Slashdot Mirror
Motorola G5 - 2Ghz 64bit
Nerdkiller writes "
An article appeared on ZDnet with some information on the G5 chip expected in 2 years. It will be competing with the Intel Merced which is expected out around the same time. A full 64 bit 2 Ghz processor. The Intel Merced will be able to support 64 bit processing, however it must be run under emulation for 32bit code. The G5 requires no change in current code with exception to some low level OS stuff.
"
QNX runs on PPC. It probably doesn't support pmacs, though, but i'm guessing you'll probably be able to find embedded altivec g4 based SBCs soon.
The "extensible architecture" is part of the previously announced "Booke E" project:
Check it out.
If it makes sense, it brings us to another question, why don't Motorola have an IA-32 unit on their PowerPC chips?
Interestinglg, IBM at one point was going to make a PPC chip with an x86 chip (the 615 and 620), but their plans fell through because Microsoft was having trouble with NT-PPC, and Apple didn't want the chip....Oh well!
>> The powerpc chip is meant for macintosh products and thats all the is to it.
Not so. IBM has been making heavy-duty UNIX workstations based on PowerPC chips (such as the RS-6000, if memory serves me correctly) for quite a while now, and Motorola has also been selling slightly-modified versions of the chips for use in embedded systems for a long time.
Apple may be the most noted user of PowerPC technology, but it's far from being the ONLY user of it.
But unlike MMX, SSE, or 3DNow, Altivec is an entirely separate vector unit, with its own vector permute unit to chop shop the data, so it is usable by the SIMD instructions. So unlike AMD's or Pentiums a PPC 7400 (aka G4) can process FP, Int and Altivec at the same time, no mode switches or register swaps.
they won't have nearly the same amount of trouble. AIM made 64-bit G3s for R&D purposes, but never marketed them and supposedly also 64-bit G4s. IBM already sells 64-bit PowerPCs, such as the Power3 - AIM is far better prepared then Intel was, so G5 will be pretty painless.
References, please? I've seen nothing to indicate that any PowerPC chip that actually made it out the door had x86 instruction support; all I've heard were claims that a 615 chip was being developed that would run x86 code, and, if I remember correctly, claims that Exponential's 704 would have x86 support.
The PowerPC architecture is derived from IBM's POWER architecture (yeah, the same POWER architecture to which you referred with "IBM Power"), which first showed up in the RS/6000's, with assorted changes, e.g. single-precision floating-point and multiply and divide instructions that don't go through an MQ register. Please explain how a POWER derivative is "a RISCified mixture of m68k and x86, just a little cleaned up".
Shit, guess I'll have to return my gross pay to my employer for the year if anything I did on my Mac wasn't work. Won't the company be surprised they haven't done anything productive in the nearly 18 years they've had Macs onsite. And our readers were just buying into the myth.
No. Check out the PowerPC Programming Environments Manuals section of this page, and the (PDF) documents it links to. The instructions are 32 bits; the addresses and data they manipulate can be 64-bit in 64-bit PowerPC processors.
And as others stated in replies to your post, that's not the case; the instructions are 32-bit in the 64-bit PowerPC architecture.
You can address more memory without 64-bit instructions; you might just have to do more work to synthesize addresses, and if the address is just a pointer you've been handed, or a pointer in a data structure, you don't have to synthesize the address.
A few inaccuracies with the post I am replying to (Most of these are just nit-picks from someone familiar with the instruction set):
:-)
* It is SIMD, in that each altivec operation does the same thing to each part of the altivec register.
* Altivec has 32 Altivec (128 bit) registers, not 64, in addition to the 64-bit floating point and 32-bit integer registers.
* Altivec registers can be treated as:
- 16 8-bit, 8 16-bit or 4-32-bit integers
- 4 32-bit floats
- 128 bits for bitwise operations
There is no support for 64-bit integers, or for 16-bit or 64-bit floats - only 32-bit floats are supported.
* There is no explicit support for fixed point operations, although there is provision for converting between floating and fixed point.
* Altivec instructions take 1,2 _or_3_ registers as source, and one as destination.
It is also rather fun to program (even if only using an emulator at the moment)
- Roy Ward.
I'd include in the PowerPC family any chip that implements the PowerPC instruction set architecture, and this press release says:
(although it's not entirely clear how the instruction set "provides the ability for thousands of POWER3 microprocessors to be combined into a single, unified computing unit"). The RIOS chipset in the first RS/6000's implemented the POWER architecture, which had a few things PowerPC doesn't and didn't have somethings PowerPC does (but there is a common subset of the instruction set), as did the RSC (which I'm told stood for "RIOS Single-Chip", and which I've heard was modified to make the 601), and the POWER2 chipset implemented the POWER2 architecture (added a few things to POWER, but also lacked some stuff PowerPC had); I guess POWER3 finally picks up all of PowerPC.
However, it, like the RS64 in some other RS/6000 models, and the models used in the AS/400's (one of which is the RS64 III, as per this paper, presumably with "tags-active" mode turned on in the AS/400's and off in the RS/6000's), are proprietary to IBM, unlike the 4xx, 6xx, and 7xx chips, which IBM Microelectronics sells.
