Slashdot Mirror
AMD to Build G4 CPUs?
the eric conspiracy writes "Bloomberg news is reporting that Motorola and AMD are in talks to include manufacturing of Motorola CPUs at AMD's Dresden facility. This could help cash-strapped AMD particularly if its x86 compatible line runs into problems. Motorola and AMD already have cross-licensing agreements - AMD gets its copper technology from Motorola, while Motorola uses AMD's specialized RAM chip technologies.
"
It seems like everyone looking for excess state-of-the-art fab capacity is going to AMD.
I'm not too sure what you mean by that title, but it sounds like you mean AMD will be stamped on the G4. From articles I've read last week seem to indicate that Motorola would buy about 1 billian dollars worth of the fab (I believe it was 50% or close) and they would be produced along-side each other, but still with their respective names. But I do agree, this can only be good. AMD get's help out of their slump, as does the G4 (Will the G5 be made there?). I do hope the best for both companies, and AMD has always been one of my favorites (Mainly because I don't play games :). I agree, good idea, and good luck. I just hope it goes through.
I wonder what this would do to competitive pricing for each respective chip. I'm no economist, so maybe someone has some more insight.
;)
But I'm hoping that this will drive g4 prices down, I'm no mac fan, but their prices are always high. And Hopefully this would cause athlon prices to go down, since amd would be bringing in a profit for a change
Course, a 500mhz athlon IS cheaper then a 500mhz pIII right now, so thats good.... plus its faster.
Also This is REALLY convient (good?) timing for motorola and apple.... since silicon valley's #1 LSD takin' CEO just stated that they were takin' a huge hit in the pocket book due to motorola's poor G4 production numbers. I'm sure the stocks will reflect this news.
Spell checking is over-rated,
Ecc
=]
The stock will go up eventually, but I'm guessing that even if the deal goes through, it'll be a few months before AMD would be able to help get G4's out the door. The stock is down mainly because Apple won't be able to fill demand for this particular quarter.
There's nothing like multiple vendors making the same hardware to help spread the knowledge. Also, as other have mentioned, control of chip supply has always been Apple's bane. It'll be nice to see continued stability for Apple, if only so that I'll still be able to buy kick-butt hardware to run linux/ppc on.
What I _really_ want is a compiler that's efficient and that builds fast code, like gcc/ix86. But I don't want to run linux/ix86. The gcc/ppc people are doing nice work, but need more support.
I am the king... of No Pants! www.penny-arcade.com
I looked over the G4 System Specs, and I can't help but notice that it uses IDE drives. So, of course, I poped over to the "Apple Store" online and checked if you could get one with SCSI drives. Not avaliable.
I thought the iMac was cute, I'll admit it, but Apple started to move away from SCSI then, and it seems they are still moving away. Apple was using SCSI drives exclusively when most companies didn't even offer them as an option, now the tables have turned again. Being that SCSI not only allows higher data flow rates, but off-loads the processing power needed to run the drives from the CPU to the controller card (or subsystem), it will clearly provide a preformance boost (that most people miss the significance of).
So, I guess I have to ask, why is it when Apple is pushing the new G4 as a "Super Computer on a Chip" and pushed the G3 as the "Pentium Beater" that they are takeing steps backwards with other componants?
I have to admit that the G4 price range (starting at about $1500, and topping off under $4000 not counting monitor) is quite impressive considering the claims of the CPU. But it's total system preformance that matters, and when even the $3500 decked out G4 system uses IDE drives, I just have to ask.... "WHY?
I guess when we actually see any G3/G4 generic systems come out, which is "suppose to happen" when IBM released mother board specs, this won't matter. But I haven't seen that happen yet, and since Apple is the primary source of the G4 systems to the consumer, it matters for now.
Lately it has been unusual to have foundries cross the great Motorola-Intel divide... but AMD is the ideal candidate for doing so.
Let's hope this will indeed happen. It's a win-win situation for everybody (except perhaps for Intel, of course)...
