***CONFIRMED*** APPLE MOVES TO INTEL - X86!

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Originally posted by: pm
This should make the Intel ZEALOTS happy after such bad news for two years now. Apple better hope Intel improves the performance of the Dothan or their machines will get beat worse than before.
I guess I qualify as a zealot, and whether I do or don't, I do know that I'm happy about Apple using x86 - whoever happens to be the specific manufacturer.

I remember years ago, using one of the original Macs in high school and compared to the Apple IIGS's and the IBM PC's, it seemed like a revolution. Then years go by and I see lots of apps for the Apple that I could wish that I could have. Case in point, iPhoto - which I am seriously impressed with right now and wish that I could find something like it on the PC (Picasa doesn't work well for me). Anyway, anything which unites Apple's stylish usability, with the PC's cost structure and flexibility is a great thing in my mind as a computing enthusiast.

Well if you dont like Zebo or me and we'r just trolls and anandtech cripples your fave chip the pentium M why you here?, p*ss off, go to toms hardware or something.
You know, I am already half way out the door. :( I only read here one or two days a week nowadays and I post far less frequently that I used to.
Nah, admit it, pm... You have a smug smirk on your face as you sit back and watch all the online Mac geeks slowly come to the realization that they have been assimilated. ;)

Anyways, I'm so looking forward to that Yonah PowerBook next year. :)

Ironically, I'm not so looking forward to an x86 iMac until 2008 or something (which is OK, since I just bought one, with a G5). All those compatibility issues are gonna take a while to get ironed out. And Rosetta'd Photoshop didn't impress. Not surprisingly, the Rosetta translated Photoshop takes significantly longer to launch on Steve's P4 3.6 with OS X than it does on my iMac G5 2.0.

BTW, ironically, I'm not a huge fan of iPhoto. It has its uses, but it's not as slick as iTunes IMO. Mind you, I'm not a fan of Picasa either. iPhoto '06 might just be killer though, cuz although I'm not a huge fan, I see iPhoto '05's virtures. It's the first version of iPhoto I'd consider using more than once a month.
 

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
Originally posted by: bjc112
Yea right, you may want to look again. First off you're comparing a mobile CPU to a desktop CPU which in itself is pretty dumb. Secondly, Anand did cripple the Pentium-M running it in the setup he did. Tech-report did a much better review:

"The performance that this motherboard enables speaks for itself, I'd say. As a desktop processor, the Pentium M fares very well. The stock Pentium M 755 at 2GHz rivals the lower speed grades of the Athlon 64 and Pentium 4 fairly consistently. Overclocked to 2.4GHz on a 533MHz bus, though, the Pentium M gets downright scary, shadowing the performance of the Athlon 64 4000+ through many of our tests, including games."

Comparing a overclocked Dothan to a STOCK 4000+ is just as stupid.. Overclock the 4000+ and watch it demolish it even more..

Dothan is a good chip and will definitely get better, excited to see the new dual core ones, once they release...

Sport, stay out of the conversion if you don't know what it is about so you don't make yourself look like an ass. We're not comparing absolute performance. The topic was clock for clock, which performs better. What clock speed does the A64 4000+ run at? 2.4 GHz. What was the OC'd Dothan running at? 2.4Ghz. Dothan was running a 533MHz frontside bus with 356MHz DDR, while the A64 4000+ has a 1000MHz hypertransport and 400MHz DDR, yet the Dothan running on an ancient platform was still able to match and often outperform the A64 4000+. If Intel further bumps the bus for Conroe it will blow by the A64 and leave no doubt which is faster clock for clock. Overclocking the 4000+ wouldn't show anything except that if it performs better, that a higher clocked A64 performs better than a lower clocked Pentium-M. What a ground breaking discovery that would be...
 

bjc112

Lifer
Dec 23, 2000
11,460
0
76
Dothan@2.0 wins:
-----------------
Business Winstone 2004
Doc Office Productivity SYSMark 2004
Mozilla 1.4
Winzip
Adobe Photoshop 7.0.1
MusicMatch Jukebox 7.10
Windows Media Encoder 9
3dsmax 5.1 Dx