Pentium chips still run 16bit code. Lets try to enter the realm of fully 32bit architecture (NT, MacOS, Linux, ect.) and leave 16bit behind. Having a fully 64bit system will be nice and all but it's not quite needed yet for the home or office, unless of course your business includes high performance servers or giant 3D workstations. As seen with many flavours of Unix-linux included-a 32bit architecture works fine for most applications. The rise in 32bit Xeon servers and Pentium workstations shows that 32bit isn't obsolete yet and can do the same job as it's 64bit counterparts sometimes for much less money. And just two things for you to think about, keeping CISC on a RISC chip and comparing it to a 32 and 64bit core complaining about performance issues is comparing Apple and oranges. Having the ability to process 32bit code on a 64bit chip if anything increases the speed because you can have two 32bit instructions run every cycle rather than a single 64bit instruction. This means the 32bit code is run at twice the speed as if it were run on a 32bit chip. Second of all Motorola already makes 64bit chips, these funny things called PowerPC chips. IBM and Apple use them, Apple runs 32bit code on the 64bit chip and now the 750 which is 128bit much faster than it ran on older chips.
I'm a loner Dottie, a Rebel.
This isn't really useful to me unless there's other PPC hardware out there besides Apple. I feel that Jobs isn't that far removed from Gates when it comes to business practices, and don't want to tie myself to a hardware platform that isn't even really open... And really, what fun is having a computer if you can't build it yourself? ;)
>In contrast, the 64-bit version of the G5 (which will also be available in a 32-bit version) will be able to run 32-bit applications "in full native mode," Swearingen said.
If the 64bit version runs 32bit applications just as good as (or nearly as good as) an 32bit version and it only takes a few modifications to make the 32bit application run, then why make the 32bit version?
Seems like a waste of time to me?!
That time could (maybe?) be used for getting the 64bit version out a bit earlier...
I agree, but since this is Intel's butter and bread, they are aggressively supporting any company that can come up with creative ways to use all that available horsepower; therefore creating demand. Microsoft, for one, will probably come up with Office 2002 with a 3D-animated Office Assistant paperclip with artificial intelligence and speech recognition..well, you get the picture.
Seriously though, so far game companies have been at the forefront for pushing the performance envelope. During the next two years, broadband, always-on Internet connections will drastically change the computing needs. Increased need for security will lead to use of stronger encryption, people would like to watch streaming videos with larger screen size and better frame rates, and so on.
And don't forget, software people will always manage to find a way of bloating their software with new features that will suck up all available CPU power. Even good old Linux is not the same thing it used to be five years ago. I mean, it was possible to do a lot of work on a 486DX-2/66 with 8MB RAM and trusty twm. Now everyone and his mother seem to run Gnome or KDE for just having a couple of open xterms.
Zigbee Central: A Zigbee weblog
Anyone bashing away on an computer trying to make music would love/needs this kind of processing power. Applying real time effects to audio production and being able to swap them in and out etc causes HUGE processing overheads.
Er, sorry, stray "e" there. Why doesn't Slashdot allow authors to edit their own comments?
Or, here to be more specific.
Thanks. I was just being lazy... ;)
--
Don't lead me into temptation... I can find it myself.
Or 1 year for a 64 bit Merced
Or buy a 64 bit alpha today
Decisions Decisions...
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
> and that the G4 makes this possible for
> wont of a RTOS.
Isn't Apple's Darwin a Real Time OS? I thought I read that in an article about QuickTime Streaming Server.
So you can play out Star Trek in your house. Dictating your report for the boss whgile listening to mp3s and surfing all at the same time on the same computer. More realistic games are another option, maybe first person shooters but also ultra real racing and flying games, detective games with super realistic environments.
I'm a loner Dottie, a Rebel.
Not if, say, a 32-bit arithmetic instruction uses an entire ALU, as I think is the case on most 64-bit machines; I don't know if there are any that'd split a 64-bit ALU up into two 32-bit ALUs and run two 32-bit instructions through them (especially given that registers generally aren't split that way, either).
The SIMD "multimedia" instructions that most general-purpose instruction set architectures have picked up may process multiple less-than-full-word-size units in one ALU and in one instruction, but that's one instruction, not two.
What chips that implement the 64-bit version of the PowerPC instruction set does Apple use? (Chips that implement the 32-bit version of the instruction set, but that have a 64-bit bus interface, don't count as "64-bit chips" here.)
No, it's 32-bit - it implements the 32-bit version of the PowerPC, without the AltiVec instructions, and according to this page at Motorola's Web site it has only a 64-bit bus interface for data.
Perhaps you're thinking of the 7400, which has the AltiVec instructions, but it also has only a 64-bit data bus, and only implements the 32-bit version of the core PowerPC architecture (note that the page for the 7400 links to the 32-bit version of the "Programming Environments Manual"), even if it also implements the AltiVec instructions that work on 128-bit registers.
I talked 'bout REAL COMPUTING. The Mac is occasionally still used for some simple Graphics tasks or Media Editing - the latter however only works reasonably well with some very pricey extra Hardware slapped on top of the System. My Word still stands: No real computing is done on Macs. They are basically lifestyle toys.
Um, I don't want to be the one complaining about new cool (hot?) processors, but to start hyping something we'll see in two years sounds a little bit extreme. I mean, that's almost the way Wintel do business, right?
main(O){10<putchar(4^--O?77-(15&5128 >>4*O):10)&&main(2+O);}
"640K should be enough for anyone" Bill Gates circa 1980.
