Intel Mac and Torque

#1 Hum well the Endian is changing. Good question I wonder if things need to be recompiled?

#2 Most likely the platform later will change just like porting TGE to any architecture.

#3 It took Paul Scott and Jay Moore 15 minutes while on the floor at MacWorld to recompile Torque for intel based Macs. We should have an official solution ready quite quickly I would imagine once we confirm nothing else is needed.

#4 Hum. Seems like Apple could ship an OSX for all X86 pc's if they wanted too lol? Would be interesting...

#5 I doubt apple would want to let OSX run on PCs as they would lose the revenue from there hardware sales.
They tried that once before ( sort of ) when they let other companies manufacture Mac clones, and they sold them the license for MacOS. They changed there minds after a year or so because no one was buying there own Macs.
I'm sure that apple would prefer that you bought a Mac and put Windows on it, than bought a PC and put OSX on it.

#6 I did some more research and wanted to caveat a bit what I said: the process is not just us (obviously), but all of the supporting libs that we use (specifically theora, unicode, ogg, etc) all have to be vetted for the new system before Torque is going to be declared "Mac-Intel ready", so don't expect the "quite quickly" I stated above.

#7 Did you see that new Apple commercial "now your intel cpu is free"....

Yea right more like the risc deathgrip on Mac's is now dead... They LOST the war...

Yea i'll adopt the enemy's CPU tech and claim we are liberating it. Brilliant :)

#8 I like the new commercial.

"Dull little boxes, performing dull little tasks" just about sums up the Windows world!

Back to the subject... I don't expect major delays in porting Ogg Vorbis & Theora for Intel Macs.

#9

Quote:Yea right more like the risc deathgrip on Mac's is now dead... They LOST the war...

Of course risc seems to have won the war on other fronts, considering that all 3 of the next gen consoles are using G5 variants...

#10 It's too bad x86 won the war. I don't like x86 architecture or the assembly language, but what can you do.

Another reason I don't expect OSX to run on all x86 PCs is that with the myriad of hardware out there, Apple would no-longer be able to maintain their standards of quality for user experiences.

#11 I suspect that one major factor for the move to Intel chips is that Apple wasn't able to get the G5 to run cool enough for their portable lines. Have you seen the cooling units for the Powermac G5's? They're massive! No way a G5 could ever safely run in a laptop.

#12 Whats wrong with Assembly? It's the cornerstone of Modern Civilisation! uhh... okay, that was a lie but it is a relativley simple language.

#13 I like RISC assembly better. I don't know PPC, however I do know R2000 (or did at one point) as well as 68k. 68k is CISC but I didn't find it as offensive as x86.

#14 When I worked for Motorola Austin we talked about stuff. I told them they should use MCM with BGA package for CPU's. They said "no one will ever want to do that??"...

#15 Have any of you seen the Macworlds's benchmark of the new intel iMac ? It sucks.

#16 I saw the benchmark, and "it sucks" doesn't seem to apply. They tested applications, which tests a whole lot more than just the CPU (eg: disk, video, etc are also tested). The fact that the CPU by itself made up to a 30% gain sounds like a good boost.

#17 It is my personal belief that 95% of benchmarking is crap. I really think that it is an artificial measurement system, and that important things are not taken into account. For example, FutureMark 06 came out. The only thing that tells me, really, is that a given computer will run FutureMark with X performance. An application will only perform as well as it is programmed. The important things, to me, aren't usually measured in benchmarks.

I would buy a Mac (though I would probably want an Intel PowerMac, whatever they will call those) not because of performance reasons, but because of OS X. A familiar, powerful shell, a stable *NIX base, and a good UI on top of it are worth way more to me than a few percentage points on an arbitrary measurement system.

#18 3DMark not only tests its own performance on a PC, but the performance of DirectX and OpenGL running on the system. It means nothing if you don't have an example PC to compare to, but if you're comparing upgrades on your current one, it will provide a nice figure of the performance to compare.

#19 Benchmarks are statistics. You must interpret them in order to have a meanigful result. If you are comparing an iMac 1.0Ghz with an 1.25Ghz model, benchmarks will be very precise on comparing the performance of both machines.

The problem is when you start comparing PPCs and x86s. In the benchmarks Jobs provided he showed that the new intels were 4x faster when processing floating points. But what kind of FP comparison was made ? A 4x4 matrix multiplication done with altivec will be much faster than one done with a dual core intel, basically because the G4 has 16 FPUs !! But this performance isn't achieved just by recompiling the code on the best availble compilers. An altivec matrix multiplication must be hand-coded specially for it. That's where Pat's personal belief that 95% of benchmarking is crap becomes true.

But Macworld's benchmarks where done at a higher level. They benchmarked specific actions that a user would usually do, importing CDs, applying filters to iMovie, etc. And the results weren't good at all.

The new iMac isn't a brand new machine with brand new functionality. It's the same machine with a different processor. Plus you'll have to run Office and Photoshop in rosetta.

I'm lucky enough not to have to buy a new mac now. By the end of 2006 there will certainly be an app like XpostFacto there will allow you to install MacOs on a Dell notebook.

Jobs did the right move on going intel for portables. But that was only like getting an extended play bonus. He hasn't completed the phase and now the enemies are much harder.

#20 After getting a lot of negative comments about the confusing methods they used to calculate their real-world comparison chart, Macworld has changed the chart to clarify things.

http://www.macworld.com/2006/01/features/imaclabtest1/index.php

The new iMac offers a nice, 10-20%, speed bump over the previous model. It's not earth-shattering, but certainly good enough to show that the move was a good choice by Apple.

Anyone who depends (for their livelihood) on an application that has not been coded for the new architecture should probably wait. Most people will buy the new iMac and find applications such as Photoshop to be good enough for general-purpose use.

Oh... And good luck with that whole Dell thing, Bruno!