Slashdot Mirror
Intel's Itanium Will Get x86 Emulation
pissoncutler writes "Intel has announced that they will be releasing a software emulation product to allow 32-bit x86 apps to run on Itanium Processors.
According to these stories (story 1, story 2), the emulator is capable of the x86 performance of a 1.5Ghz Xeon (when run on a similar speed Itanium.) Who said that no one cared about x86 anymore?"
I haven't seen enough info on the new IBM PowerPC 970 CPU expected shorty.
watch who you're calling shorty, farm boy.
The wise follow a damned path, for to know is to be forsaken.
http://www.macslash.org/comments.pl?sid=03/04/23/1 82250&cid=3
First off, the disclaimer: this is my pet theory, i.e., a total, wild, pulled-of-out-my-*ss speculation, okay. I have no inside info or contact with people who might know this, but here is my speculation of why this AMD thing keeps coming up despite the fact that the use of the IBM 970 is almost a certainty.
Put this AMD thing into a bigger context of recent events.
We've heard that 10.3 will include a more integrated Classic environment where Classic Mac OS apps will be given many of the benefits of Aqua.
Apple quietly releases their implementation of XWindows system, X11. Despite the fact that this news set Slashdot buzzing for days on end and probably should have had some mention from Steve Jobs in the keynote he'd given a few days prior, it was released very quietly. Interesting.
Next, the somewhat unexpected news that Microsoft was buying Virtual PC. What on earth could Microsoft want with VPC? We can speculate that they want greater control over emulation of Windows on the Mac, but that sounds weak. They still control the operating system that gets installed on VPC so from that perspective they've gained nothing by buying out VPC.
And then these weird, peristent, inexplicable rumors that Apple is in talks with AMD about something or other. Who knows what. It's very doubtful that it's about a chip that would replace the PPC since we've read many, many well-informed examinations of such a move and the technical hurdles would likely ruin Apple.
So what could all this possibly point to? Apple has given us a system that can basically run software from three different operating systems: the classic Mac OS, Mas OS X (the Next OS), and Unix. They recently brought the Unix world closer with the release of X11. Wouldn't it be amazing if hardware in the near-future included an "add-on" chip (something like Altivec that works in conjuction with the PPC processor) that emulated the x86 hardware? Maybe it would give Mac users the ability to run Windows and PC software, not via software emulation, but with hardware assistance. Imagine the interest Apple could draw if they presented the world with a machine that runs the Classic, OS X, Unix and Windows applications... all in one environment and almost seamlessly.
Now does Microsoft buying VPC make sense? Maybe? Maybe not. Maybe MS Mac Business Unit caught wind of this and wants to one-up Apple somehow. Any thoughts?
AMD would be a likely partner is such a move since one could imagine the problems with Intel assisting Apple with this. If it was popular, Intel would be killing their own business. AMD, on the other hand, wouldn't, if I understand the situation correctly.
Anyway... like I said... wild speculation, but that's what all this says to me.
Comment removed based on user account deletion
What's more, the PPC 970 is not shrouded in secret, (at least from an apple hardware point of view) If you think the 970 is shrouded in secret and is vaporware, I wonder what you think of the Moto G5.
Quid festinatio swallonis est aetherfuga inonusti?
Africus aut Europaeus?
Remember, Motorola didn't even release the PowerPC 7400 (G4), much less any information about it, until apple introduced the Power Mac G4. Does that mean it didn't exist?
Apple probably can't play those same kind of "keep it secret until we announce our product" games with IBM, but keep in mind the only thing IBM has really done was introduce the concept for the processor at MPF. Judging from how Apple has rolled out new processors in the past, it wouldn't be surprising to find if further information about the 970 is being withheld at Apple's request (Apple being a potentially huge customer for this chip).
Also, remember that before CeBIT, IBM posted press releases on its German site talking about 970, the fact that it featured AltiVec, and how IBM was going to be demoing several 970-based blade servers at CeBIT. The press release suddenly got pulled and there was no further information about the 970 from IBM.
One way this could be interpreted is that the 970 is vaporware.
The more likely scenario, however, based on how apple has done releases of new processors in the past, including several iterations of the 7400 family, is that more information is being withheld until Apple releases a system featuring this processor. Then the floodgates will open.