I don't wish to be an apologist for Bill Gates, but this quote appears to be an urban legend. No one can find a source for the quote.
More info can be found at http://www.urbanlegends.com/celebrities/bill.gates /gates_memory.html
Every two years the speed of software halves.
So all this stuff makes not a damn bit of difference.
Now look - you've gone and made me grumpy.
I stand corrected. Thx.
Now if we could just get some additional, good, free, open-source operating systems ported to run on this chip, it would be even better.
LinuxPPC is there, and good; I've been very impressed by what it can do. But wouldn't we all like to see FreeBSD, or other Linux distributions ported to run (and run well) on PPC chips? And even better, now on the G5. I can just see the performance of a full 64-bit native OS running on that chip... *drool*
But then again, we have plenty of time before they come out to do the work!
---
Tim Wilde
Gimme 42 daemons!
This is all fine except that future 3D card have their own geometry engine, which means the only job left to the CPU is AI and a bunch of small non-intensive tasks. This means there will be hardly any difference between a 400 Mhz Celeron and a Athlon 700 if you fit a card with geometry engine. My opinion is that the video chipmaker (nVidia) are going to get the big bucks Intel and AMD used to have from gamers. Now it is common to pay more for your 3D card than for your CPU, and according to price quotes from nVidia on the new GeForce 256, it is going to be even worse.
Why was this posted under the Apple heading? Slashdot needs a seperate topic for PowerPC related articles.
To increase my rc5 stats. That's probably about it. Pathetic existance, isn't it? -Saxton
_________
My name is Aaron Landry, and I approve this message.
I don't understand what they mean by saying that the new processor will be able to run 32-bit applications 'without emulation'. All processors have a certain amount of legacy stuff they have to support. Journalists would have us believe that while Merced will 'emulate' x86 instructions, the Intel 686 CPUs can run them natively. But the i686 uses a RISC core and translates instructions before executing them.
-- Ed Avis ed@membled.com
So what is REAL computing?
I could've sworn that i've used Mathematica on a Mac.
I could've sworn that i've done alot of CAD work on my Mac.
I could've sworn that i've done circuit layout on a Mac.
I could've sworn that i've done digital simulation on a Mac.
I could've sworn that i've done logic optimization on a Mac.
I could've sworn that i've used PSpice on a Mac.
Did I not do these things? Or do you define REAL computing to be something else, like recompiling the Kernel.
FYI - A *HUGE* chunk of graphics and Media work is done on the Mac. We're talking in the 60% market share figures here. That isn't just occasionally used in media work.
Tom
Actually, the AltiVec instructions do operate on vectors. The cool thing is, it can operate on a 128-bit vector with components ranging from 8-bit to 64-bit. There are references for the instructions in AltiVec at http://www.mot.com/SPS/PowerPC that describe all of the built-in functions available. I imagine there are quite a few more vector instructions they could have added, but adding 162 new instructions is not something to be taken lightly when your instruction word size is fixed (like it is in all 'pure' RISC ISAs).
Can anyone explain how much difference there is between the architecture of the G4 and G5? I was reading just yesterday that the theoretical limit on cycles for the G4 was 4 gigaflops -- will the G5 be just a faster version of the G4, or will there be much change in architecture?
I could've sworn that i've used Mathematica on a Mac.
I could've sworn that i've done alot of CAD work on my Mac.
I could've sworn that i've done circuit layout on a Mac.
I could've sworn that i've done digital simulation on a Mac.
I could've sworn that i've done logic optimization on a Mac.
I could've sworn that i've used PSpice on a Mac.
Did I not do these things? Or do you define REAL computing to be something else, like recompiling the Kernel.
ALL of these are NOT REAL COMPUTING. I was talking about wheather forecasts, hydrodynamic simulation, nuclear plant process control, space flight navigation, bank and insurance accounting, nuclear test simulation, data warehousing, civil flight control and dispatching and the like. No man with half a brain would trust a Mac (*shudder*) for REAL TASKS like those.
A *HUGE* chunk of graphics and Media work is done on the Mac. We're talking in the 60% market share figures here. That isn't just occasionally used in media work.
And where else are Macs used????? It's the Macs *only* niche, and it has been losing share constantly over the last years. But perhaps the Mac can open up a new niche as your little sisters internet terminal (see iMac).
And then you go up exponentially from there if you want to apply those real time effects to video, or render massive 3D spaces.
...
Computers aren't fast enough until there are no wait cursors or progress meters
Agreed. Thankfully, IBM just released the specifications for the PPC motherboards for anyone to use and create freely. (I'd say they open sourc'd it, but I feel that I'd be using the word improperly--plus it's really a noun, not a verb).
ARGH! I tried to find a link to the slashdot article or something on macweek/techweb/etc, but I couldn't find it. Anyone have the info?
Good point. Hopefully the availablility of PPC motherboards in general will improve as a side effect of IBM's announcement at LinuxWorld that they're giving away the design to one of their mobo's. With any luck this should help jumpstart the PPC mobo industry and set the stage for supporting future processors in that family.
