Andy Ihnatko's rumor might be true after all.. - Page 11

At this point OS X and a great deal of software has been optimized to use two cores. So their is considerable gain from only using one.

I'm not saying I think Apple should not use quad's across the iMac line. What I am saying is don't think its imperative because it won't offer much improvement for the average computer user.

Things don't work like that. You cannot just will something into being simply because you want it to. Of course everyone is working on improving technology. But it takes the time it takes and it won't happen any faster.

The graphic user interface had been around for some years before Apple used it. It took the time it took to make it into a viable product.


Technology doesn't just come about simply because you want it to.

The iPhone is many times more powerful than the computer that sent a rocket to the moon. That happened incrementally over 40 years, their was no big sudden leap.



Again you are using unrealistic expectations, things don't work like that.

I'm not saying that Apple should not use quad processors.

I'm simply saying quad processors are not automatically better just because their are four processors. The performance gains come under certain circumstances.

How else is technology moved forward? Someone must be the first to try a new part/component/technology.

More and more it seems businesses just patent ideas with no intent to actually ever use them. They want patents only so they can sue others who are bold enough to actually apply the same idea first.
It is no longer about keeping others from using your ideas, but rather about keeping others from using new ideas altogether - while you yourself never intended to use it in the first place. At that point it is stiling progress. Companies should be forced to lose patents they don't use. And not decades later.

But boring patent issues aside, Apple does have a lot of great ideas patented, they just never seem to make it to happen.

I understand the conservative business approach, motto: "Let's see someone else do it first, and then we can do it better by learning from their mistakes" (while we sue their asses off because we hold the patent too).
But that way the Macintosh would have never happened. You cannot introduce radically new technology with a guaranteed market.
There is no success guarantee in new technologies. That's not how things work either.


As a middle ground I think Apple should start releasing 'Special' or 'Limited Editions' where they can test new technologies on a small run (with a premium price). There will always be people willing to pay a premium for the latest and greatest, the true 'consumer pioneers' (and it shouldn't be hard to find many of them among Macintosh folk).

This approach hedges potential losses, yet gives Apple a chance to introduce radically new technology without upsetting everyone automatically. So Apple can be bolder while testing the waters for things that could eventually roll out to everyone.

All electronics companies have research and development labs where they are testing new technology. They are not going to ship a product with new technology until its ready, reliable, and cost effective.


Patents are used for different reasons. Sometimes its a defensive measure for a product that you may not release for awhile, if ever.


What you call the conservative business approach is the line between remaining profitable and going out of business.


The goal is to release a product that does what it does as best as you could design it. Shipping new technology just because its new does not necessarily accomplish this goal.



This is not a good business strategy. You have to understand business is not a game. A company cannot afford to ship experimental products. They have to ship reliable, functional, cost effective technology when it becomes available.


Shipping radically new technology for the sake of shipping radically new technology is not a mission statement of Apple. This strategy would not be supported by Apple's board of directors or major stock holders.

I'm totally with you on the general assessment. Yet I think Apple should think differently.

- If everyone waits for others to make the first step nothing will ever happen. Someone has to be the first using a new technology. Apple did so with quite a few products and I really think it should be their mission statement. It is at least their legacy.

- While a misstep might mean going out of business for many technology companies, Apple surely is in a different league. With 23 Billion dollars of cash reserves they can make total losses on a number of products and still not go out of business in a long time.
I really think Apple should start putting this cash to work. At least part of it.
And more daring product releases would be one way. And it would be in line with Apple's legacy.
I'm not saying they should release technology for technology's sake. Release something that's immature. But I have the impression that they're sitting on the fence with a lot of technologies, unsure if they would succeed or not, 50:50. At that point, they should be bolder IMHO and go for it.


Anyway, back on topic.

How if Ireland would get in contact with Andy Ihnatko and exchange 'ideas'. Just to see if they both talk about the same thing.

Or, Ireland, are you 100% sure you've got the same thing in mind as what Andy was alluding to?

Simply being the first to implement a new technology does not automatically make you the winner. Their are plenty of examples where a company improved on a technology that was pioneered by a competitor.