The only reason we may know anything of it at all is that IBM felt it fundamentally important enough to present at MPF - we haven't heard a peep since.
So Apple says this is the year of the laptop, right? If I'm not mistaken, AMD chips run hotter than just about anything out there. So who wants a laptop with 15 minutes of laptop life and the capability of burning your wang to a small, blackened stump of carbon (or for the ladies a sizzling fajita)?
Besides, are they or any of the Mac software vendors going to support two versions of their Mac products? No.
Yup, that's right. Keep perpetuating the myth. Have you looked at the Apple Store lately? iBooks are <$1K ... compare to a Dell or similar notebook and you'll find that Apple matches or beats their pricing.
... as are the Dell, Compaq, HP, IBM and Gateway machines.
True - the desktops are still somewhat pricey. $1000 more? No. Not if we're talking iMacs, and if you're comparing them to a machine purchased from a major manufacturer like Dell or Compaq - If you're talking beige-boxes, well then yes. Apple computers are $1000 more than a beige box
But keep in mind, Apple is really focusing on the portable market segment this year, so that's where most of the value is going to be.
Does nobody remember that both AMD and Apple sit on the consortium for Hypertransport? If you look at Apples current lagging hardware specs you'll see a need for two things. A faster Bus and a faster CPU.
AMD == Hypertransport && IBM == P970
Actually I believe that the complete opposite is true. Apples market share could go up 10x overnight if they released Mac OSX for x86. Hardware is a tough place to make money, the hardware COSTS money. Lost of it, profit margins are slim.
And even if Apple made the OS free-standing for any x86 machine, that would not stop Apple from being able to build ultra-stylish, high-priced, boutique x86 machines and put their own OS on it now would it? They'd probably sell exactly the same # of machines that way as the die hard Apple aficinados would probably still prefer to buy apple branded machines, AND it would grow their OS market share tremendously, maybe even driving their branded machine sales higher in the future.
Frankly I think not porting OSX to x86 was a huge blunder that will hold the company back another 5 years till the next actual achetecture change is forced upon them.
Contrary to popular belief, coding is not all free blow-jobs and beer. Those things cost MONEY!
Apple has heavily optimized OS X and the so-called iLife apps (iTunes, iPhoto, etc.) for AltiVec, the special vector instruction set that the G4 has. That's why OS X runs much more nicely on G4's (which have AltiVec) than on G3's (which don't). The reason all the buzz started about Apple migrating new Macs to the 64-bit IBM chips in the first place was that IBM introduced AltiVec workalike instructions for their new chips, so Apple could move up without sacrificing the AltiVec optimizations. Moving to x86-ish hardware would mean that they'd lose all the AltiVec optimizations they've made, so it seems unlikely to me.
Or, instead a bunch of wild speculation, why not realize that Apple and AMD are both a part of the HyperTransport consortium and are (presumably) both very interested in 64-bit computing on the desktop, and that:
1. One of HyperTransport's most commonly supported speeds is 6.4GB/sec;
2. Apple is desperately in need of a revamp of the entire desktop architecture, especially memory and system bus (aside from processor itself);
3. The IBM PowerPC 970 cooincidentally supports a system bus speed of 6.4GB/sec.
Doesn't the HyperTransport relationship seem a bit more logical than all this off-the-wall stuff about Marklar, Apple switching/adding processors, etc.?
If Macs could suddenly run Windows applications (without something like VPC), why would anyone write anything except Windows apps? The big companies that now target both platforms could just drop their Mac software and tell Mac users to buy the Windows version. Companies that now specialize in the Mac market could start making Windows apps and sell to both platforms. Apple would totally lose control of the integration that has made the Mac experience what it is today. I just can't see any other reasonable result of what the poster suggests.
Hardware is a tough place for Windows PC makers to make money. Apple has been doing pretty well there. Dell and Gateway have problems and losses because they're in competition with each other and with your cousin who makes PCs in his garage. Your suggestion that Apple would make more money by competing with Dell and your cousin is strange.
Apple need another supplier so they limit their risk. They maybe getting AMD to fab a PowerPC type chip.
Alternatively....
Maybe they are just going to use AMD64 chips to build 8 and 4 way XServes?