Now last I looked, TI's 'C6x architecture was headed that way with speed and scalability, but no one is ever going to write or port a general purpose OS (ie, Linux, Windoze, MacOS, Be) for/to that chip.
Uh, why would you port a general purpose OS to a DSP chip? DSP chips are special purpose devices that fit a lot of niches (well, okay, niche really isn't the word since they outsell CPUs by orders of magnitude) that don't need general purpose processors. Much better to have a controller chip of some kind (say x86 or PPC) interfacing with a DSP farm for processing.
Apple never used hardware emulation in the switch to RISC. It was all software.
Most 64-bit RISC architectures continue to use 32-bit-long instructions, and one (ARM THUMB--interesting!) 32-bit architecture also includes 16-bit-long instructions (compressed instructions) to reduce the sick code size of the RISC applications for embedded purposes. It makes little sense in the RISC world to use longer instructions because this would increase the logic complexity and actually decrease performance. An exception is the long-instruction-word architectures, which change the compiler writers' obligation to generate pipelined/parallelized code into a requirement, hoping for better speed. The Merced follows in this line, and includes an obligation for explicit parallel computation. In the MIPS IV ISA, for example, there are 64 64-bit registers, 32 of which are used for floating point. The address space is 64-bit. In the Alpha and the UltraSPARC, the data size and address space is 64-bit. There's really no reason for MOT to try to escape any of these parameters, because technology isn't there that requests "better" capability.
Did you forget about OpenBSD or NetBSD? They both run on the ppc architecture.
that was about the best comment i've heard all day
Personally I think hardware and software innovations go hand in hand. If you build a processor with that kind of power, it opens up a whole LOAD of new possibilities as far as the things you can actually consider doing with it.
The easiest example is games. Can you imagine a fully raytraced outdoor renderer??? Imagine running around Jurassic park in real time.
Or how about doing it all with a headset and a VR suit on???
There are always bigger and better places to go, and its the hardware rather than the software that limits us. There is only so fast you can optimize an algorithm, but by increasing the speed of the hardware, the sky is the limit!
Dave.
G3 B&W are now upgradeable via a fix from Powerlogix !!!
Ummm... The statement in this article is not correct. Merced has already taped out and if it takes them 2 years to produce Intel has serious problems. By then it will be Madison or McKinley. Also the Merced processor has two different parts to its logic. 64 bit and 32 bit architecture, so the processing between the two is seamless. Also I saw someone refer to Intel as Wintel. I don't like Microsoft either, but when you make processors, you have to make them work with the software everyone is using. Reguardless of what everyone thinks, the people who run this company are not idiots, they support what is used. As far as Linux is concerned, now that it is being used more... Intel is starting to support it. If you didn't hear, Linux ran great on the Merced at an expo a few weeks ago. :)
...because CrackOS is so damned lame, slow, idiotic, and of piss poor design.
Just like your not so well thought out comment.
--
Don't lead me into temptation... I can find it myself.
There are no motherboards for StrongARM, it is an embedded processor.
With Apple's current business ethics (suing all makers of translucent all-in-ones, disabling upgrade potential, discontinuing free/semi-free software upgrades for new purchasers) as well as their dwindling market share and frankly overpriced, underpowered lineup.... they simply WON'T be around come the release of this so-called G5. Hopefully IBM's open-ended logicboard designs will still be around so I can run linux/*bsd on a G5-based system....
That will be out in two years and will require your computer to running off of 220 volt services. spoon
The difference between an ``embedded processor'' and a ``general purpose processor'' is as much marketing as anything else
At one point, the StrongARM was being strongly promoted as a Network Computer (aka ``X-Terminal'') device. Note the announcement of 1997 of the Digital Network Appliance Design.
And note that it is the processor used in the Rebel/Sidewinder that Corel Computers used to hawk.
The point of all of this is that the CPU is clearly not so ``embedded'' that it would be inherently useless in a ``desktop'' role.
It ought to have been possible to build motherboards integrating a CPU, video chipset, and Ethernet that could retail for less than $150, and this could have brought us $300 computers a year or so ago, and provided slick little boxes to velcro to the sides of 17" monitors.
If I could have bought a StrongARM motherboard for $100, I probably would have built a machine by now.
But no motherboard leads to no systems. Note that exactly the same reasoning may be used with MIPS...
If you're not part of the solution, you're part of the precipitate.
...that by the time G5's come out, Apple will have 750Mhz buses. Likely, once G4s hit 1Ghz, the boards will have 500Mhz buses.
Wish I could cite you something, but I can't find it right now. That's what makes the most sense, though.
J.
damned vulpine http://sb.drtwister.com/
For the L2 cache on the 500 MHz models (half-clocking on all models) . It should hellp a lot.
You can't bend the laws of physics, just work around them. I believe the main bus (processor card to memory) is 125MHz. Processor to main short-term storage will be a problem for many years to come, to say the least.
Your point is correct. But the ability to use a computer you already have as a general-purpose DSP is very powerful. It allows you to do it without having to add a card to your box, deal with driver / compatibility issues, etc. It lets software designers skip the step of creating hardware (that most people won't want to buy, anyway). And most importantly, it saves on costs: a software-only DSP could be downloaded free, but creating a hard-wired DSP solution means nobody will ever get the result for free.