Apple has been on both sides of this equation.

Apple was the first company to launch a PDA with the Newton, only to be beat three years later in the market by Palm.

In 1997 SaeHan Information Systems shipped the first mass produced mp3 player. Four years later Apple came with the iPod and conquered the market.

Ah don't be so touchy. As the old one goes; if you didn't care, you wouldn't need to let us know.

I agree to an extent, you only notice the difference if you do multiple intensive tasks and keep working. For example, you can do a Shake render and a Final Cut encode at the same time and the quad will still let you do basic tasks whereas a dual core would stutter.

But now that desktop quad processors are so cheap, in the interests of price vs performance, there's no sense in using dual core chips unless they use desktop dual core chips of course but they run too hot (95W TDP).


So if someone was sitting next to you on a train with a foghorn blasting in your ear, you wouldn't tell him to knock it off? After all by doing so, you're saying that you care about his reasons for making the noise.

This goes back to what I was saying before about your SNR approaching zero. Look back at all the posts you made in this thread and weigh up the content relevant to the discussion. You're welcome to join in the discussions at any time but choose to simply make empty posts, which as kickaha correctly points out, most of us can't really be bothered with.

The impressive part of whatever Apple are up to is the implementation, not the idea. The idea that Apple were making a phone was terrible. The implementation was very different.

Whatever ideas Apple are toying with right now are meaningless until they start implementing something.

Days since updates:

Mini = 546
iMac = 281
Mac Pro = 392
Cinema displays = 671
XServe = 392

As I've said before, at this point, unless they come out with time travel and anti-gravity, the reactions won't be 'WOW' but 'it's about f* time '.

i'd have bought a new iMac already in october/november ... but i'm still waiting.
i don't want to buy a 1-year-old system ... update the iMac (and every other system) twice a year: once for a new hw-design and once with updated cpu or gpu

Is it possible the delay is due to some significant architectural changes?

Has anyone figured out what the heck Andy Ihnatko was so impressed with, maybe he should come out and explain.

AppleInsider Express. Choo-choo!!!

lol

This thread certainly has been *de-railed*...


OH SNAP!

I didn't think it was going anywhere in the first place. That's what happens when a thread starts based on a wild, empty, factless, baseless rumor. Anything can happen.


Hmm, another clue. What could it mean?

BUY STOCK IN TRAIN MAKING EQUIPMENT!! Instead of having you haul your broken iMac screens back to the Apple store, they will send out the Apple Train to pick them up. At most 48 hour turnaround time.


Nah, I think it's partly Intel's fault combined with Apple's design choices. Good desktop chips have been out for ages but they can't go into Apple's consumer lineup. The mobile chips haven't really improved or dropped price so there's little that would sell an upgrade model besides it being new but the desktop chips are racing ahead.

This is why the new laptops are pretty much the exact same speed as the old ones. The selling point they can make on the laptop is durability, better cooling etc. This doesn't fly on a desktop where it's about performance and value for money.

They can't shoehorn a 95W CPU into an iMac that is only meant to hold about 55W tops so they had to wait until Intel landed the new low power desktop quads running at 65W. Still a bit of a technical challenge as the iMacs are still not much more than laptops.

This shouldn't affect the mini if they are crippling it again but they won't push it out ahead of their favorite in case, god forbid, people actually prefer them. If they put quads in the Mini too, well that would be awesome but they probably won't.

Either way, since Intel only released these chips recently (last week):

http://www.anandtech.com/cpuchipsets...oc.aspx?i=3505

they couldn't really have put anything significant out earlier.

The Mac Pro and XServe share the same chips and AI reports these won't come until the end of March from Intel. Gainestown is already out though but not all the chips so it will depend on which ones Apple chooses.

The Cinema display line will come with the Mac Pro I imagine but they're probably ready to push out an all glossy lineup and smart people inside the company holding them back saying 'are you crazy, the 517 people we manage to sell these expensive displays to worldwide won't ever buy from us again, we'll lose thousands of pounds'.

Not to mention on top of this Apple may be fumbling around like a headless chicken now that number 1 is out of action.

