[chris v]

Originally Posted by rozwado1

Not a direct 'NO', but more than one salesman teaming up and letting the guy know that it wasn't a good idea to purchase a G5 now.

If I was a shareholder, I would have made a scene.

Jesus, what dumbf*cks. There's not a damn thing wrong with the G5 towers right now. The quad is a friggin' beast. They should be FIRED!

[rozwado1]

Originally Posted by chris v

Jesus, what dumbf*cks. There's not a damn thing wrong with the G5 towers right now. The quad is a friggin' beast. They should be FIRED!

it's been my observation that the majority of apple retial workers are dumbf*cks. I feel bad for the uninformed people that go in there for tech advice.

Long live 'NN.

[isao bered]

? personally, i'm not yet seeing much problem with the rosetta emulation. the software manufacturers that i've purchased from certainly make no performance guarantees - which is understandable; however, but i've yet to experience any problems attributable to the emulation. so far i am much more impressed with this transition than the one from os9 to osx.

is my dc as fast as an imac g5 for any given application? i don't know and don't really care. i do know that even under emulation it is tremendously faster than what it is replacing and that is all that matters to *me*.

why did i adopt early? firstly, because for a variety of personal reasons it was appropriate timing. secondly, for me it is only a personal computer and not the end of my livelihood or my life if there are some growing pains. for those whose livelihood is at stake it probably is more prudent to wait it out some. so i can't run my filters as fast as someone else - i'm not in a contest or competition. but since it seems relatively certain the direction things are going i've decided to go ahead and get on the path now.

other early adopters please consider contacting the software manufacturer's of your emulated applications. they may be interested in your experiences in running their programs on the intel machines - i know a couple of mine are.

be well.

laeth

[goMac]

Originally Posted by CharlesS

Maybe people are pissed off because the Intel Macs chop the Mac's backward compatibility from 22 years down to only 5. Maybe if Apple would integrate the PowerPC emulation from SheepShaver into Classic, it would win over more users (especially educational customers). Sure, they'd have to GPL Classic (note: not OS 9, just the Classic environment itself) if they did this, but under the current plan it'll be effectively killed off anyway when PowerPC eventually becomes no longer supported by OS X, so who cares?

Microsoft Windows: Prime example of what happens when you try to keep backwards compatibility going too long.

[Chuckit]

Originally Posted by CharlesS

Maybe people are pissed off because the Intel Macs chop the Mac's backward compatibility from 22 years down to only 5.

I have never had 22-year-old software run on my PowerBook G4. The ones I have tried have failed. Can you give me an example of one which with current Macs are compatible?

[CharlesS]

Originally Posted by Chuckit

I have never had 22-year-old software run on my PowerBook G4. The ones I have tried have failed. Can you give me an example of one which with current Macs are compatible?

StuntCopter isn't quite 22 years old, but it's close, and it works, although without sound.

Granted, there's a lot of old stuff from the 80s that doesn't run, but a lot of that is the fault of the app, not the OS. A lot of them cut a lot of corners to try to fit into the tiny amount of disk space and memory available on the 128K Mac. But there's a surprising amount of System 6 software that works fine.

[CharlesS]

Originally Posted by goMac

Microsoft Windows: Prime example of what happens when you try to keep backwards compatibility going too long.

Uh huh. Because adding a PPC emulator to the Classic environment and keeping it around would surely make OS X just like Windows. Even though it's not installed by default, even if it's not physically present on your hard disk, its GHOST would HAUNT your system!!!

[goMac]

Originally Posted by CharlesS

Uh huh. Because adding a PPC emulator to the Classic environment and keeping it around would surely make OS X just like Windows. Even though it's not installed by default, even if it's not physically present on your hard disk, its GHOST would HAUNT your system!!!

Why Shapeshifter? Apple already has a PowerPC emulator on the system.

I would think it's obvious why Classic was dropped. Apple doesn't want to have to supply OS 9 with all new Macs as a separate OS.

[CharlesS]

Originally Posted by goMac

Why Shapeshifter? Apple already has a PowerPC emulator on the system.