You can use a computer you already have as a general-purpose DSP without using a DSP chip as the CPU. However, you're not going to get as good of performance (read: You get what you pay for). And the majority of DSP chips are not compiler friendly (although this is changing in some of the newer generation chips) because DSP chips have structures that are difficult to map into HLL's like C. Putting a general purpose OS on a DSP chip will only end up hurting performance. If you want a quick prototyping base for DSP work, use your computer. If you want peak price-performance, you're going to do it on a DSP chip and with a minimal RTOS.
The G4 is not a SIMD processor. All instructions are fixed length (good thing if you ask me). They added 160-some odd vector instructions, which allow the 128-bit vector registers (all 64 of them) to be treated as either 16 8-bit integers, 8 16-bit integers, 4 32-bit integer, 2 64-bit integer, 8 16-bit floats, 4 32-bit floats, or 2 64 bit floats. There are also quite a few fixed point operations. All instructions take either one or two sources, and one destination, which is the most efficient instruction design since there are so many registers on the chip (you don't have to deal with registers being overwritten by single machine instructions if you don't want to.)
I had heard that the G5 will have multiple AltiVec logic units on it, much the same way as the current generation has multiple Integer logic units. You may want to check www.mcg.mot.com -- they build PPC boards (ATX, VME, cPCI) and have some flavor of Linux supported that has an embedded taste.
They talk about Apple working on 266MHz SDRAM implementations, probably for future G4s and in anticipation of G5s.
Someone else mentioned that Alpha gets it's 200MHz/250MHz bus by multiplexing a 75MHz 256bit wide bus. I can imagine 8 64bit memory buses running at 100MHz, which is can be worked with as a single 800MHz 64bit bus. But that's a darned high wire count. More likely you'd get 8 32bit buses at 100MHz or even at 133MHz.
Excuse me if I don't make much sense, I'm speculating here =)
-AS
-AS
*Pikachu*
A clock pulse can only travel a limited distance given the limited speed of electrons. Am I right in assuming there will be a time when the laws of physics put a halt to ever increasing speeds? In one billionth of a second, light only travels less than a foot. And that's in a straigth line.
Bill Gates circa 1980.
Please...
You do realize that by the time the G5 ships Mac OS X will have been out for at least a year an a half? You might want to go read up on OS X and them come back and let us know what your specific objections are to a Mach 3 based real-time OS with a full BSD 4.4 layer, a graphics engine more powerful than anything ever seen before, most of the APIs of NeXTStep, and ease of use better than the current Mac OS.
Yes, the current Mac OS sucks, but Apple knows this, and it has less than a year (and possibly as little as 3 months) to live.
--
This space unintentionally left unblank.
Apple was testing 233 or 266MHz RAM and has it working (accoring to the article) but has not committed to implementing it commertially yet because it is so hard to find. This is just for the G4's also, don't know what's cooking for G5. http://MOSR.com, you will have to look to find the article (may instead have been on appleinsider, I can't remember.
- Sig
Thou shalt not cast pearls before swine.
I've read that the G4 is a 128 bit processor. I know this sounds too good to be true, but according to the Apple site "The secret of the G4's revolutionary performance is its aptly named Velocity Engine. It's the heart of a supercomputer miniaturized onto a sliver of silicon. The Velocity Engine can process data in 128-bit chunks, instead of the smaller 32-bit or 64-bit chunks used in traditional processors." Is this just marketing hype? If the G4 is a 128 bit processor, why wouldn't the G5 jump up to 256 bits? What gives?
i selected the main link, and couldn't see the
comments
So DEC Alphas and AMDs use a 200MHz FSB, or whatever they call it, and Intel/VIA/etc will be using 133MHz FSB as well.
I'm not sure how the RAM and bus speed correlates, but Apple seems to be actively researching 266MHz RAM, most likely in anticipation of G5s. It has also been documented that PC133 outperforms RAMBUS, so PC266 may be what you're looking for.
Look at http://www.macosrumors.com/8-99.html under the 8/16 update.
-AS
-AS
*Pikachu*
604e should be enough for everybody.
---
Have a Sloppy day!
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
-F.
I'd actually believe that Motorola has no problem selling it's PPC and 68k CPUs in embedded systems, and it was only recently with the advent of $600 systems that the popularity of architectures and CPUs becomes an issue.
I mean Motorola sells a 68k CPU with every Palm and Visor out there. If they have an embedded processor in a cell phone, I'd think they would be selling more cell phones than PCs. Yes, PPC CPUs in computers would be more popular, but it wasn't necessarily the most profitable or intelligent thing to do; it would require that Motorola(or someone else) support AGP chipsets, PCI chipsets, memory chipsets, etc, for a single system to make a profit.
If you haven't noticed, Apple does this all by themselves. If you search IBM's website for PowerPC, you'll also see that they had PowerPC systems for sale since 1996 or something, but at $6,000 costs.
I can't imagine anyone just picking up the PowerPC platform until LinuxPPC stabilizes and matures, because I don't think anyone can compete with Apple on a design standpoint and no one can compete with Intel on a price standpoint.
Apple may be our only hope here =(
-AS
-AS
*Pikachu*
By the time the G5 is out, the MacOS as you know it today will have been dead and buried for at least a year and a half.
so why hasnt anyone posted in here yet? Do you all just have nothing to say or what? This is the first time I've seen 0 comments for an article that is atleast a few hours old. :)
Joseph?