Still, the positive things to note are that the low powered quads are out and there is a timeframe for Mac Pros. Given that they don't usually update these machines together, we can only hope that the Mini and iMacs will be out within a matter of days. If so, it would be a new strategy for Apple not to have an event for a redesign but you saw what happened to their stock the last time they did that. I think it's safer to just put the updates on the store and let people figure out the spec for themselves. The US dollar prices they quote are always pretty meaningless worldwide anyway given that Apple use their own tax calculators.

You really think Jobs is the only smart person at Apple?

Former Apple Exec: Tim Cook's Been Running The Company For Years ...




I think its pretty clear Jobs doesn't care for the Mac mini. Someone else in the company is saving it, and why it has been in a strange limbo.

Just throwing this one out there in what has been a thoroughly entertaining thread.

The, err, spirited discussion Steve Jobs had with Sony Music in December got me thinking back to a few tiny rumors that emerged twelve or so months ago.

Apple to start record label?

I've always been a Monorail guy myself.

Smart people are not always leaders. I find a lot of company bosses are not all that smart. But smart doesn't have a universal definition. Is Einstein smart because of his intellect but died in poverty or is Donald Trump smart because he is a billionaire with a model wife half his age yet far from intellectual?

There are a great many smart people at Apple, engineers etc. but they don't keep the company ticking over. That report indicates what I had previously thought that Jobs is more a marketing face for the company but other reports had consistently indicated that certain key individuals reported directly to him and he has been involved in major negotiations.

I think it's a good thing that Cook is more of an unknown but in control. The media will be less likely to tie Apple's success to an individual with him in charge.

One thing I found odd is Steve Jobs' involvement with the financial reports:

http://www.apple.com/pr/library/2009/01/21results.html

Maybe his break starts after that.

I presume that was sarcasm?

I don;t know if this is necessary - artists can use services like CD Baby to get their tracks on iTunes. Perhaps Apple might get rid of the middle man and turn iTunes into a pay-to-own YouTube of sorts, allowing people creating content with Apple products (Logic, Logic Express, Garage Band, iMovie, Final Cut) to post content for sale in the iTunes store.

This could be a HUGE step to win over the people who made Mac important, and have been neglected by Apple with the lack of pro updates. If this is something Apple has planned, I would definitely welcome it, I just hope they have some quality control, or else it could be flooded by people who think that someone will want to spend money for a webcam rant or their cat doing something cute.

I'm afraid somehow farting will involved. As per the iPhone App Store iFartFiesta.

That's a scary but not-too-remote possibility.

I think what is most frustrating for many is that Apple has had Oct, Nov, Dec of 08, Jan 09, ... four months to get the 17" into Unibodyness.

Edit: Oops wrong thread. LOL.

If they can't even deliver on the things they said are ready (17"MBP) then why are people still expecting another announcement?

That's what I said on the 17" thread. No 17", no Mac Mini, no iMac... Analysts are pulling things out of their ass to keep interest in AAPL.

Something is going on at Apple HQ. I am concerned.

Thank you for the response. Any ideas as to what Andy Ihnatko found so jaw dropping? I haven't seen anything from Apple that would be considered jaw dropping.

Hell NO! SJ and Apple just don't give a rap about the 'faithful'. They'd rather play mind games with us. They expect us to be loyal to Apple, but that's a one way street. Isn't that apparent from all the crap they've foisted on us? When has Apple EVER had it's users in mind? They are up in their ivory tower with NO regard to users who keep them in business. Do I sound bitter? You bet your a__ I am.

Do any of you fanboys think my opinion is wrong? Do you think Apple actually cares about us?

Doesn't the support cut off of G4's tell you what Apple thinks of users?

Everything has to end, they cannot keep support for older systems in indefinitely. There may be very good reasons for dropping it like performance gains on the newer systems that break G4 support.

By the reports they are updating the Finder for better performance and that update might break PowerPC support right there. Should they not update core components that benefit the OS just to keep in support older systems?

You have a choice, you don't have to update to the latest system and you G4 will work just fine. Apple needs to advance the OS to keep in competitive, and the extra time writing code to ensure support for systems that are at least 3 years old may not be cost effective for them.