I'm assuming that Rosetta, being a dynamic binary translator, wouldn't work to make a virtual machine running an entire OS à la Classic. Maybe I'm wrong. Either way, it would have been within the realm of possibility for Apple to do it.

I would think it's obvious why Classic was dropped. Apple doesn't want to have to supply OS 9 with all new Macs as a separate OS.

They don't package it as a separate OS currently. They supply OS 9 as part of an installer for the Classic Environment. I don't think it blesses the System Folder it puts in, so I don't think it would work as an installer for a separate OS anyway.

[MindFad]

Originally Posted by greenamp

Why should they have re-rendered the MacBook when it looks almost completely identical to the powerbook?

Render? Like, 3-D model render, is what everyone is saying?

If that's what you mean, here—I'll go ahead and say it: no Apple product promo image has ever, EVER been rendered using a 3-D model. It is photography. All of it.

That is all.

[Leia's Left Bun]

Originally Posted by MindFad

Render? Like, 3-D model render, is what everyone is saying?

If that's what you mean, here—I'll go ahead and say it: no Apple product promo image has ever, EVER been rendered using a 3-D model. It is photography. All of it.

That is all.

Why are you sure about that?

[goMac]

Originally Posted by CharlesS

I'm assuming that Rosetta, being a dynamic binary translator, wouldn't work to make a virtual machine running an entire OS à la Classic. Maybe I'm wrong. Either way, it would have been within the realm of possibility for Apple to do it.


They don't package it as a separate OS currently. They supply OS 9 as part of an installer for the Classic Environment. I don't think it blesses the System Folder it puts in, so I don't think it would work as an installer for a separate OS anyway.

I would assume they could somehow bind Bluebox into Rosetta, feeding Rosetta the PPC calls that it's VM makes.

They probably don't want to have to supply a Classic environment installer anymore.

[Chuckit]

Originally Posted by goMac

Why Shapeshifter? Apple already has a PowerPC emulator on the system.

It has a translator for PowerPC binaries. I don't think it fully emulates a PowerPC.

[greenamp]

Originally Posted by MindFad

Render? Like, 3-D model render, is what everyone is saying?

If that's what you mean, here—I'll go ahead and say it: no Apple product promo image has ever, EVER been rendered using a 3-D model. It is photography. All of it.

That is all.

I have no idea if they do or not. My comment was purely and obviously hypothetical.

[analogika]

Originally Posted by Chuckit

I have never had 22-year-old software run on my PowerBook G4. The ones I have tried have failed. Can you give me an example of one which with current Macs are compatible?

Crystal Quest, Daleks, Tetris 1.1, Lode Runner, Shufflepuck, Shufflepuck Café, Leather Goddesses of Phobos, Banzai, Artillery... There's more, and they're not *all* from 1984, but most of those still work in Classic.

Unfortunately, most of them lost sound support when we switched to the G3 PowerPC. But they still work fine (of course, stuff like Shufflepuck and Crystalquest gets difficult with mousing about, since we can't actually limit our resolution to 512x384 anymore, but that's not the point.

[goMac]

Originally Posted by Chuckit

It has a translator for PowerPC binaries. I don't think it fully emulates a PowerPC.

Right. But Bluebox could be regeared to use it and feed the PowerPC instructions from it's guest OS into Rosetta.

[Railroader]

I'd be surprised if Apple is around next year.

[analogika]

Originally Posted by Railroader

I'd be surprised if Apple is around next year.

[Big Mac]

Originally Posted by Eug Wanker

Think of it this way... As of January 2006, an iMac is as fast as a dual Power Mac for native applications, costs less, and comes with a free screen.

You and goMac are both making this claim, and every time I read it I have to call bullshi†. Show us the application benchmarks if you wish to contend otherwise. The Intel iMac is not on par with the Power Macs, and it's damn annoying to read this empty Intel fanboy propaganda.

[Gee4orce]

The sky is falling !

The sky is falling !

[Jan Van Boghout]

Originally Posted by Gee4orce

The sky is falling !

The sky is falling !

Will it help if I put a paper bag over my head?!

[Eug Wanker]

Originally Posted by Big Mac

You and goMac are both making this claim, and every time I read it I have to call bullshi†. Show us the application benchmarks if you wish to contend otherwise. The Intel iMac is not on par with the Power Macs, and it's damn annoying to read this empty Intel fanboy propaganda.