AMD may be gone in two years, given today's news that Intel just cut a sweetheart deal with Gateway. If Intel can reverse the trend among the big players towards using AMD chips, it's Goodnight Irene.
A 750MHz FSB? What are they planning on using to do this miracle? On-chip fiber optic? Current technology tops out at a 200MHz FSB and Motorola is going to try and achieve a 3.5X improvement in the state of the art in two years? I think not. Most likely this will be running at 2GHz on a 266-300MHz FSB - although possibly this 266MHz FSB could be double-pumped.
If you can cite something - anything - to back up this claim I would be extremely interested in reading about it.
I'm typing this from a G3 running MacOS X Server, and it's no slouch. Random sample info that probably means little:
well, in all fairness the G5 would not be the first 64bit PPC. That distinction goes to the PPC620, released in '94. If you expand the PPC family to include the POWER chipsets used in RS/6000s, then you can also include the POWER3 as a 64bit PPC...
since I'm a nice guy I'll even look up a link for ya:
www.chips.ibm.com/news/1994/94101810.html
You have no fucking clue. Legacy, are you kidding? Can't believe you compare x86 to PPC. Do some research on the PPC before you start yelping about its "legacy support".
I know people still will want to run Doom and Hugos Great Adventures on their G5's and Merceds, but I think it is time to cut the fat(32-bit) off and move on to the future. Trying to find a way to use 32-bit and 64-bit instrucitons just wastes time (I guess I just want the processor as soon as possible). I for one would be willing to give up my 32-bit programs for faster 64-bit programs on a 64-bit OS. Hell us linux users who left the clutches of microsoft to embrace the wonders of Linux gave up all our Windows possessions and now run Linux programs. People can do it. It's just like when Apple started converting from CISC to RISC. They kept that ungodly slow CISC emulators on thier RISC chips which dragged the processors speed down. All I am saying is "To Intel and Motorola: JUST GO PURE 64-BIT. Thank you."
In two years, we'll all have huge, digital, plasma televisions on the living room wall. Some of us will also have them in the bathroom.
In two years, we'll have Linux on our credit cards, in our running shoes and trainers, in our hats, heads and toasters.
In two years, we'll be living under the sea and speaking with the dolphins.
In two years, we'll have an outpost on Mars.
In two years, we'll all be two years older.
In two years, the world will collide with an asteroid and only the bacteria will survive. Some Linux boxen will still not need to be rebooted.
http://www.worldnewyork.com
Wordnik, a dictionary project which aims to collect
I was unaware that gcc isn't available for LinuxPPC.
Wow...I'm just chomping away at all the troll-bait that comes my way this morning. Must be a Monday.
pooptruck
No, not really. Despite the recent dabblings with the Alpha, the focus of the FreeBSD group has always been on getting it to work well on Intel hardware. Look to the other BSDs for PowerPC support:
"The invisible and the non-existent look very much alike." -- Delos B. McKown
According to the current information from Intel and companies working with them, Merced is "only" 1 year away, not 2.
Although the x83 family is not a very efficient design because of the enormous amount of resources occupied by backward compatibility, it just isn't bad enough to allow competition to overcome intel's main advantages: compatibility and price/performance ratio (and perhaps FUD-proofness).
Throughout the past decade and a half, many competitors have made products that were superior to intel's in every way but compatibility and price/performance ratio. None of them ever crawled out of their niche market.
Now this Motorola announcement not only promises to beat intel at both counts, but also to do this by an incredibly wide range
When something sounds too good to be true...
Given their history of accurate journalism, Nerdkiller should not repeat news from Ziff-Davis without double-checking first. Merced is NOT expected to come out in two years; it is expected to arrive in Q3 2000. In two years, Intel will have McKinley. If it takes that long for Motorola to beat Merced, it would be a good idea to short Motorola stock.
Maybe the G5 will finally make the MacOS run as fast as linux and some windows machines. Macintoshes have always had great chips, but the software's overhead has been killing the speed and making most people think that Macs are slow and a joke. May the G5 prove the power that Apple can wield.
Dan
Intel expects to release Merced earlier than in 2 years, at least I hope they do, otherwise they will lose all the market in powerful and SMP machines, ATHLON(K7) is so much better than PIII, and Merced seems to be a little hyped. I heard Merced it's booting Linux, WinBlows NT and Monterey already, but what about benchmarks?? Is that thing fast? Well Athlon is here and it's fast, and who needs 64 bit now? This G4 chip seems to be nice too, but running AIX, MACOS doensn't attract me,maybe to run linux but I prefer x86,hell I prefer FreeBSD, and 2 years to release G5, well that's a lot of time...
I wonder if they'll have the same problems intel is having (had?) building it's 64-bit chip. And there's the question of multiple core G5s... What's up with that? BTW: Did I just get first comment? naww...
As Apple could license AGP from Intel, I expect HP/Intel to license SOI (or having developed its equivalent at the time for McKinley). Thus, status quo.
NetBSD 1.4 runs on the Mac. You have to dedicate an entire hard drive to it (Apple's partition map isn't yet supported). External Device support is kind of sketchy, but then it's a work in progress. I'd expect the G4 Macs to be supported relatively soon.
