IBM’s new $3 billion microprocessor factory in New York – where G5 processors are reportedly manufactured – is losing money, the company announced this week. Reuters reports that production problems are affecting the plant.
IBM’s new $3 billion microprocessor factory in New York – where G5 processors are reportedly manufactured – is losing money, the company announced this week. Reuters reports that production problems are affecting the plant.
Apple’s not selling enough G5s. C’mon IBM. Make a new PC type of platform out of the G5. Start selling chips on the consumer market and encourage motherboard manufacturers to jump on board. Maybe even get a cheap nForce board or something.
I’ll buy one if I can put it together for less than a couple thousand dollars. But I won’t buy an Apple G5 until at least the second generation release and some price-drops. I ain’t payin for beta hardware and software, sorry.
Yet another excuse for Apple not being up to par? A shortage of chips eh?
First I had ot blame Motorola, now IBM?
The Reuters’s article seems to be a lot of FUD. IBM has the very best engineers in the world, the best fab process, deep pockets, and most importantly, the desire to be the best. Aside from AMD who got most of their expertise from Motorola, IBM is the only real competition Intel has when it comes to developing new fabrication processes. Motorola seems to have given up, as they don’t seem very interested in future Semiconductor technology right now, at least outside of cellphones, Analog Devices, National Semiconductor aren’t doing anything new, and outside of chipset production Via doesn’t offer anything noteworthy either,(sorry Cyrix). IBM fabs a lot of important microprocessors. Not only Apple and Nvidia, but Nintendo, Crusoe chips, and themselves as well. I’m probably forgetting other customers too. My point is, I think IBM is committed for the long haul, because this just seems to me to be a part of the heart and soul of the company, to innovate in these kind of things.
…And Microsoft is losing money on the X-box.
Apple’s not selling enough G5s.
Come on, they’re not even out yet…
But I won’t buy an Apple G5 until at least the second generation release and some price-drops. I ain’t payin for beta hardware and software, sorry.
Beta hardware? How can you judge something if it’s not even out…? I totally agree with you that IBM should make a new PC platform, it would be really great, but please don’t spread FUD!
Apple’s not selling enough G5s.
No, the plant is having production problems, which, if you had read the article, you would have known was the problem with the plant, not poor sales.
C’mon IBM. Make a new PC type of platform out of the G5.
What do you want, an ATX form factor PPC970 motherboard? To run YDL on or something? Why not just buy a PC, it would be cheaper…
Start selling chips on the consumer market and encourage motherboard manufacturers to jump on board.
Yeah, everyone knows IBM and Asus have an excellent relationship…
Maybe even get a cheap nForce board or something.
A cheap nForce board? The nForce being a north bridge designed by nVidia for the Athlon? Yes, I’m sure nVidia will custom design a north bridge for IBM, considering the large volume of processors they will move because of the initiative.
I’ll buy one if I can put it together for less than a couple thousand dollars.
Of course you will, you’d sacrifice the extra money instead of building a crappy PC system for the fraction of the price
But I won’t buy an Apple G5 until at least the second generation release and some price-drops. I ain’t payin for beta hardware and software, sorry.
I’m not sure what you’re incinuating here, something about Apple’s hardware quality, something for which they have an excellent reputation? Does Apple sell PowerMacs that just break after a few weeks/months? I don’t think so…
RE:bascule
lets see G5 uses an hypertansport from the controller to the system (I cannot think of this chips name it has the memory bus hanging off and talks to both cpus and uses hyperytransport to talk to a set of AMD chips). Nforce 3 is hyper tansport channel to cpu. Ergo a G5 Nforce 3 board with COTS hardware. sigh.
I personally think the next Apples will split the memory chip into two parts so they can go NUMA and have more CPUs.
Donaldson
then they could begin sellign their desktop systems at low margins.
there are rumors about systems for the big movie houses rangng from 12k and 500k and have up to 64 G5s in them.
It is funny how people who know jack about the industry make such statements, fact: most IC plants take years til they are payed off/are profitable. It is the nature of the beast….
The G5 is brand new, not even intel can pay off a new design in less than a year sheesh. Once the design is payed off, then you can make a profit.
It is called investment, but hey… do not let that affect the PC trolls….
My basic understanding is that factories take several years to become profitable. Is this really news to be all hyped about???
nah, cmon man its IBM they werent profitable on day one of opening the plant, they are going out of business!!!!
And well, apple is always going out of business.
(Im posting from an ibm powered ibook, man im so screwed cause of this horrible news)
At least IBM are profitable, AMD is STILL in the red. though getting closer to zero. They are still losing money at the end of the day.
this is not an article or story about the g5. this is a story about a chip foundry not measuring up. a chip foundry. IBM built this plant to make chips for themselves, and to make chips for designers who could not afford to build their own plants. and it was built mainly for the latter. it is a rough time for the technology industries, less companies are putting money into designs of microprocessors, so obviously the foundry that makes the microprocessors is going to have less business.
there are rumors about systems for the big movie houses rangng from 12k and 500k and have up to 64 G5s in them.
