I'd like to see GMA X4500 benchmarks from more sources. The only ones I have are on beta drivers and months old now.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Prepping for Notebook Refresh?
- Thread starter MacRumors
- Start date
- Sort by reaction score
And there lays the invalidity of your arguments. You don't know how these things work. You don't develop for operating systems or perform multicore process optimization. Some of us do, and we are telling you it's not going to be what you think it is.
Will it be better? Yes.
Will it rule the PC world? No
Will it rule the Universe? No, but maybe the Milky way.
Which is misleading as Intel is catching up to AMD in terms of IMC and actually having a real multi core system versus dies being "taped" together.
Like I said earlier Grand Central is sounding like a scheduler not unlike what is used in Cell (PPE controlling what SPE's are doing) just software based and can have any core actually do the scheduling for the others.
A year before the iPod came out, would you have assumed the iPod would come out? Think: the iPod. When the iPod came out, would you have assumed that Apple would sell 100,000,000 units in five years?
You can tell me that it's not going to be as good as Apple says, and for the most part, I will definitely agree.
But only if I can tell you what I know: Never underestimate Apple.
There never was a MhZ myth. There was a common misperception.
You want proof that there never was a myth? It's simple. Take one processor and perform a benchmark of its performance. Take the exact same processor, increase its clock speed and repeat the benchmark. The higher clock speed benchmark will win (with the exception of a very few cases).
Higher clock speed = faster on comparable architectures.
People attempted to compare Apples to Oranges with the P4 and the K7; The difference in their architecture was significant, equal to that of a x86 to PPC different. Yes, they ran the same code, but how they did so was not the same.
Pentium 4 was built for high clock rates. Period. The only reason Intel held on to the P4 architecture was because of how heavily vested they were into it. It was unprofitable to change then.
Point is, apples to apples, higher mhz wins.
Not an apples to apples comparison so it cannot apply.
I haven't underestimated them. In fact, I have given them the benefit of the doubt.
A scheduler can only do so much and is not the end all or single important part of optimizing multicore systems. A scheduler can help significantly, no doubt, however it will not give you system wide performance benefits. Just look at the new CFS in the linux kernel. It works well, in many cases better than the one before it, but it also in cases performs more poorly.
If Apple is smart (and I assume their core developers are), then they will work on optimizing their back end technologies, their libraries, API's, et al., along with optimizing the kernel, its scheduler as well as every existing application that benefits from multiple threads (which in most cases is nearly everything).
Most people seem to think threading is only done for things that need to happen in order, which is why it is so hard to optimize lots of applications. However that is not correct. Virtually every application can benefit from lots of threading, especially GUI applications with lots of objects.
I personally hope Apple does a good job, as I don't have access to core technologies to help improve them.
Here's a question for everyone
Since the discussion has moved on to Grand Central, let me ask you this:
If you could french kiss a gopher while downing a six as you are riding a camel without falling off, does this make you a God among men, or does it really matter unless you actually see it. My point is, does Grand central matter right now?
Since the discussion has moved on to Grand Central, let me ask you this:
If you could french kiss a gopher while downing a six as you are riding a camel without falling off, does this make you a God among men, or does it really matter unless you actually see it. My point is, does Grand central matter right now?
No, they're not (by this I mean Apple)
Not going to happen.
Seriously, I'm not sure what pixie dust you think is going to come in your Snow Leopard box, but it's not going to provide the performance improvement you're talking about. The idea that current PC OSes are only using ~half the potential of their current hardware, or that Apple is going to completely and utterly revolutionise multicore computing in 12 months, is simply laughable.
oh well, going to sleep now guys, good night, I hope new MBPs are presented tomorrow.
Yeah the wait is killing me.
Seconded!
People need to stop looking at just article titles instead of the article. It doesn't matter whether "by" means "before" or "before and including." According to the original article, LED displays were to come in 2008 or 2009.
And I wonder. Why didn't anyone else do that before me? Maybe because they were too involved in discussing the subtleties of "by"? Maybe because they wanted LED screens on the MacBook in 2008 and not 2009?
I think it's Nehalem vs. Core/P7/P6.
Nehalem is about the same as K10 in terms of design, if you know what I mean.
Not now, but in a year's time, and current computers can be updated with Snow Leopard when it is released.
Apple is not going to simply refresh a re-designed laptop. It will come at an event.
I wouldn't expect any refreshes tomorrow unless it was software related.
I wouldn't expect any refreshes tomorrow unless it was software related.
Not trying to be funny or difficult, but didn't the last revision of the MB/MBP have a quiet introduction? I honestly can not remember.
Yes they did but they weren't a complete redesign. Which is what's rumored this time around.
Apple has always shown off new completely redesigned hardware at some kind of event.
Minor updates like new processors or even just small things like LED screens (with the Macbook Pro) don't require events. However I think that they require events for case revisions.
Did you miss the redesigned MacBook in 2006? And don't tell me it wasn't because it stayed plastic, just about everything else changed.
I see iLife '09 and iWork '09 coming out with the redesigned laptops at an event not unlike last year's.
Windows and Linux are already there. They've been supporting multiple CPUs for ~15 years already and had years and years of optimisation to take advantage of multiple CPUs. When Apple delivered the 8-core Mac Pro at the beginning of this year, we'd already had 8-core Windows and Linux servers in production for nearly a year.
Heck, in 1996, Windows NT 4.0 support 32 CPUs.
So first you make a statement about what Grand Central will be, and then you say we can't make any such statements?
But it's not true. OS technology only proceeds at a certain pace, like hardware technology (as you implied at first). It is highly unlikely that Apple will invent something that nobody has ever dreamed of before. And as someone pointed out earlier, Apple has actually been behind what other OS developers have been doing with SMP. "Grand Central" is most likely going to be their way of catching up so they're on a par with the latest OS technology, not making any grand new leaps.
I understand that's what you'd like.
And I'm out.
Untrue, we can look at what everyone else has been doing optimising their platforms and conclude that the - essentially overnight - improvement you are suggesting just isn't going to happen.
Apple have done nothing of the sort.
The index is never higher than the value of the slowest component. So a quad-core, 16G RAM machine that only has onboard video, will have an index of only about 1.0 because of the video card. Your methodology is flawed.
But that's not what Grand Central is meant to do, it's meant to improve the efficiency of multiprocessing across larger numbers of cores, not fewer.
Well I hope to God you are right but I still remain skeptical, I think that it may but again I hope I'm wrong. I'm listening to the Financial report hoping something may be leaked