Also, the Mac OS X BSD foundation is based on FreeBSD 3.x sitting on top of a Mach 3.0 microkernel. It should be out by next spring. I like what I've seen so far.
Macs rock (the user experience and hardware, that is, not the core OS). BSD rocks. Linux.... is all right.
Hi, :(
I've been working with the ti c6x for about a
year now. The reason why no-one will port linux
to it is that it doesn't have an on-chip MMU and
I'm not aware of a simple way of adding an
external one. Also my gcc port is coming along very slowly
Greg
a 64-bit processor that can run its 32-bit predecessor binaries natively? UltraSPARCs have been doing that for years _already_. The 80386 could run 32-bit instructions, and the 80186 16-bit instructions natively. Same thing, bigger numbers. x86 instruction set has been extended to death, though, so Intel had to go for a whole new ball of wax for a 64-bit architecture. POWER architecture was meant to be extended like this.
Well, if you have been paying attention to posts in the past weeks concerning the EV6 bus (Alpha / Athlon), you would know what I am writing here.
What Alpha busses do is have a 256 bit wide bus running at 87MHz, and multiplex the data to a narrower, faster bus. The switch bus operates at 64 bit / 350 MHz or faster. The dual processor Alphas have 2 64 bit PCI, 2 RAM _AND_ 2 CPU busses, all switched, interconnected and multiplexed. It is somewhat like a siamese computer, any part of the computer can ask any other part for any data. I believe that the dual 500 MHz systems have a theoretical 5.6GB/s bandwidth.
When I can by a PPC system from somewhere besides Apple, then maybe /. will start using a different logo.
It is true that Motorola sells PPC CPUs for embedded systems. For example, they are fairly common in switches and routers.
However, the reason why they don't sell any PC style motherboards is because they are restricted by an agreement with Apple that dates back to the beginning of the Power Computing consortium. I believe that the same agreement or similar one prevents IBM from selling PPC motherboards, and that is why IBM is publishing a reference motherboard design to get other MB makers into the business. IBM is permitted to sell AIX PPC systems, but there are restrictions on what they can sell to keep them from competing with Apple in the PC market.
Apple is not our only hope here. Apple will never sell motherboards, only systems, and they will never let IBM and Motorola out of their agreements.
My impression was that the "extensible architecture" quote simply meant that it would be offered as a reusable core. So you could throw on your MPEG decoders, ethernet interfaces, etc.
You may not see this capability used much in the desktop server area, but for Motorola's embedded products - this would just rule!
For those of you not familiar with cores. Its analogous to building a circuit board. Where you've got lots of cores (like chips) with defined interfaces and, when manufactured, they're all wired together at the silicon level - no off-chip drivers to slow you down or draw extra power. Also makes things incredibly compact, especially if 1 chip does everything.
Tom
thats true and thats also why its replaced with MacOS X now which is BSD.
"The most amazing achievement of the computer software industry is its continuing cancellation of the steady and staggering gains made by the computer hardware industry..." -- Henry Petroski
Wait a minute... Does that really make sense?!?
If you look at the various architectures on which Linux runs, there are three varieties, in general:
There are a boatload of IA-32-based Linux systems.
There are a fair number of such systems.
PPC is hard to assess; it is easy to buy a PPC Mac from Apple, but fairly difficult to buy just a motherboard.
If PPC motherboards were readily available, PPC would be vastly more popular...
If you're not part of the solution, you're part of the precipitate.
I'm not sure how fast a 2 GHz CPU will seem unless we start seeing some really fast RAM and BUSes. Right now the fastest bus I've seen is 133 MHz, and the fastest large scale RAM on the horizon is still RAMBUS. Neither of these will allow a 2 GHz processor to run anywhere near it's full potential. Even with lots of cache you would only approach some upper bound that wouldn't be anywhere as high as 2GHz.
There needs to be some other advances, other then jacking up the CPU speed, before processors like this become useful. Maybe it'll be a new RAM or BUS design (RF anyone?), or a new way of dealing with code internal to the CPU to make memory more effecient then using it for cache. Which ever it's time to start looking at them.
The G4, with the AltiVec additions, makes a fantastic DSP development platform. In case you didn't realize, Altivec supposedly completes a multiply-accumulate every clock cycle minumum (It can actually do 4 per clock cycle because of the SIMD architecture). Doing a single-cycle MAC is almost the definition of a DSP. But to really take good advantage of the DSP aspect, you need an honest-to-goodness RTOS.
Now last I looked, TI's 'C6x architecture was headed that way with speed and scalability, but no one is ever going to write or port a general purpose OS (ie, Linux, Windoze, MacOS, Be) for/to that chip.
The G4, however, already runs at least MacOS and Linux. Unfortunately, neither of these are real-time systems. (Has RTLinux been ported to PPC? A great project, if not!) And it's too bad that the situation with Be prevents them from taking advantage of this great chip.
Anyway, I guess my point is just that having a hosted (versus embedded) DSP development system would be just the coolest, and that the G4 makes this possible for wont of a RTOS.
I have not seen anything yet that state explicitly that the G5 will have AltiVec on it. God, I hope so, though.
(drool-mode off)
As long as people don't whine and bitch about not being able to upgrade their already-too-fast-for-their-own-good G4 machines to G5s, that is.