A64 3200 939 @2.0 wins:
-------------
Comm Office Productivity SYSMark 2004
Data Office Productivity SYSMark 2004
ACD Systems ACDSee PowerPack 5.0
WinRAR 3.40
MCC Winstone 2004
3D ICC SYSMark 2004
2D ICC SYSMark 2004
Web ICC SYSMark 2004
Mozilla + Media Encoder
Adobe Premier 6.5
Roxio VideoWave Movie Creator 1.5
DivX 5.2.1 with AutoGK
XviD 5 with AutoGK
Windows Media Encoder 9 HD
Doom 3
Wolfenstein: ET
3dsmax 5.1 Ogl
Visual Studio 6
SPECviewperf 8 3dsmax
SPECviewperf 8 CATIA Viewset (catia-01)
SPECviewperf 8 Lightscape Viewset (light-07)
SPECviewperf 8 Maya Viewset (maya-01)
Pro/ENGINEER (proe-03)
SolidWorks Viewset (sw-01)
Unigraphics (ugs-04)

Dothan gets dominated by a 512 cache budget chip, pretty sad really Intel with 1000x the budget getting beat by little AMD, and with K10 AMD's lead should grow even more.

----------------------------

Sport, stay out of the conversion if you don't know what it is about so you don't make yourself look like an ass. We're not comparing absolute performance. The topic was clock for clock, which performs better. What clock speed does the A64 4000+ run at? 2.4 GHz. What was the OC'd Dothan running at? 2.4Ghz. Dothan was running a 533MHz frontside bus with 356MHz DDR, while the A64 4000+ has a 1000MHz hypertransport and 400MHz DDR, yet the Dothan running on an ancient platform was still able to match and often outperform the A64 4000+. If Intel further bumps the bus for Conroe it will blow by the A64 and leave no doubt which is faster clock for clock. Overclocking the 4000+ wouldn't show anything except that if it performs better, that a higher clocked A64 performs better than a lower clocked Pentium-M. What a ground breaking discovery that would be...



It doesn't matter that it's a mobile chip either, if you don't want it compared with desktop CPU's, don't put it in the same lineup.. So of course it will have a slower FSB, and slower overall DDR speed..

Clock for clock, at the present time, the A64 is still faster.

And please, don't use the HyperTransport arugment, because we can knock it all the way down to 400, and performance is damn near the same, Sport!



 

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
You're comparing a 400MHz bus when the Pentium M has a 533MHz version. Why? Looks like you're intentionally trying to cripple the Intel offering to make your viewpoint true. The P-M 770 runs a 2.13GHz on a 533MHz bus. Look how many tests it beats the 3400+ running at 2.4 GHz. It even beats the 4000+ is some tests despite a 267MHz deficit.
 

bjc112

Lifer
Dec 23, 2000
11,460
0
76
Originally posted by: Pariah
You're comparing a 400MHz bus when the Pentium M has a 533MHz version. Why? Looks like you're intentionally trying to cripple the Intel offering to make your viewpoint true. The P-M 770 runs a 2.13GHz on a 533MHz bus. Look how many tests it beats the 3400+ running at 2.4 GHz. It even beats the 4000+ is some tests despite a 267MHz deficit.

Blah..

The 3400+ isn't even Dual channel. And runs a lower HyperTransport speed, which you portray as a big deal..
 

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
After a recount, the 2.13GHz P-M beats the 3400+ running at 2.4GHz in 15 out of 25 non-overclocked tests.

A clear win for the Pentium-M.
 

bjc112

Lifer
Dec 23, 2000
11,460
0
76
Now, granted I didn't see any 533's on newegg, and assume the prices are the same, if not higher..

@ $600, I'll take my chances on the 3200+

And please don't argue it's a mobile chip.. Because again, if you don't want it compared on all aspects, don't put it in the same lineup...





I believe that's a clear win for the 3200+
 

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
I miscounted. I was using Anand's review, since that's the one Zebo quoted.

Out of 33 tests, the 770 won 16 and the single channel 3400+ (2.4GHz vs 2.13GHz) won 17.

Pretty even, once the clock speed is taken into consideration, advantage P-M.

Vs the dual channel slightly slower 3200+ (2GHz va 2.13GHz), the 770 won 28 out of 33 test.

Not much to conclude from that as the 770 should have been faster in the vast majority of tests.

The 2 architectures look pretty evenly matched clock for clock. If Intel can bring the P-M into the 21st century with the release of Conroe, whatever you call the architecture, it should return the undisputed speed crown back to Intel.
 

bjc112

Lifer
Dec 23, 2000
11,460
0
76
Originally posted by: Pariah
I miscounted. I was using Anand's review, since that's the one Zebo quoted.

Out of 33 tests, the 770 won 16 and the single channel 3400+ (2.4GHz vs 2.13GHz) won 17.

LOL, I am still confused, I thought Anand's review was slanted, and the Tech Report was better?

The 3200+ seems to best the 2.0 by quite a bit in the TEch Report... And running 533FSB..