It has been well known for years that on average Pentium M keeps up more or less with the G5, clock for clock. It's not a big jump to think that a dual Pentium M design would be competitive with a dual G5.

Anyways, since you are willing to claim bullshi† without bothering to look this up yourself, below are some benches of actual Core Duo and G5 Macs. (Note that some of the descriptions got a bit muddled in the translation, but you can just look at the graphs.)

1.83 GHz iMac Core Duo vs. 1.9 GHz iMac G5 vs. 2.5 GHz Power Mac Quad








As you can see, many benches do in fact show a Core Duo blows the single-core G5 out of the water. However, in many unoptimized applications, it doesn't. Then again, even a quad may not do all that well either if the application is not optimized for multi-processor.

Originally Posted by greenamp

I have no idea if they do or not. My comment was purely and obviously hypothetical.

[Leia's Left Bun]

Originally Posted by Eug Wanker

As you can see, many benches do in fact show a Core Duo blows the single-core G5 out of the water.

Um, for most of those benchmarks it is only about twice as fast not 3-4x like Steve went on.

Also, you are putting a single chip against a dual chip and that doesn't seem fair. But it against a dual G5 and it won't do so well.

[Eug Wanker]

Originally Posted by Leia's Left Bun

Um, for most of those benchmarks it is only about twice as fast not 3-4x like Steve went on.

"iMac has always made it fast and easy to do the most amazing things. Now the fast and easy part literally doubles overnight — because the newest iMac computers are powered by the Intel Core Duo. This revolutionary bit of technology is actually two processors built into a single chip, giving iMac up to twice the horsepower it had previously. So the wows will come faster than ever before."



You're probably thinking about the MacBook Pro. And indeed, for a some apps it would be over 3X as fast, since the G4 is very weak at certain types of code. For example, Cinebench rendering:

1x G4 1.7: 150
1x G4 2.0: 170
1x Pentium M 1.8: 241
1x Pentium M 2.0: 266
1x G5 2.0: 286
1x G5 2.5: 359
2x G5 2.0: 522
Yonah 2.0: 548
2x G5 2.5: 655
4x G5 2.5: 1101

In other words, a 1.83 GHz Core Duo MacBook Pro would have over 3X the performance of a single G4 1.67 PowerBook in this app.

Also, you are putting a single chip against a dual chip and that doesn't seem fair. But it against a dual G5 and it won't do so well.

There's a single G5 1.9 and a Quad G5 2.5. The iMac Core Duo 1.83 falls in between, not surprisingly. Also, see above.

[TETENAL]

[analogika]

Originally Posted by Big Mac

You and goMac are both making this claim, and every time I read it I have to call bullshi†. Show us the application benchmarks if you wish to contend otherwise. The Intel iMac is not on par with the Power Macs, and it's damn annoying to read this empty Intel fanboy propaganda.

Dude, you've been spewing empty PowerPC fanboy propaganda since the very first day the Intel switch was announced.

Even back then, the average Intel laptop was already faster than the most high-end portable Apple had to offer. Except the Intel side has actually seen progress since then.

Nothing you have claimed so far has been true, except that Intel is some sort of evil empire. Dare I remind you that so was IBM?

[Eug Wanker]

Originally Posted by TETENAL

Point taken.

I agree. The "3X" claim would be very... err... optimistic. To claim "up to twice" (like the iMac site says) is quite reasonable however, even if many apps are not multi-processor optimized.

[Leia's Left Bun]

Originally Posted by TETENAL

Exactly, burn on Eug. I thought I also saw a 3-4x during the keynote.

Also that slide says 2-3 times as fast not UP TO 2-3 times as fast. By the benchmarks posted some are only 10% faster on optimized apps which is really really bad considering it is 2 chips against one.

[goMac]

Originally Posted by Eug Wanker

Point taken.

I agree. The "3X" claim would be very... err... optimistic. To claim "up to twice" (like the iMac site says) is quite reasonable however, even if many apps are not multi-processor optimized.

Right, for an application that can fully use both cores it would be 3x faster.