:-]
Hey, anybody want to sell me their "useless" "non-upgradable" blue and white G3? Since the machine is obviously so out-of-date and nonexpandable, I'll give you $500 for it.
Now, as for the G5, provided it supports the same kind of multiprocessor architecture as the G4, I can only guess at what this could mean for applications like Photoshop, Final Cut Pro, After Effects, and Quake. "Omigod!" would be my guess, though.
I use Macs for work, Linux for education, and Windows for cardplaying.
That is, if you believe the IA-64 application developer's guide on intel's website (developer's section). It's a 1.9MB PDF document. For full detail, check out chapter six. You can see the glorius mapping of IA-32 registers, ala PII's MMXregs on FPregs.
What would be interesting is how much extra silicon/cost will it take up? Will the heat dissipation levels be acceptable?
If it makes sense, it brings us to another question, why don't Motorola have an IA-32 unit on their PowerPC chips?
Later
Hasdi
Seeing as overall system speeds double every 18 months, a 2GHz new-generation PPC chip, probably running on a "brand new" 200MHz bus at launch, sounds about right two years from now.
If it's going to be sold as a high-end workstation/server CPU, it should benchmark about 2.6 times faster than the high-end Alpha. If it's going to come out at the mainstream price point, figure on it benchmarking 2.6 times faster than an Athlon 650.
Wake me up when Moore's Law finally breaks. Looks like it still has legs. And no, the zippy faster-bus Mac G4/500s won't do it when they ship in a month or two. Certainly not at twice the price of an Intel box.
Ok, 2GHz, cool...
But what are the everyday uses for a 2GHz processor in your home? I have a 330MHz Celeron in my laptop, and I really can't see why I should want to have a processor six times faster than the one I've already got. Everything above 400-500MHz seems a little bit wasted in a PC.
I can understand why one would want to have a couple of CPUs like that in a number crunching mainframe though...
The glory of this version of the G5 is that it is aimed at your and my desktops! The 64 bit Intel chips are for extremely high end servers, etc.
Incidently, one of the things that got IBM into the ppc game initially was the ability of the design to efficiently use cheap memory. This has evolved and improved over time, so that the G4 can run at 8 times its memory speed without an outrageous performance hit. Intel has problems with multipliers over 4 or 5. Hence, the 2 GHz chip should only need 250 MHz memory bus, and there may be some tricks to play there (?) which will keep the actual memory requirements to less than 250 MHz.
There may be some G5's shipping before 2 years too, only for high end servers, and limited production runs because of this. This Motorola announcement is about production in quantity, not about existence. (The current G4 Mac's are probably pre-production-in-quantity, since Motorola announced opening up the line to produce them about the time Apple announced the G4 Macs, which are shipping even before the G4 comes on-line in the production cycle. The G5 will probably also have a pre-production issue of significant size, but aimed primarily at hardware vendors first intro products.)
CHRP spec is not open. yet. its due in fall99 and you will see mobos soon after.
And AltiVec instructions are not Cray FP type instructions. The only similarity is that they are both SIMD (single instruction stream, multiple data) instructions. Some Cray FP instructions operated on vector data, which is primarily what made them so useful for crunching scientific data very fast. AltiVec instructions are merely Motorola's effort for not missing the SIMD extensions train. Almost all leading architectures have SIMD extensions now for increasing performance in multimedia applications, so Motorola did not want to be left behind, and came up with AltiVec, which is better than most of other companies extensions because it has some instructions traditionally exclusively encountered in the DSP domain...Otherwise it is just a SIMD extension, much like MMX and SSE for Intel, 3DNow! by AMD, VIS by Sun, MVI by Alpha, MIPSV(MDMX) by MIPS, etc..
Zigbee Central: A Zigbee weblog
In the marketting world, yes. Of course the G4 is 128 bit... ;) However, the definition of "how many bits" is in a particular processor is very loosely defined. My rule of thumb: Its the largest integer that the processor can use without taking a performance hit. Usually, its sizeof(int)*8. So the G4 is 32 bits. Now, i'm sure someone will get picky with my def - thats fine. I freely admit its not perfect. But I think it works for most normal cases/users. Tom
Motorola sells a 604e + ATX board for the embedded market, and it costs $2500. The 750(G3) is only available on their VME and CompactPCI boards. They don't (and can't) sell anything for the PC market.
I am more interested in faster system overall then cpu. Except for computation, 2Gig CPU is useless if there is only a 200Mhz bus. I hope that the bus specifications are atleast 1GigHz by then, with faster memory. If they become affordable and graphics and internet connections are fast enough then maybe we'll finally move to video conferencing to the average user.
Only 'flamers' flame!
Where can I get a ATX G3 motherboard TODAY? With the opening of the CHRP spec I would have thought these would be available by now. Any news on were to get one of these beast?
I'm not sure that running 32bit code natively would be advantagious for the IA-64. I think that it is time to say goodnight to x86. It is tied up in so much cruft that it is choking the potential of newer software.
In support of this argument I invoke the memory of the PPro, which sucked when running both 16 and 32 bit code. (Okay, obviously not at the exact same time, in fact switching back an forth is what caused the slow-down.)
Wouldn't you rather run Linux on a box optimized for 64bit operation?
-P