Home Page Forums Archives MacMAME OT - Intel inside Apple?

On the PSX hardware games in MAME (aka "the ZiNc games"), AMD chips do in fact run in line with their true clock speed and not their "xxxx+" marketing rating. Interpretive emulation is a marketeer's nightmare.

The lack of 64 bit support doesn't surprise me, since from what I understand (which is very little), there's really not a whole lot that OS X is doing with 64 bit now, anyway. I think there was a lot more hype to it than Apple is actually able to deliver on at the present time. I think it's something they still want to do, but it's not going to happen overnight.

But...

I suspect that Apple knows something about the Intel chips they're actually going to be putting in their Macs that they're not able to talk about yet. If they've been planning this switch for awhile, I think they could get Intel based systems to market much sooner than a year from now. My guess is they're planning to stick an as-yet-unannounced Intel CPU into their Macs. Those may not be the first chips Apple uses, but I suspect there's more there going on than just "let's use Pentiums".

Jobs seemed to indicate as much, since he kept referring to "future products" that Intel's chip roadmap fit, that IBM's didn't, and not so much about what the chips are doing right now. I think that Apple's "projects" will extend beyond computers as well, to include more consumer products. So there's more to this than meets the eye (insert Transformers sound effect here).

Also, I think RB is absolutely correct - this is a lot more about laptops than desktops. The way Jobs presented it, it was obvious that it was much more than just part of it. It's a big part of it. If I had to guess, I would think their first Intel computer may be the next gen PowerBook.

I'll be interested to read Ars Techina's take on all of this. They ususally have very good technial analysises... uh, analysese... well, whatever.

We aren't using G5s in our labs. We're using G4s (dual 450s). We want to switch to G5s, but I'm questioning that decision based on whether we can still keep running them with the latest software for five years or not. I'm concerned that Apple may not be supporting the PowerPC platform that long, and we may have a couple of years where we have to run out-of-date software. Not good at a film school. Not good at all.

Five years for us is mandatory, too. We just don't get that level of financing all that often.

So I have to mull over the possibility of us slogging through another year on our G4s (which run everything we need, albeit slowly), and buying Aptels next year in order to gain better future compatibility.

I'm pretty sure he said "all products". He didn't specifically mention the XServe, but he didn't exclude it.

But you bring up an interesting point I didn't consider. If Apple is going to continue to have a server business, they're going to have to maintain PowerPC support for quite some time, to keep the boxes that are out there now running. Servers get changed out far less frequently than desktops do, but it's essential to keep the software updated. Even though it's a small part of their business, that's one area that's harder for them to abandon than desktops, since it's not as easy to swap out a server that supports hundreds of users, as it is to swap out a desktop Mac. So by extension, they may keep support for other PowerPC Macs around longer, as well.

At least I hope so. We've got seven XServes in our area that we rely on every day.

This is an over-simplification of the modern x86 chips. Read about register renaming:

http://www.cs.swan.ac.uk/~csneal/HPM/reorder.html

Also the Pentium does have vector instructions (SSE and SSE2), this is the first couple of links I found:

http://www.tommesani.com/SSE.html
http://www.answers.com/topic/netburst

K

Haha, no I wasn't being sarcastic. I realize you can't really compare Mhz between chip families which is why I was thrown back when reading about MAME primarily wanting more Mhz. I wasn't sure if you were talking in a general sense, or just between G5 and P4 or what. Thanks for the carification.

I hate to break this to everyone, but the moment that Apple said "we won't stand in the way of using Windows on our new Macs" (which, I assumed would be the case by Apple killing OpenFirmware, as that's the only reason you would even want to), I realized how unnecessary MacMAME will be.

I had hoped from the minute that the Xbox 360 was launched that J Allard at MS was working on a special NT kernel for Macs that you could just reboot and run Xbox 360 on Mac. He's a big Mac fan, there's certainly a market in (what would have been) emerging G5 Macs, and the advanced state of Xbox 360 would have allowed for things like high-resolution displays (like, a computer display above 640x480).

I have always maintained that if OS X (advanced multitasking, sharing of resources, etc) goes against everything that a good gaming OS needs, why not just reboot out? It only take a few seconds, and the benefits are well worth it. Windows allows the game to "take control", OS X so far hasn't, and don't get me started about the wonders of Direct3D on my future PowerBook.

Now, I don't think this is going to hurt Apple. Apple will never bundle Windows or make it easy at all to install Windows (enjoy repartitioning a hard drive with both NTFS and HFS+, for example). But, let's be blunt here; the userbase for MacMAME has always been the same enthusiast that would be able to install that with even a basic set of partitioning tools (I can see iPartition selling big here if they step up to the plate and add NTFS).

I'll be happy to compile MacMAME in Xcode for years to come, but when it comes to playing that high-end Model 2 (and soon Model 3) game that gives even today's games a run for their money.... it's off to Windows I will be happily going.

Here's an article on the switch at Ars Technica . It's not one of their usual ?�ber-technical pieces, but it's a good read.

Well, the article misses a few things, for example, not stacking up SSE3 against AltiVec. Does AltiVec do some things a lot better? Yes, but nothing necessary and nothing that would warranty going back...

It also fails to mention a critical advantage for going with Intel: Chipsets (and options). Apple needs x86 Macs now. Where are the AMD chipsets? In the dust. If Apple went with AMD, they would have had to go elsewhere (nVidia, VIA) to get a high-end chipset that did everything they need today. Buying them isn't an issue, coding drivers, battle-testing, and supporting them is.

Going with Intel now doesn't preclude Apple from going to AMD, and that's the key. If Intel gives Apple not enough of a pricebreak, they can take that time later after the migration is done to start a price war (imagine, an actual pricewar over Apple's business, and no, Freescale vs IBM doesn't count).

And about x86-64, please, Apple knows it's essential (because they know Power Mac users love their 8 GB of RAM), but it isn't necessary today. I wouldn't be surprised at all if Leopard is x86-64 clean (that's right, clean) and would be Apple's first 64-bit OS. But, the amount of work needed now (porting apps) can be simplified by telling developers the truth; don't worry about it right now.

It's even simpler than that. Remember, laptops (and other small devices, a la the Mini and maybe the long-rumored Mac Tablet) are the key to this deal. AMD has no story there, while Intel's Pentium-M is so well regarded in the PC hardware community that companies are building full-blown desktop overclocker motherboards for it.

Actually, that was my point, Centrino can fit into a 12-inch PowerBook (13" assuming HD compliance) tomorrow. Down the road Apple can assess chipset divergance from Intel.

The Pentium-M is the key to Intel's success... it is being diverged into just about everything Apple will sell. I don't doubt when the transition is done that we'll be looking at every single Mac having an M-derived processor (with the potiential exception for eMac, Xstation, etc).

On a bit of a sidenote, If all goes well though, and IBM pulls through with their long-term Power strategy (unifying Power and PowerPC processors) we could see super-high-end (Xstation) computing going back to the Power family... Apple is certainly not telling people to stop building PowerPC binaries once things are done in the transition...