There are rumours USA is annexing China as a state any day now… too…
I really hope they are able to turn this around. I think it is important for small chip designers (companies)to be able to get their chips built (like transmeta). With only two or three companies running the entire show, you could say goodbye to innovation.
So if a tidal wave hits a coast some where it is Apples fault?
The G5 is a new chip getting built in a NEW foundry, what IBM is facing is a situation of EXCESS capacity, HOWEVER, when PowerMac sales rise, this slack will be absorbed.
For me, after now looking at when my tax return will come back – yeah, I know, if they owe you money they take them time but by hooker by crook, if you owe them money they wanted it paid yesturday not today – I am strongly looking at the 1.6Ghz PowerMac. I’ve removed the SuperDrive, and priced out a moderately priced Mitsubshi Trinatron monitor, 19 inch, and it will work out to be around $3,400 for the lot. Pretty good value. Only $150 cheaper than my last computer.
The story is this:
IBM based their pricing on one yeild level, and are getting another. They’re contractually obligated to the current customers at a price, and because of that are making less/loosing money.
nothing else.
Gee, already and just to think they annexed Australia quite recently. Governor Howard is thrilled to be part of the Stars and Stripes and the Australian public are too apathetic to do anything about it. Huray for the lazy state, now we can really sit on our asses and drink Fosters (yuk).
I believe apple makes superior hardware. But i don’t see a Mac club where i can go buy a new board and ram and add a new HD. for a relatively low price. what about modding it? i dont think so, on a MAC. If IBM is having troubles is just a minor sign. I think they will succeed with the people that have enough MULA to dish out for a nice dual g5. BTW g5’s are awesome, just wish i got paid more. Install nice and real *nix.
You know there are also rumors that rajan is a victoria secret underwear model too
“I really hope they are able to turn this around.”
Its not a propblem that needs to be turned around. They don’t anticipate making a return on investment until they can start shipping product in mass quantities… something that hasn’t happened yet because THE PLANT JUST OPENED.
This story is plain FUD.
OSNews shouldn’t have even linked to it.
“I believe apple makes superior hardware. But i don’t see a Mac club where i can go buy a new board and ram and add a new HD. for a relatively low price.”
1) Every new Mac since the 603s has had a motherboard that has been upgradable.
2) Macs use the same ram used by PCs.
3) Macs use the same Hard Drives as PCs
“what about modding it? i dont think so, on a MAC.”
I know so. I’ve bodded my Macs for years.
“If IBM is having troubles is just a minor sign.”
They’re not having troubles. They spent 3 billion dollars on a new chip fab and JUST opened up to start making product. Its going to take some time to recoup that invesment.
“I think they will succeed with the people that have enough MULA to dish out for a nice dual g5. BTW g5’s are awesome, just wish i got paid more. Install nice and real *nix.”
Agreed. Thankfully they are priced incredibley cheap when compared to the competition.
The sky is falling! The sky is falling!
Why should I buy a G5 64 bit, when I can get an AMD 64 bit for 1/10th the price?
And, from what I have hear is that the G5 was built partially on AMD technology.
“IBM’s New G5 Microchip Plant Not Measuring Up”
Hehe. Glad I don’t have a dirty mind. But then IBM makes more money in one year than some countries. Guess they have the measuring stick.
I was sacarstic. Annexing AUstralia is quite possible (remember, they once tried to annex Phillipines), but China is impossible. My point is that Apple haven’t yet came out with quad XServes, it’s a pipe dream thinking about those big big high money supercomputer jobs. And why would IBM allow Apple with their processors try to make inroads in one of their key markets?
Quite unlike. Once they said that it is Motorola that is slacking, IBM would be just alright. And now…?
RE: Gajendra Deshmukh (IP: —.si.rr.com)
One tenth the price? so you’re saying to me now that I can pick up a AMD 64bit workstation for $199.50?
RE: rajan r (IP: —.time.net.my)
For all we know there could be a 32bit G4 coming out by IBM. They’ve added AltVect to Motorollas line, Apple could simply get that added to IBM’s 750 range and voila, IBM becomes Apple supplier of chips.
The IBM PowerPC 750FX would be a good replacement for the G4 in their consumer line. 200Mhz bus and a fairly reasonable life expectancy.
Btw, isn’t the same plant going to be used for the AMD Opteron? And a future small server line of IBM based on the PPC970?
THE PLANT JUST OPENED.
This story is plain FUD.
No, it’s an interesting article. It says IBM’s shares fell despite meeting analyst profit expectations. For god’s sake, you’ve posted far more than the so-called PC trolls, who’ve only posted three times. You’ve even gotten yourself moderated!