SOI drastically increases performance, but there is little hard data other than IBM's own tech reports. I believe IBM is the leading authority on SOI(silicon on insulator). Similar to the situation of Motorola with copper technology, there is no company that "owns the method to SOI"-companies have patents on their ways of implementing SOI or copper interconnects, but since the underlying physical phenomenon is out in the open, you can go and invent a process which uses copper or SOI. Just like copper, other companies will follow suit and introduce SOI processes in a couple of years..
Zigbee Central: A Zigbee weblog
I personally find this very strange. Both AMD and are suffering from capacity shortage problems. AMD's stock is at its pathetic levels because analysts slammed the company when they could not provide enough AMD K6's to the market. Just recently, Gateway announced that it would not be using AMD processors since AMD can not supply them in volume. How can AMD consider sharing its production capacity with Motorola at this critical time when Athlon is just out and in great demand? If they can not meet the demand for Athlon and provide enough Athlons to OEMS, Wall Street will KILL AMD this time. And unfortunately for us techies, technical excellence does not make a company successful by itself. AMD is having a lot of financial trouble, and I don't think taking the risk of not being able to produce enough Athlons is worth the money that Motorola will provide them in return for this favor..
Here is what I believe: There must be something else behind this, if it is true. Motorola must have made a really attractive offer for AMD to have taken this risk..
Zigbee Central: A Zigbee weblog
hmm everyone seems to say copper is good but I don't see any specs
:-)
Electricity passing through copper encouters less resistance and therefore produces less heat than electricity passing through aluminum (the current standard). This is why aluminum wiring in houses caused so many house fires and has been outlawed in the U.S. This is also why copper interconnects on a chip are better than aluminum -- less heat.
Less heat allows faster processors (ask any overclocker), but how much faster depends on specific factors of the implementation. Anything that says "copper processors are x% faster than the same design using aluminum" is BS unless accompanied by three pages of conditions and explanations.
Me, I want gold interconnects
Slashdotters--
It's been my belief that Apple doesn't do well with competition. For example, I remember the "Power Computing" cases where Apple managed to "show" that they had a copyright to anything remotely "Macintosh". This is one reason I was hesitant to believe in the stories that Apple had devoted site space to MkLinux. It doesn't seem to me that Apple is interested in open hardware, software, or standards for that matter.
I've secretly been lusting after a G[2-4] chip for a while now. I'm have yet to try LinuxPPC, but I have heard very good things from graphics artists and designers. This sounds to me to be a PowerPC UNIX box for cheap, a luxury which only the rich normally have available to them.
I sure hope no problems are run into by AMD, and I *really* hope AMD doesn't just put out chips that are merely poor mirror images of an existing chip.
I like AMD as much as the next guy, especially now, with the new processors available and impressive benchmarks galore. But, let's face it, most of us wouldn't have bought AMD if our lives depended on it a year or two ago.
For now, long live the AMD G4.
aÍÍ©ÍÌÍ£Ì'̽ͩÌÍzÍYÌÍÌY
Uhhh, in the short run, AMD may have problems doing anything with the Athlon...all of the recommended MoBo manufacturers for the Athlon are in Taiwan, and those factories are in varying degrees of earthquake-induced non-functionality.
The earthquake in Taiwan is going to put a crimp in many manufacturers Christmas quarters...
Also, AMD licenced the Cu technology from Moto, so it's not all that out of line for this deal to happen.
As the other people said, SCSI *is* a BTO option using a high-end PCI card. But Apple is pushing FireWire as their high-end hard drive solution. As I understand it, it is a higher-performance option, and it is much more flexible and simpler.
One other thing to keep in mind is the Unified Motherboard Architecture. Apple's new computers will soon all be using the same basic motherboard. The iBook and the high-end "Sawtooth" G4's are UMA-based right now, and the soon-to-be released iMac revision will be as well. The Powerbook will be UMA-based early next year.
A big reason for this is to allow economies of scale, both to spread R & D around and to allow cheaper production costs. They took SCSI off the motherboard to save money on the low-end machines, and they're puching FireWire as the replacement on the new G4's.
