we'll start seeing 4 and 5GHz chips over the next couple years?

PS - can someone put the "?" at the end of my post title. Left it out by mistake...

[Moogs]

More info

From NYT Many executives in the industry say that Intel is still recovering from a strategic wrong turn it made when the company pushed its chips to extremely high clock speeds — the ability of a processor to calculate more quickly. That obsession with speed at any cost left the company behind its competitors in shifting to low-power alternatives. Now Intel is coming back. Although the chip maker led in the speed race for many years, the company has in recent years shifted its focus to low-power microprocessors that gain speed by breaking up each chip into multiple computing “cores.” In its new 45-nanometer generation, Intel will gain the freedom to seek either higher performance or substantially lower power, while at the same time increasing the number of cores per chip. “They can adjust the transistor for high performance or low power,” said David Lammers, director of WeSRCH.com, a Web portal for technical professionals.

This quote would also seem to indicate that once again clock speed will be moving forward from generation to generation, instead of being stuck between 3GHz-3.5GHz as it has been.

[torifile]

Moogs, you know that if you double click in the "thread title" field on the forum index page, you can edit your title in-line. As long as you do it within a certain time, you don't need to have a mod edit it. Be sure to click on the whitespace as opposed to the actual title otherwise you'll end up in the thread.

[Moogs]

Oops. Nope didn't realize that. Used to be that if no one had responded the Edit function would bring the title field back for changes, but when I went back right away and clicked edit, the title field was not available.

Anyway some presentation for Babar...

...Nearl thinks maybe waiting another 6 or 8 months for next gen Mac Pro, better than getting current upcoming generation. Thanks Ken!

[Eugene]

65nm wasn't a wall, but whenever you go down in transistor size you have to deal with more leaking current (read lower efficiency) and increased power density...more transistors in a smaller space.

[Enki]

All other things left constant yes. Seems Intel tried some new multi-material transistors and cut the loss by 30% rather than a proportional increase in the loss.

The rules be a changin'. Once upon a time shrinking process size and materials was taboo. When you did that all hell broke loose and you ended up just like Moto did, IBM hosed that up too. Seems Intel must have got it right.

[psmith2.0]

All I know is that it's so nice to read stories like this and know that all this cool, good news applies to us Mac users now? Or at least the possibility is there.



We're no longer locked out in the cold, watching Intel and others making these types of advances, while we sit there with G4 and G5 machines that have pretty much hit their ceiling.

Honestly, 18 months ago would any of us have imagined something like the iBook ever sporting as much muscle as the $1,099 MacBook does? We would've been "thrilled" with 1.8GHz G4 in the 14" iBook probably.

This Intel switch was the best thing Apple has ever done, IMO. Makes it about the OS and overall user experience now, more than ever. And on that front, the other guys are gonna have their hands full.

[Partial]

Originally Posted by Eugene 65nm wasn't a wall, but whenever you go down in transistor size you have to deal with more leaking current (read lower efficiency) and increased power density...more transistors in a smaller space.

Wouldn't you have to deal with less leaking current? Thus, the improved efficiency from the transistors closer together?

[Chinney]

Originally Posted by pscates2.0 All I know is that it's so nice to read stories like this and know that all this cool, good news applies to us Mac users now? Or at least the possibility is there. We're no longer locked out in the cold, watching Intel and others making these types of advances, while we sit there with G4 and G5 machines that have pretty much hit their ceiling. Honestly, 18 months ago would any of us have imagined something like the iBook ever sporting as much muscle as the $1,099 MacBook does? We would've been "thrilled" with 1.8GHz G4 in the 14" iBook probably. This Intel switch was the best thing Apple has ever done, IMO. Makes it about the OS and overall user experience now, more than ever. And on that front, the other guys are gonna have their hands full.

Unless of course, Apple had just waited a bit more for the Power 6. But as the article points out, all this is academic now. Heck, I am the owner of an Intel Mini in addition to my G5 - who would have thought?

[Paranoid666au]

Apple can still go back to PowerPC. It's simple really, IBM didn't have the right CPU for the laptops so that was a primary reason for Apples switch to Intel on x86. But OS X runs on both and apps are universal. So Apple could easily drop the Power 6 into any of it's line up.

It's fantastic, both Apple and ultimately us it's customers, have a super wide choice between x86 processors and PowerPC processors. And it makes Apple far less dependant on any one CPU manufacturer.

[Enki]

Originally Posted by tensdanny38 Wouldn't you have to deal with less leaking current? Thus, the improved efficiency from the transistors closer together?

No, it usually gets worse as there isn't as much insulator between layers and traces. Effects like quantum tunneling begin to show up where single electrons slip through empty spaces between atoms. Given layer depths were only about 10 atoms thick at 90nm that would get worse without specific mitigating techniques as it got thinner.

[Partial]

Originally Posted by Enki No, it usually gets worse as there isn't as much insulator between layers and traces. Effects like quantum tunneling begin to show up where single electrons slip through empty spaces between atoms. Given layer depths were only about 10 atoms thick at 90nm that would get worse without specific mitigating techniques as it got thinner.

Ahh. Cool.

[ezkcdude]

Although an increase in processor speed is still important, the real progress in computing power is going to come from additional cores. Hopefully, the software catches up with the hardware.

[Enki]

Originally Posted by ezkcdude Although an increase in processor speed is still important, the real progress in computing power is going to come from additional cores. Hopefully, the software catches up with the hardware.