I personally find OpenCL jaw dropping. Some uninformed reporters say things like 10.6 will be 100 times faster, which are quite funny to watch but I believe it will be very impressive.

This is not some advance like auto-vectorization in Alti-vec, this is using the processors in the GPU to do intense calculations and even on the new low end 9400M, it's capable of 54 Gflops, which is double what a Core 2 Duo can do.

It's not that surprising when you think of them as 16 x 450MHz cores vs 2 x 2GHz cores.

Now yes it will be limited to certain tasks but this isn't some language that came from somewhere else, this is Apple's language developed alongside GPU manufacturers. I think they will leverage this a lot at the core level of the OS. Not for use in the OS itself (Finder etc) but in the Core APIs.

They could render text with this, PDFs, encode video or at least process intensive parts of the encoding. Right now, the hardware can handle normal tasks just fine. All machines including the high end Mac Pro struggle when it comes to raw encoding and calculations.

I would bet that every model they bring out will have a GPU that at least equals the CPU in that machine. So the low end Mini will have Core 2 Duo with 9400M. The iMac will use the core 2 quad and a mid-range Nvidia card and the Core i7 Mac Pro should come with the 9800GT with the GTX 280 as BTO. This means for the same price, you effectively get 2-3x the performance for certain tasks than if you didn't have OpenCL. I think that's impressive.

I have a suspicion that Apple could eliminate the 20" iMac and go all 24" and use the 3 mobile quad chips. This means they can share the display panel from the 24" Cinema LED and economize the inventory - it then makes sense why they didn't make a 20" model. If you look at the 2.8GHz Core 2 Extreme 24" iMac, that CPU costs $851. The price of the quad 2.33GHz chip is $245.

This means that a 24" iMac with that chip would drop to:

$1799 - $606 = $1193

The current low end on the iMac line is a 20" display with 2.4GHz Core 2 Duo and is priced: $1199

So to summarize, what I think they will do:

Core 2 Duo Mini, still with mobile chips but 9400M, maybe drop back to the original lower price point.
Core 2 Quad iMac, possibly 9400M on entry model and going up to a mid-range Nvidia 9 series card.
Core i7 Gainestown, possibly 8-core chips but likely dual quads, with a single quad on the lowest end and high end Nvidia 9 series cards with the pro cards too.

I'd like to see them make an entry level headless quad but this is the lineup that seems most likely to me. As I say, combine these hardware specs with OpenCL and I will be impressed. Still annoyed about the lack of the mid-range headless machine as always but I would acknowledge these as very good performance and competitive machines. I will buy a Mini for home and a Pro for work and complain that I would rather buy two middle options.

I think the iMac and Mini updates will come first given that the processors for the Pro probably aren't ready yet although there is a possibility they could organise an event some time this month and launch them all together with redesigned enclosures (the chin is definitely going if it's 24" all the way), then ship the Pro later on.

Apple are also supposed to be working on high end graphics software to replace Shake and their recent update after a number of years suggests they at least haven't forgotten about it. If there was a new Final Cut Studio that was OpenCL enabled and rendering say 4 times faster than the old Mac Pro, it would be very appealling.

This is why rumors are best to be fully detailed because the above is what I find impressive but others will always be looking for gold at the end of the rainbow instead of just stopping to admire the rainbow itself.

There doesn't need to be something that is beyond everyone's imagination. When it comes to highly parallel tasks, we will see as much as 10x increase in performance. This will destroy PC benchmarks for at least a few months.

The G4 is 10 years old now, how long do you expect them to continue to support it?

Apple has never been sentimental for legacy technology. At some point you have to let go of old and antiquated technology, to then support optimized and lightweight technology. This is exactly the same as when Apple switched from OS 9 to OS X.

In my experience, anyone hysterical enough to use terms like 'fanboys' can safely be ignored.

That would be foolish IMO.

While I don't know for fact, I would bet the 20" iMac is the best selling iMac.