I often see people comparing ultra66 ide devices to ultrawide or u2 devices, and say "Look, the ide drive goes at 66MB/s, whereas the expensive scsi disk only goes 40MB/s" (or 80 as the case may be).
/older/ scsi drive doesn't even affect my load average, it still hovers around 0.01 or so! Granted, cd writing is only one specific example, but it is still an example.
This is almost completely wrong! These speeds are for the device interface, the speed at which data can be moved on the chain. The speed of the actual drive is independent of this! Most current ide drives have around a 6--10 MB/s sustained transfer rate. Nicer scsi drives are in the vicinity of 20+ MB/s. Also note, that for these speeds, your cpu for ide is having to work very hard, whereas with scsi it may not even be doing any work.
So, if you can only have two ide devices per chain, that means maybe 20MB/s transfer rate. Further, ide is single-tasking: when you do a request to a device, all other devices on the chain must sit and wait for the other device to find what it needs, read it, and send it back. SCSI is multithreaded: you say "hey, give me data", while grabbing data from two other drives, a fibrechannel array, send data to your 8x dvd writer, do ip-over-scsi, and receive the data you just requested!
Lastly, since scsi drives are often targetted at higher-end markets, they tend to be of higher quality than their ide counterparts.
And, finally, a (/very/) brief example:
Burning CDs on my box to my scsi cd writer from an ide drive pegs me around a load of two or three! The machine (a dual 166 pentium) is barely able to respond to mouse events. Doing the same task, but with the data coming from an
Don't get me wrong, ide drives definately have their places, but for large work loads they come nowhere near scsi drives and arrays or fibrechannel drives (which haven't started to really crop up yet).
Just my thoughts on the subject (hope they help).
As for wanting gold interconnects, no you don't. For one thing, copper is a better conductor than gold. Besides that, gold is a disaster in silicon processing (it diffuses like lightning and scavenges carriers. Low transconductance and high leakage everywhere = slow and hot.)
Lacking <sarcasm> tags,
What else were they supposed to use, ST-506? Or should they have simply delayed the introduction of the Mac for six years while they waited for IDE to come out? ;-)
The real question is why did the PeeCee switch from ST-506 (and then ESDI) to IDE, when they could have settled on SCSI instead. The reason IDE is cheap is mainly due to economy of scale, which didn't exist in IDE's early days. I never understood that... but I never understood WinModems either.
---
Have a Sloppy day!
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
After reading the coverage of this story on cnet, I have decided that they are a completely worthless source of news. No where are any real facts given, as the story is based on alleged talks between the chip makers that is loosely based on an agreement of a few months ago.
At the bottom of the article however, is this interesting link to a previous story ran by cnet. Guess what this one says. That Motorola is going to produce chips for AMD. Well, this article is equally as worthless as the current story, as they even in there own article state that its not true. The article is about how Mot is going to make AMD's chips, and then it states that the agreement did not include manufacturing of any of AMD's PC processors.
It is difficult to say whether or not it really is good for the giants of the computer industry combining to form titans. The computer industry right now is screaming Standard Oil, Carnegie Steel, and Pulman Rail. When those conglomerates formed, they didn't work for the good of the industry, but rather increased their profits since they had a monopoly. Granted, some say (not me) the M$ doesn't have a monopoly, and that AOL doesn't either, but they technically don't have to.
Once things settle, they'll all have their own territory. MS and Sun will still make servers and Motorola and Intel will still make processors, but Windows NT could, say, own the small business market while Sun controlls the high-end server/workstation market. Likewise, all PCs might have Intel CPUs while all cars have PowerPCs. So they technically compete, but are stil able to gouge the consumer.
I prefer to see many (ie 5) companies working in the same industry. If Intel, AMD, Cyrix, WinChip, and (maybe) Transmeta were all making compatible chips with each adding features here and there, I think we'd see a much more consumer friendly market.