[Eug Wanker]

Originally Posted by goMac

Right, for an application that can fully use both cores it would be 3x faster.

Yonah clock for clock is in the same ballpark performance-wise as Dual-core G5. Sometimes Yonah is faster, but It would be uncommon to see Yonah Core Duo running 50% faster than dual-core G5 970MP at the same clock speed.

Thus, claiming Yonah is twice as fast as single-core G5 is reasonable, but claiming Yonah is 3X as fast as single-core G5 is really stretching.

[Chuckit]

Originally Posted by Leia's Left Bun

By the benchmarks posted some are only 10% faster on optimized apps which is really really bad considering it is 2 chips against one.

Some programs are just poorly suited to multiprocessing, so the 2 vs. 1 thing is not necessarily relevant. If a program is optimized for a dual processor machine and doesn't come up much better than on a single-processor system, odds are that's what you're looking at. Like I said before, the real world is an awful big place.

[Leia's Left Bun]

"Based on our initial tests, the new Core-Duo-based iMac seems to be 10 to 20 percent percent faster than its predecessor when it comes to native applications, with some select tasks showing improvement above and beyond that."

http://www.macworld.com/2006/01/feat...est1/index.php

The rest of the benchmarks don't come near to 3x faster but only 1.2x on average.

That is ****ing pathetic. Good thing Apple isn't using single core chips from Intel or it would be hard to convince anyone moving to intel and having it 50% slower is a good move.

[Eug Wanker]

Why do you keep repeating the same Macworld tests over and over again? And why don't you make these same arguments about dual Power Macs over G5 iMacs? It's essentially the exact same argument.

Your argument doesn't prove that a Yonah core is worse than a G5 core. It proves that dual-core isn't hugely helpful when apps haven't been optimized for dual-core.

For example, iMovie:



Note the scores of the Power Mac G4 1.42 vs. the Mac mini G4 1.42 (shorter is better):

Single Mac mini G4 1.42: 88
Single Power Mac G4 1.42: 85
Dual Power Mac G4 1.42: 83

[goMac]

Originally Posted by Leia's Left Bun

"Based on our initial tests, the new Core-Duo-based iMac seems to be 10 to 20 percent percent faster than its predecessor when it comes to native applications, with some select tasks showing improvement above and beyond that."

http://www.macworld.com/2006/01/feat...est1/index.php

The rest of the benchmarks don't come near to 3x faster but only 1.2x on average.

That is ****ing pathetic. Good thing Apple isn't using single core chips from Intel or it would be hard to convince anyone moving to intel and having it 50% slower is a good move.

MacWorld is using tests that use other components like the hard drive. These aren't processor only tests. The test will get dragged down by stuff like the hard drive. If they really want to verify Apple's claims, they should use a pure CPU test. They're also using a QuickTime test which creates two problems: a) you're writing to the hard drive, this limits your benchmarking of the CPU to the speed of the drive and b) QuickTime's Altivec functions probably haven't been traded out for their SSE equivalents yet.

I've been optimizing my code on these things since WWDC. They really do fly, especially when you start using that second core.

[Leia's Left Bun]

Originally Posted by Eug Wanker

Why do you keep repeating the same Macworld tests over and over again? And why don't you make these same arguments about dual Power Macs over G5 iMacs? It's essentially the exact same argument.

Where else have I posted them repeatedly? Can you post some links that show real world tests matching Apples claims?

At any rate, yes hard drive speed etc. make a difference. Unfortunately that is just part of using a computer in real world tests. If you use iTunes and iMovie there is no way for you to set your computer just to use the CPU and nothing else. As much as I would like it to iDVD will not ignore my hard drive and use Intel chips to the max.

[TETENAL]

You are forgetting that if you are running a program that is not optimized for for multiple processors and you are running two of them at the same time they are almost as fast as when run alone on a dual core machine while they are half as fast when run on a single processor machine. This is practically never considered in benchmarks, but it's real world reality for some to run multiple applications at once (other people work in Photoshop only and Photoshop only, so it might not matter as much to those).

[production_coordinator]

Originally Posted by Leia's Left Bun

Exactly, burn on Eug. I thought I also saw a 3-4x during the keynote.