:confused:


If reading the Anand review that Zebo quoted, it was 25-8.. And that's @ almost 1/3 of the price.. Even 1/2 to be generous..


But I do agree that the updated core will be awesome, and can't wait for it tob e released..





 

Pariah

Elite Member
Apr 16, 2000
7,357
20
81
Originally posted by: bjc112
Now, granted I didn't see any 533's on newegg, and assume the prices are the same, if not higher..

@ $600, I'll take my chances on the 3200+

And please don't argue it's a mobile chip.. Because again, if you don't want it compared on all aspects, don't put it in the same lineup...





I believe that's a clear win for the 3200+

Yes, but the Pentium-M packaging looks a whole lot cooler which is just as relevant when comparing clock for clock performance.
 

bjc112

Lifer
Dec 23, 2000
11,460
0
76
Originally posted by: Pariah
Originally posted by: bjc112
Now, granted I didn't see any 533's on newegg, and assume the prices are the same, if not higher..

@ $600, I'll take my chances on the 3200+

And please don't argue it's a mobile chip.. Because again, if you don't want it compared on all aspects, don't put it in the same lineup...



I believe that's a clear win for the 3200+

Yes, but the Pentium-M packaging looks a whole lot cooler which is just as relevant when comparing clock for clock performance.



Come on now.............. SPORT ;)

 

Childs

Lifer
Jul 9, 2000
11,450
7
81
Originally posted by: miniMUNCH
Software porting (if it could even be called that) will not be that bad at all. Many, Many Cocoa programs will compile on x86 and run without error. Others will require some clean up but we are talking about 5% or less of the code that would need modification. In the case of high performance codes where significant optimizations for PPC are hard coded rewrites will be required...but I can't think of too many pieces of HP software that are Mac only.

This will be tougher than you think. Jobs's demo was a bit misleading. You only saw the functions in the apps that worked. Getting it to run is one thing, optimizations and no loss of functionality will be harder. Hopefully Intel will beef up the Vector in SSE and general FPU. Porting the software with no new features will still take time. Making sure everything works, and works at exceptable speeds. Remember, Apple still has to support all the existing OSes and hardware. They basically doubled QA and engineering with no increase in resources. Rough times ahead.

The software I'm really thinking about is the Apple software...FC Studio, Shake, etc, and any Mac only software that relies heavily on AltiVec or low level code.

The Intel vs. AMD thing comes down to two things: comparing Intel and AMD performance now can be a bit misleading. Apple is looking at desktop, server chip performance 1.5+ years from now...many folks think Intel has some seriously ass kicking chips lined up on that timeline.

Second, AMD is pretty much just chips. Intel is chips, chipsets, MB's, wireless, HD audio... near complete system design. Apple is sick and tired of having to carry most of the load of designing their own MB's and writing their own drivers. They just don't have the volume of sales to justify custom MB design/production...the design and tooling overhead is too high. Intel can and will provide, at a minimum of design and tooling cost, a robust, slightly customized, high-tech MB with solid drivers at a price point that Apple could only dream about previously.

I doubt Apple will use Intel's standard chipsets. Apple has a lot of hardware engineers, and I'm sure they'll do something to differentiate themselves from PCs you can get at Best Buy or from Dell. Until they fire those engineers, I think Apple will actually design their own chipsets. They already have the expertise, and its not uncommon considering the number of alternate chips on the Intel side. I haven't looked at the Intel Road Maps lately, so maybe their chips will be more competetive with AMD. If done right however, switching between the two may not be that difficult ala NF4.

So maybe Apple will pay more for chips relative to their IBM discount...but Apple's internal hardware engineering costs and MB prices will be reduced significantly.

I don't think Apple will go the generic PC route anytime soon. In fact, I wouldn't be surprised if the prices didn't change all that much. They will still claim they have the fastest machines on the planet. The performance difference maybe the same as an Intel reference system vs NF4, but there will have to be differences. They are already claiming you won't be able to install OSX on regular PCs, and if they have their own chipset, then its easy to see why. And it would be easy for Apple to claim these new macs can run Windows if they write the drivers.

Its funny, but I'm been screaming for this switch for awhile. Now that it happened it doesn't seem like that great of a thing...probably because I use Apple hardware/software more than I ever have in my career.



 

CrimsonChaos

Senior member
Mar 28, 2005
551
0
0
Didn't Steve Jobbs try this before? Remember his little PC line called "NeXT"?

Let's see, if memory serves me correctly.... NeXT was an x86-based machine. And oh yeah, it was also a UNIX-based machine.

If at first you don't succeed, try try again...