Apple didn't do anything to the processors themselves. What they did do was include a check for the PPC7400 in the latest version of the Blue&White G3 firmware, mostly (I've gathered) to make certain that no one jumped onto the market with a G4 upgrade before they announced their G4 machines. (There have also been some reports of power supply issues related to the B&W/G4 combination--apparently the blueberry boxen don't take so well to the G4's increased power requirements or somesuch.) This is a moot point, however, as at least one upgrade maker has already developed a firmware patch of their own to bypass Apple's block, as everyone knew would happen from the beginning anyway.
Almost every Mac at the office has external drives connected to it (Audio, Video, and Graphic work). Swapping 18 GB external HDs around is common practice here. Some even have arrays.
Apple has gone with ATA becuase it's cheap and for most purposes just as fast as SCSI, Macs (new Macs at least) don't suffer the same kind of CPU usage with ATA drives that many PCs do. For Apple the low-end get's IDE, the mid-end get's IDE and Firewire ,the high-end gets SCSI (via a PCI card) and Firewire.
Firewire is going to be very cool, the problem right now is there are no native firewire drives, all the firewire drives you see for sale have a SCSI or IDE drive with a IDE/SCSI--FireWire converter in the exclosure. The Sawtooth G4s have an internal Firewire port and once two things happen you're likely to see internal firewire offerings. First someone has to start shipping firewire HDs, second Open Firmware has to support booting from FireWire.
They cheaped out on the Disk subsystem and that's that.
I'd rather see SCSI offered as an Ultra2 LVD dual channel PCI card then the 5 MBps SCSI port seen on the 9600...
"Grab them by the pussy" -- President of the United States of America
Backwards in your opinion. A step forwards in my opinion, and apparently that of Apple's. Sure, SCSI is faster and less CPU-intensive. However, it's also much more expensive. For $150, I can get a 20 gig IDE drive (or bigger). For that same $150, you can't even break 10 gigs on a SCSI drive. So if you want 40 gigs of hard drive space, you have a choice of around $300 for IDE or over $800 for SCSI. IMO, the extra speed isn't worth the extra $500.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Apple is using EGCS/GCC 2.95 as its compiler for Mac OS X [client] (not OS X Server; that uses gcc "2.7.2.1" for now).
They've submitted a large quantity of code (mostly from the work done at NeXT) to the GCC maintainers, and work proceeds to integrate the two source bases.
Furthermore, Apple would be nothing short of braindead to release OS X [client] for G4 systems without using an AltiVec-aware vectorizing compiler to generate their code. Since GCC Is the compiler that they're using, it seems more than likely that they will expend considerable resources to making GCC's PowerPC codegen as good as possible. (And LinuxPPC will see the benefits, too. Very cool.) Perhaps they can even integrate some of their work from MrC/MrCpp, Apple's fantastically good optimizing PowerPC C and C++ compilers for MPW.
- Mali
Actually, the ATA spec has all the multitasking capability of SCSI(disconnect/reconnect(called overlap in ATA spec), command queuing, command reordering). & some ATA drives have the commands which use these features(IBM GXP & GP series, Fujitsu MPE3 series). overlap & tagged command queuing are implemented with ATA's r/w dma queued commands(overlap is a step of the commands & tcq is used if que length>1). If a driver were written to use r/w dma queued, you'd have an ATA driver which supports disconnect/reconnect & tagged command queuing. From looking ATA driver source, it doesn't look like Linux supports the commands. Perhaps I missed it? The ATA spec can be found at t13.org.
Actually the 64 bit extension to PCI is part of the regular PCI spec. The fastest PCI bus is 64 bits wide running at 66 MHz. Some Wintel server machines implement this, as do many Sun SPARCstations. Other RISC workstations probably do too, I just don't keep up with them all.
You may be thinkng of PCI-X, which is an extension to PCI to up the clock rate to 133 MHz. PCI-X may require royalty payments, I'm not sure.
Intel intended to make NGIO an open spec. NGIO and FutureIO recently merged into a single proposal, and I'm not sure where they are going wrt licensing.