Hopefully the damn memory latency catches up with the processor.

[Chinney]

Originally Posted by Paranoid666au Apple can still go back to PowerPC. It's simple really, IBM didn't have the right CPU for the laptops so that was a primary reason for Apples switch to Intel on x86. But OS X runs on both and apps are universal. So Apple could easily drop the Power 6 into any of it's line up. It's fantastic, both Apple and ultimately us it's customers, have a super wide choice between x86 processors and PowerPC processors. And it makes Apple far less dependant on any one CPU manufacturer.

Ack. I floated this theory a while ago, but I was pretty much shot down on these boards. I suppose that there would be two main reasons against any Mac in the near future being offered with PowerPC:

1. Bootcamp. It is selling computers. I wonder if a robust Parallels will reduce the practical attractiveness of Bootcamp, but the fact of the matter is that there is a substantial subset of new Mac purchasers who like the idea of being able to boot natively into Windows, whether or not they will ever actually use it, and a substantial subset of those who do, in fact, use it and and expect to be able to continue using it.

2. Contractual tie-up with Intel. About this I am only speculating, but I imagine that going back to PowerPC any time soon may not be something that Intel would tolerate. Also, it may not be in Apple's interests to be seen to make a public about-face after making a very public switch.

IBM/Freescale would have to make something pretty special for Apple to consider a move back, even partially.

[DMBand0026]

Could it be all bad if the GHz began to increase again? I completely agree that clock speed is the wrong way to judge the worth of a chip, but certainly you'll agree that a 3.5ghz chip is faster than one clocked at 3.0? So if we make it closer to 4.0 again, or even creep above that, I for one won't be complaining. It sucked before when all we had was overclocked single core chips that did little but overheat, but I think manufacturers are smarter than that. The chip makers (Intel, AMD, Freescale, IBM) have shown that their current road maps involve more cores, not necessarily higher clock speed, but if higher clock speed is a byproduct of regular R&D than it can't be all bad.

I doubt that Apple would go to a two processor type system, they've got a great thing going on right now. Intel can make more than enough chips to supply their every need. I really think that a big reason behind the initial switch was that Freescale/Moto were unable to provide enough chips. Furthermore, it wasn't just one time, it was several times and it caused huge delays for Apple (I'm sure they were forced to push back product launches because Moto wasn't ready) and even bigger delays for consumers who had to wait several months for their new machines. On top of all that we'd wait nearly a year with one of Apple's lines completely dormant only to see a 200mhz increase in power. I'm not saying that Apple wouldn't give them another chance if they were able to provide proof of adequate production abilities, but I don't think that's a road Apple will be going back down anytime soon.

However, if you think that Apple hasn't been continuing to develop the PowerPC version of OS X, I firmly believe you're wrong.

The future of processors is in cores, not clock speed, however, I think that we'll see higher clock speed along with more and more cores.

[specter]

At first I should state that I'm a big fan of Parallels. I prefer using Parallels instead of boot camp, because I just need windows and Mac OS to run simultaneously on my machine. But I still think that Parallels, being "robust" as you say, will not replace or bring down Boot Camp, at least in the nearest future. The main issue that keeps Boot Camp afloat, when compared to Parallels desktop, is 3d graphics. We all know that Parallels is a VM, it need great amounts of memory to run properly. I think 2GB is optimal for Parallels. Even if the Parallels team makes 3d support available with the next beta, it will remain impossible to run serious games. That's why, I think, people will continue to use Boot Camp - mainly for games. The other thing is that Boot Camp is free, while one will have to pay about 80$ to become a happy user of Parallels. In fact, Parallels is somewhat complicated, and some new Mac user may choose Boot Camp, because it's more simple, while it is less convenient to reboot each time you want to switch your OS.

[20newUser]

Originally Posted by specter At first I should state that I'm a big fan of Parallels.The main issue that keeps Boot Camp afloat, when compared to Parallels desktop, is 3d graphics. We all know that Parallels is a VM, it need great amounts of memory to run properly. I think 2GB is optimal for Parallels. Even if the Parallels team makes 3d support available with the next beta, it will remain impossible to run serious games. That's why, I think, people will continue to use Boot Camp - mainly for games.

Once I visited Parallels official blog and I read there that Parallels' next beta will definitely include 3d graphics support and it will make playing games under Parallels possible. I just can't remember where exactly it is written. Do you mean that playing games will still remain impossible under Parallels, when they speak about it on their official site? Thanks

[specter]

In fact, yes and no In theory 3d graphics support makes running games under Parallels possible. But on the other hand, in practice, we have a Parallels Desktop VM that needs at least 1GB of ram just to run simple windows apps at native speed. It means that you will need a very serious machine to play games, and RAM requirements will be enormous! But still in theory, it is possible I think, that 3d graphics support will benefit those who will run complicated 3d design programs under Parallels. Of course, 3d graphics support will let people run Vista with it's graphics on their Macs. That's my personal opinion.

[20newUser]

In any case, I think that when new beta of Parallels with 3d graphics support comes out, it will be an absolute boom in virtualization. I also like Parallels very much, and though I use it less than a month, I got addicted to the program, because it makes my life much easier now. At first they released Coherence mode, which is a complete wonder, and now I await 3d graphics support.

[specter]

20newuser, we don't swim in the current of this thread I think that if wish to discuss advantages and disadvantages of Parallels desktop, Boot Camp and virtualization it would be better to post new thread somewhere instead of littering other people's threads that are little related to Parallels...