Also that slide says 2-3 times as fast not UP TO 2-3 times as fast. By the benchmarks posted some are only 10% faster on optimized apps which is really really bad considering it is 2 chips against one.

YOU WEREN'T LISTENING TO THE KEYNOTE:

Steve was talking about the CPU being 2-3 times faster... the CPU. The test they ran were on raw CPU speed (SPECint and SPECfp)... that's where he came up with the number.

IN FACT... he even mumbled something like "Oh, well the Hard Drives are the same..." which means that the CPU is faster, but many of the components (RAM, Hard Drive, etc. etc.) haven't changed.

[Eug Wanker]

Originally Posted by Leia's Left Bun

Where else have I posted them repeatedly? Can you post some links that show real world tests matching Apples claims?

At any rate, yes hard drive speed etc. make a difference. Unfortunately that is just part of using a computer in real world tests. If you use iTunes and iMovie there is no way for you to set your computer just to use the CPU and nothing else. As much as I would like it to iDVD will not ignore my hard drive and use Intel chips to the max.

Well, as you probably already know, dual-core is also much more responsive UI wise.

Even if certain actions in iLife are not sped-up much on dual Power Macs, those same dual G5 2.0 Power Macs "feel" much faster when working complex projects than my iMac G5 2.0. Expect the same UI feel for dual-core iMacs. (I tried testing this out the other day, but CPUsed only has G5 iMacs on display right now.)

This is similar to TETENAL's statement in a way. Multitasking rocks on multi-processor machines. The reason I bought a G5 iMac though was because I didn't want to spend the extra coin on a Power Mac plus monitor, and I didn't want to use up all that space either. The iMac is the form factor I want, and having a dual-core iMac is perfect.

So, back to my previous statement... For native applications, the Intel iMac is on par with the low end Dual G5 Power Macs, but costs less, and includes a screen.

P.S. I'm very impressed by the Handbrake performance tests. Handbrake encoding is not hard drive or DVD speed limited, and is slow as molasses on my G5 iMac. However, it is twice as fast on dual-core machines, because it is well-optimized for dual-core, and it performs extremely well on Yonah.



It seems to be more than twice as fast compared to the G5, clock for clock, and it's a real-world app I use. It's even getting close to gomac's (and Steve's) 3X claim. The frame rate in this test is around 20 for the iMac Core Duo, and only around 7ish for the iMac G5.

[goMac]

Originally Posted by production_coordinator

YOU WEREN'T LISTENING TO THE KEYNOTE:

Steve was talking about the CPU being 2-3 times faster... the CPU. The test they ran were on raw CPU speed (SPECint and SPECfp)... that's where he came up with the number.

IN FACT... he even mumbled something like "Oh, well the Hard Drives are the same..." which means that the CPU is faster, but many of the components (RAM, Hard Drive, etc. etc.) haven't changed.

SWG: Steve Jobs mentioned in the keynote that the hard drive would drag that sort of thing down.

However, think about games. They don't use the hard drive that much. They could easily be 2-3 times faster.

[Leia's Left Bun]

Originally Posted by production_coordinator

YOU WEREN'T LISTENING TO THE KEYNOTE:

Steve was talking about the CPU being 2-3 times faster... the CPU. The test they ran were on raw CPU speed (SPECint and SPECfp)... that's where he came up with the number.

WHO THE HELL CARES if you can't use this speed.

It is like buying a car because it has an engine that can do 300MPH but the tires will blow up at 100MPH.

[goMac]

Originally Posted by Leia's Left Bun

WHO THE HELL CARES if you can't use this speed.

It is like buying a car because it has an engine that can do 300MPH but the tires will blow up at 100MPH.

You can. I already noted how above.

Also, Photoshop and 3D renderers would be able to use the speed without being dragged down by the drive.

[Leia's Left Bun]

Originally Posted by Eug Wanker

Well, as you probably already know, dual-core is also much more responsive UI wise.

I can't verify that because I haven't used one. Do people who say this take the better video card into account? I doubt that highly.