NeXT used to have fat binaries compatibility across NeXT Black hardware, Intel, Sun, HP and Alpha.
Anthony
I have a better reason: AMD can fab those CPUS easily and IBM has better things to do than fab chips for Macs. Apple needs to make sure it doesn't have to stop the assembly line for IBM to fab more CPUs. My guess is that Apple will have AMD produce IBM 970 chips alongside IBM. IBM probably doesn't want to be the first in line for Macintosh CPUs, there's not enough money in it for a multi-faceted operation like IBM. AMD can produce ample chips and they might be able to make a profit doing it.
This has nothing to do with Macs and x86, AMD produces a LOT more than just athlon chips, they'll be pumping out AMD-970s with their extra capacity.
"Sometimes, I think Trent just needs a cup of hot chocolate and a blankie." -Tori Amos on Nine Inch Nails
As strange as this may seem at first, it makes sense: Microsoft is now in the process of stopping support for Windows 95 and Windows NT 4.0. But some customers still have to run applications that requires these operating systems, and VPC will allow them to do just that:
Quote from an article in Server Watch:
The PowerPC isn't a proprietary processor. If you'd like to design a motherboard that uses it, go ahead. No one's stopping you (unless it infringes on an Apple design, that is). The hard part would be selling it...
Apple will not destroy its PPC customer- and developer-bases by tossing them aside after all the time, money, and effort expended on Mac OS X. Apple will adopt the PPC 970, take Motorola out of the CPU development loop, and provide Mac OS X for a tightly proscribed x86 configuration (including its own branded boxes--almost everything but the CPU in a Mac is now commodity parts, so that perceived barrier is long gone).
Steve Jobs is a patient man when it comes to the world-at-large. He knows that Apple probably won't ever replace Microsoft as the dominant player in the x86 market, but he also knows that this is probably the perfect time to give them some competition. Microsoft faces a number of challenges to its dominance: its attitude toward DRM, its "trusted computing" initiative, the quiet debacle it's weathering vis-a-vis virtual weekly security updates to XP and other critical software, the growing popularity of open source software, its enterprise licensing scheme, and the increased scrutiny it's under after losing the anti-trust case (like IBM before it, the loss itself will prove more damaging than the punishment).
Apple will continue to produce Mac OS X for PPC. The x86 version would be--in the beginning--a loss-leader. It would get noses into the tent from every market segment. That interest would fuel developer interest (notice how quickly the "there's no software for the Mac" discussion abated in the flood of Open Source offerings it now enjoys).
Once that interest is cultivated, Apple has a whole slew of products/ideas "on the shelf" that would benefit from this renewed interest. There's an advantage to being ahead of your time if you survive long enough.
I have a better reason: AMD can fab those CPUS easily and IBM has better things to do than fab chips for Macs. Apple needs to make sure it doesn't have to stop the assembly line for IBM to fab more CPUs. My guess is that Apple will have AMD produce IBM 970 chips alongside IBM. IBM probably doesn't want to be the first in line for Macintosh CPUs, there's not enough money in it for a multi-faceted operation like IBM. AMD can produce ample chips and they might be able to make a profit doing it
I think you have this backwards. AMD just recently signed up to use IBM's new manufacturing plant to increase production yield on chips and allow for higher process manufacturing (.09 micron.) IBM wouldn't be disrupting anything to "just" manufacture chips for Apple. Since AMD will be booming in the embedded business when/if HyperTransport takes off, they'll need the extra manufacturing space to produce their chipsets.
You're also overlooking a very obvious clue to the PowerPC 970 being the chip of choice for Apple: the fact that IBM has included an AltiVec engine (and by that name, too.) IBM has stated before and stated again that they will not be using AltiVec, that it's simply there for 2nd and 3rd party vendors to take advantage of.
Can you name one practical vendor that utilizes AltiVec other than Apple? I highly doubt IBM is catering to Amiga.
The whole thing about Apple being in talks with AMD is more plausible if it's put in terms of HyperTransport chipsets and software compatability, and not switching their entire platform over to AMD64. As noted before, IBM and Apple are both on the HyperTransport consortium, it's only reasonable that they need to talk to each other now and again regarding HyperTransport issues. If you see on The Register or some other place about Apple being a purchaser for chips from AMD, please keep in mind that it's most likely HyperTransport chipsets and not Opterons.