This has to be what Andy was so excited about... http://www.macrumors.com/2009/02/08/...stars-tv-show/

Radeon HD 4800 support in Snow Leopard:

http://www.macbidouille.com/news/2009-02-09/#17950

think this is interesting, the 4800 is from 2008... maybe a low power version of this card would be new and good for the next iMac ;-)

It will take a few years for OpenCL to really mature in a lot of Mac OS X apps... But yes, it is promising. Just look at the Folding At Home stats:

There are much less Nvidia-GPU clients but they destroy the CPU clients:
http://fah-web.stanford.edu/cgi-bin/...?qtype=osstats

Yeah hopefully we're talking a lower power version Radeon 4850 or 4870. Maybe Radeon 4650 or better. Radeon 4650 at minimum. But the iMac will very likely use the Nvidia chipset so no Radeons in iMac, I think...

This is the thing we have people over selling what OpenCL will be capable of doing. People that don't understand will be disappointed.

The big but here is that yeah it will be impressive in limited ways. Currently my biggest concern is that Apple back ports GPU acceleration of movie playback on my early 2008 MBP.

Auto-vectorization is a compiler technology. You may be confusing people here because the processors in the GPU are similar to vector processors. Also those GFlops are all but useless if your code and compiler can't fined an arrangement of code and data that these units can execute against. Effectively that is.

Remember that almost never are all of those cores available at once. Do even a little bit of graphics or other tasks that tie up the GPU and your free cores begin to shrink. In fact there is some evidence out that that indicates that one can actually put to much of the work load onto the GPU and actually lower performance. This often involves heavy 3D but the point remains in order for those processors to help they need to be free for use and not impact bandwidth that the GPU needs.

The other problem is just how many of those GPU's is Apple going to leverage? I could actually see them eventually having a couple in use just about all the time. There are several possiblilities here including sound processing, TrueType processing, Much more of Quartz or even all of it. It all depends on just how aggressive Apple wants to be and how well the GPU can actually accelerate these subsystems.

Actually I think you are half right above. Eventually they may accelerate Apple supplied apps themselves. Many of the API's are a given.

In any event i'M surprised that you discount finder as there is good potential there for acceleration. Just rendering all those icons could be an area of improvement.

Yep, lots of potential but I don't see the huge speed ups that some imagine. YOu will see exceptional performance when your app and the API's can make use of all the hardware threads available but you shouldn't assume that every app will leverage those GPU cores.

Potentially get that sort of performance increase. For specific apps though you need to know how or if the app can leverage all those cores. A so called embarrassing parallel app might actually see a 16X speed up, but those will be few and far between.

Well hopefully Apple won't degenerate into the highly questionable benchmarking that was common when the G's where being marketed. OpenCL is great and all but the last thing we need is some sort of garbage bench mark being promoted b Apple the way they did some of the G5 benchmarks. Generally what they did was to concentrate of the good things that maybe didn't have universal usage by Apple hardware buyers.

Along these lines I wouldn't expect Apple to have 10.6 highly accelerated with OpenCL at introduction. In other words what we are likely to see is a slow but stead improvement in the libraries and apps as Apple finds more and more uses for GPU processing. Note that I'm not calling this GPU acceleration because in many instance the benefit will simply be the reality that there is another hardware thread acting in parallel to heavier threads.

Dave

I really hate to see references to a signal processing program like this which marries up to GPU processing all most ideally. The problem is that it sets up an expectation that all programs will benefit in a like manner. This isn't the case in the least. What we should see is a wide range of acceleration from almost zero to performance numbers like Folding at Home imply.

I still see the bigger advantage i the idea that you will have many parallel hardware threads to work with. In some ways it won't matter if the GPU acceleration is a bit slower than the main GPU for some tasks, you benefit more from doing many things at once. At least for some work flows.

In other words; even for tasks where GPU acceleration isn't faster than the main CPU you may get a significant advantage simply from off loading the main CPU. Of course it would be nice if every task that got sent to the GPU ended up running 3X faster or even faster, it just isn't a requirement to be considered an advantage.

Dave

my thoughts are more like:

mac mini - nvidia (desktop version of the MacBooks)
iMac - something "bigger" with quad (real desktop)

but who knows, maybe that's just wishful thinking