I also think the GUI speed is one of the least important right now as the GUI is more than responsive on a G5 with Core Video. People who use their computer for work care more about encoding speed and not window resizing and counting bounces for apps to launch.

[Eug Wanker]

Originally Posted by Leia's Left Bun

I can't verify that because I haven't used one. Do people who say this take the better video card into account? I doubt that highly.

I also think the GUI speed is one of the least important right now as the GUI is more than responsive on a G5 with Core Video. People who use their computer for work care more about encoding speed and not window resizing and counting bounces for apps to launch.

Not just GUI rendering speed, but multitasking responsiveness. It's much, much nicer on a dual, for obvious reasons. I've been lusting after a dual for just about forever, but for various reasons I haven't gotten a dual, the most important of which are size and cost (of the Power Macs).

Now, we iMac types can get dual too, and that's truly a step forward for Apple customers. Well, once we get past the transition that is.

[Leia's Left Bun]

Originally Posted by Eug Wanker

Not just GUI rendering speed, but multitasking responsiveness. It's much, much nicer on a dual, for obvious reasons. I've been lusting after a dual for just about forever, but for various reasons I haven't gotten a dual, the most important of which are size and cost (of the Power Macs).

I guess I am just spoiled with my 2.5 year old Dual G5. Shame it took that long before it is affordable to the lower end market.

[production_coordinator]

Typical Leia's Left Bun.

First you post an opinion piece regarding iMac sales... (notice how it was all hearsay, turning "lower" into "sluggish" for a OMG!1! factor)

In your 2nd post, you say:

I had 3 people express interest in the Intel Mac's and I told them to wait to see what real world was as I knew those stupid Benchmarks Steve was spewing out were total bullshit.

And we have since proven you wrong.

The Ars benchmarks are correct... Steve's benchmarks were correct [granted, probably a little marketing thrown in there]... but you simply don't understand the benchmarks.

So to review
- The CPU is 2-3 times faster.
- The Computer is anywhere from 50% slower to 50% faster depending upon a number of items.

A Clue:
The fact that many programs (Such as Photoshop) are running near native speeds in emulation says VOLUMES about how bright our future may be.

WHO THE HELL CARES if you can't use this speed.

It is like buying a car because it has an engine that can do 300MPH but the tires will blow up at 100MPH.

YET again... you don't understand. We will get 2x-3x speed improvements with cpu intensive applications.

Example:
http://media.arstechnica.com/reviews...dia/itunes.png

The iMac Intel nearly beat the Dual G5 (with 5x the ram). Considering Quicktime, iTunes, OS X, etc. will only get faster with Intel optimization... the future looks good. PS, the PowerMac has a 2X faster optical drive.

Not bad IMHO. Can't wait to see OS 10.5 is like... hopefully with Intel optimizations for all core technologies.

[Eug Wanker]

Originally Posted by Leia's Left Bun

I guess I am just spoiled with my 2.5 year old Dual G5. Shame it took that long before it is affordable to the lower end market.

Yup. You definitely got your money's worth.

[Eug Wanker]

Originally Posted by Leia's Left Bun

Also, you are putting a single chip against a dual chip and that doesn't seem fair. But it against a dual G5 and it won't do so well.

Here ya go.





The iMac Core Duo 1.83 is faster than a Dual-core G5 2.3 at Handbrake, but slower at iDVD.

[analogika]

Quoth Steve Jobs:

1:07:00
"So what's different? Well what's different is that The new iMac Intel processor is two to three times faster than the iMac G5."

1:08:38
"On the most important computer benchmarks: Two to Three X. Now everything isn't going to run two to three X faster; you know the disks aren't two to three X faster etcetera, but on the most important benchmarks: Two to three times faster."

All of which is true.

[Leia's Left Bun]

Oh oh it is Think Secret vs. Appleinsider:
http://www.appleinsider.com/article.php?id=1490

They were both 100% WRONG for Macworld this year.

[goMac]

Originally Posted by Leia's Left Bun

Oh oh it is Think Secret vs. Appleinsider:
http://www.appleinsider.com/article.php?id=1490

They were both 100% WRONG for Macworld this year.

Actually it's ThinkSecret vs. American Technology Research.

Guess which one I'm inclined to believe.