Okay, regardless of whether or not Apple ever designs X86 computers, if they did it wouldn't lower their profits. Why? They can still charge more their computers if the OS only runs on their version of x86 hardware! If it runs faster, the Mac faithful will be pleased. Sure someone is likely to hack it to work on a white box PC, but as far as average end users are concerned, it is not a big issue and any piracy issues would be easily offset by the number of new people buying slightly cheaper Windows compatible Macs. (Heck, I might even consider it.)
I also suspect that OSX (if written properly for a small set of sound/video cards) would be faster than Windows on the same machine. Even if it isn't, people crave the Mac experience. Mac users have never minded paying more. They don't even care that Macs are the slowest on the block right now. It's about the user experience folks. Plain and simple.
And I thought it was just going to be a space heater.
Social Contract? I don't remember signing any Social Contract!
Ultimately, all an emulator does is convert instructions from one architecture to another. It's almost always more efficient to translate instructions in blocks
To come up with a really primitive, simple example, imagine a simple instruction set with a load, add, and branch if zero-set.
Code might look like this:
lda avar
add bvar
bre label
Now imagine we were translating to an instruction set that had mostly the same instructions, but needed a compare instruction to set our conditional flag
Instruction-by-instruction conversion might turn out like this:
lda avar tstz
add bvar
tstz
bre label
Now if the conversion was done on the entire block, we might end up with this:
lda avar
add bvar
tstz
bre label
Granted, this is a pretty simple example, but I hope it makes my point. Block conversions allow a great deal more optimization than instruction conversions.
This optimization might sound like a lot of work for the host processor, but if the block in question is a tight loop you more than make that up.
This actually reminds me of when Apple's emulation strategy back when they migrated from the old 680x0 series to PowerPC. It was well orchestrated and was actually something of a triumph for them. I hope that bodes well for Intel's attempt.
For Intel to have a long term future without the embarassment of junking the whole architecture, they need Itanium x to run IA32 credibly. Advances in x86 performance keep coming at such increasing development costs that I think they would have to be able to migrate the market to IA64 within 5-10 years from now.
I would like for both the IA64 and the Hammer architectures to flourish, but Intel's taken an extremely bold step with EPIC, and I don't want to see them get punished in the market for that alone. I like the spirit of aiming higher.
-- John Truong
Actually, when we created Merced (1st Itanic) it was designed to be able to be FULLY backwards compatible (i.e. boot MS-Dos 1.0). 25%-33% of the chip was actually a HARDWARE ia32 to ia64 translation engine.
You could put the chip is EPIC (ia64) mode and everything would run though the normal pipeline or ia32 mode and things 1st ran through the ia32 translator then most of the normal pipline. Yeah, you took a performance hit in ia32 mode, but it was the price you paid for "100%" backwards compatibility.
So, I am not sure why the change to a software emulator, unless:
1) they ditched the hardware emulator to get back some real estate of the die, or
2) they didn't like the switching the chip between ia32 & ia64 bit modes.
Also, you can tell I've been out of the Itanic design loop for 5 years now. So, some information is out-of-date or lost in the fog of memory. And, I'd like to say that Merced was such a horribly managed project I left engineering.
I think that's a different situation. For starters, Itanium already does IA32 in hardware (it's just really crappy apparently).
DEC wasn't in the x86 market to start with so FX!32 extended their market by making NT/Alpha more attractive. With the 21164, Alpha introduced data handling functionality in hardware that was intended to accelerate x86 emulation. It was probably too late by then.
There must still be major management/direction problems with the Itanium project for them to resort to this kind of hack. It's embarassing that they can outperform their hardware implementation in software.
The only software emulation I can think of that was successful was Apple's 68k emulation for PPC, but their approach was brilliant and well thought out IMO (smooth transition, fat binaries including code for both chips). At the time, PPC was compelling. I don't think Itanium performance is as compelling even though Itanium 2 is pretty decent from what I've seen. I think for a straight 64 bit Linux system, Itanium 2 is a much better chip.
I suspect Intel and friends (oops almost typed fiends!) will be back with improved hardware support for IA32 because people won't be satisfied with the emulation performance. AMD has to feel pretty good about having Intel/HP in this position.
-Kevin