Believe me, I'm no Microsoft fan. In fact, I actually bought a NeXT back in college (1991 or so). But their extremely short lifespan left me rather bitter toward Steve Jobbs. I love UNIX, and I really do hope this succeeds, because frankly I can't stand Windows.

I just wish he would've went with AMD.
 

Pabster

Lifer
Apr 15, 2001
16,987
1
0
Originally posted by: bjc112
Comparing a overclocked Dothan to a STOCK 4000+ is just as stupid.. Overclock the 4000+ and watch it demolish it even more..

LOL. I never took you for an AMD zealot bjc112, but crap like that is going to change my mind.

The comparison was fair because Dothan was clocked up to match (as closely as possible -- note Dothan STILL with a far slower bus) the A64. And surprisingly, it was able to BEST the A64 in many areas. This despite two HUGE disadvantages -- single channel DDR -- and a slow FSB. The A64 has dual-channel as well as HTT running much faster than the FSB on Dothan.

This wasn't a desktop against mobile comparison or anything of the sort. It was designed to illustrate CLOCK-FOR-CLOCK performance of these processors. Clearly, Dothan IS faster than Athlon64 at an equal clock -- despite the huge disadvantages facing it.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Originally posted by: CrimsonChaos
Didn't Steve Jobbs try this before? Remember his little PC line called "NeXT"?

Let's see, if memory serves me correctly.... NeXT was an x86-based machine. And oh yeah, it was also a UNIX-based machine.

If at first you don't succeed, try try again...

Believe me, I'm no Microsoft fan. In fact, I actually bought a NeXT back in college (1991 or so). But their extremely short lifespan left me rather bitter toward Steve Jobbs. I love UNIX, and I really do hope this succeeds, because frankly I can't stand Windows.

I just wish he would've went with AMD.
NeXT the OS was cross platform.

In fact NeXT is what OS X is built from. They even released Rhapsody as a developer preview. Rhapsody was the precursor to OS X, but the GUI looked like OS 9ish, and it ran on x86. However, nobody seemed interested.

IOW, I think Steve Jobs wanted to switch to x86 before OS X was even released, but his hands were tied by the software developers. Plus he had the deal cooking with IBM, and he hoped IBM would come through. IBM didn't, so Steve "switched".
 

Anubis

No Lifer
Aug 31, 2001
78,716
417
126
tbqhwy.com
why does it only have to work with 1 type of CPU?

why cant OSX work with any x86 CPU like windows or linux does?
 

jose

Platinum Member
Oct 11, 1999
2,076
0
0
Eug,

Is there any info on the operating system that will be used when they switch ? Is apple sticking w/ its current version or they changing to windows or Linux ?

Regards,
Jose
 

digitalsm

Diamond Member
Jul 11, 2003
5,253
0
0
Originally posted by: jose
Eug,

Is there any info on the operating system that will be used when they switch ? Is apple sticking w/ its current version or they changing to windows or Linux ?

Regards,
Jose

They will continue to develop their own OS, which is a good alternative to Windows. It will be able to compete better vs Windows than Linux ever will.
 

dullard

Elite Member
May 21, 2001
25,059
3,410
126
Originally posted by: Anubis
why does it only have to work with 1 type of CPU?

why cant OSX work with any x86 CPU like windows or linux does?
Most likely, Apple will be using a proprietary motherboard that happens to accept an Intel chip. AMD chips just don't fit in the same slot anymore. AMD simply cannot meet Apple's production needs (even though it is meager).
 

Anubis

No Lifer
Aug 31, 2001
78,716
417
126
tbqhwy.com
Originally posted by: dullard
Originally posted by: Anubis
why does it only have to work with 1 type of CPU?

why cant OSX work with any x86 CPU like windows or linux does?
Most likely, Apple will be using a proprietary motherboard that happens to accept an Intel chip. AMD chips just don't fit in the same slot anymore. AMD simply cannot meet Apple's production needs (even though it is meager).

yea i know they only fit 1 type of mobo, but again i ask why limit it to a propriety anything

IMO it woule be much better to have it work with any cpu and any hardware so there would essentially be a 3rd OS as a choice
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Originally posted by: jose
Eug,

Is there any info on the operating system that will be used when they switch ? Is apple sticking w/ its current version or they changing to windows or Linux ?

Regards,
Jose
??? They demo'd Tiger running on a P4 3.6. Apple running Windows or Linux is not Apple.

Anyways, Maxon has just announced that Cinema4D is ALREADY running on OS X x86. It's a fat binary, that supports both the G5 and the P4.