My fellow apple fans, I’ve been reading osnews for a year now and many times I’d have to reply to some zealot x86 high school kid about how powerpc prices are not going to drop to x86 prices or how apple will never port os x to amd64 for wide consumer consumption. Many times I get the infamous “apple should… cater to the commodity based consumers and change their business model to suit my specific needs, bargain basement prices for quality not found in x86”
Now it’s gotten to the point where I think it’s becoming rather useless. The more and more I read these replies from those x86 thrifties, the more I find that it’s better not to argue with them (not because they’re right, but because they’re so badly wrong). So why don’t we as a community stop arguing with those guys, and let them post whatever the heck they want. In fact, when you happen to glance over a post like that, just laugh and say, ok, I’m not going to argue with a 12 year old.
It’ll save you a lot of anger and time, because they are millions of those trolls being generated from god knows how many wintel bridges.
from my opinion, it’s probably better that apple charges an arm and a leg for workstation class computers, it’s better for the environment long term because there’s less supply/demand.
IBM seems to innovate just to innovate. Perhaps that is why they are still at or near the top of all they do. Each year they just about get more patents than the rest of the businesses put together. One day management at IBM will ask to see short term returns on their investments and that will be the beginning of the end for them.
…the boys make software. Computer hardware is a killer industry…they’re working with one of the smallest fabs in the world here kids. What’s worse is that hardware works or doesn’t…there’s no patching it after the fact in general. A single mistake costs a hundred shiny new G5 processors. ouch. Now realize that they’re at the bleeding edge of their industry, and the whole article is a non-issue. There’s only a half-dozen fabs in the world that are in that league…IBM’s just a little behind schedule. These things happen.
It’s the Wall Street analysts that are the crackpots here. You’ll notice none of them are actually rich from starting anything, the only money they have come’s from percents of other peoples hard work. For the economy IBM’s doing great…again earning the name Big Blue. The market is crazed thinking they’ll get .com gains in this economy. Were in a giant January after Christmas slump folks…even big companies have to pay the credit card bills after holiday shopping [.com] too! That wall street dosen’t understand this is very “Marie Antoinnette” of them…and look what happened to her.
Why are people surprised that a new plant creating a new processors has not reached full production yield.
Cut IBM some slack they know what they are doing and they have been making microprocessors for a long time.
Besides the PowerPC 970 is part of IBM’s long term strategy for PowerPC Linux so its not like its a problem they don’t want to address. For the time being Apple will probably be the largest consumer for the 970.
Shouldn’t people instead focus on AMD’s losses which are more significant and look to continue? People blabb about Opteron but it looks to me that if AMD tanks further it won’t help. The inherit cheapness of the whitebox PC vendors and community does not help AMDs bottomline and it shows. They in fact need to charge more for the processors.
As far as AMD technology used in PowerPC 970, anyone care to guess who bought out the Exponential patents?
If IBM are having problems it could affect Apple badly. IBM may lose interest in the G5 like Motorola did with the G3/G4. Alternately low yields may affect supplies, slow the roadmap or limit future cost reductions.
Regardless of the merits of the G5, Intel can produce vast numbers of very cheap x86 processors.
Why should I buy a G5 64 bit, when I can get an AMD 64 bit for 1/10th the price?”
Becayse the G5 offers you so much more for the price.
“And, from what I have hear is that the G5 was built partially on AMD technology.”
You’re talking about Hypertransport… and it was only an AMD technology under a consortium
“My point is that Apple haven’t yet came out with quad XServes, it’s a pipe dream thinking about those big big high money supercomputer jobs.”
Why is it a pipedream?
“And why would IBM allow Apple with their processors try to make inroads in one of their key markets?”
IBM is just a chip supplier, Apple can do whatever they want with the chips once they get them.
“Once they said that it is Motorola that is slacking, IBM would be just alright. And now…?”
IBM isn’t slacking. The story isn’t about IBM slacking.
Geeez! This is a perfect example of an article that feeds the trolls who choose only to read the headline.
“For all we know there could be a 32bit G4 coming out by IBM. They’ve added AltVect to Motorollas line”
IBM added AltVec to Motorollas line?
1) Its called AltiVec… not AltVect
2) IBM didn’t add AltiVec (or AltVec) to Motorola’s line. I don’t even understand what you’re trying to say here.
“Apple could simply get that added to IBM’s 750 range and voila, IBM becomes Apple supplier of chips.”
IBM has already added AltiVec to its G5s… and AltiVec is not a Motorola technology anyways…
“The IBM PowerPC 750FX would be a good replacement for the G4 in their consumer line. 200Mhz bus and a fairly reasonable life expectancy.”
Motorola is already in the process of making such a chip actually.
Despite the woes Apple experienced from motorola, the company is still benefitted by getting its chips from two different suppliers.
“No, it’s an interesting article. It says IBM’s shares fell despite meeting analyst profit expectations.”
I didn’t say that it wasn’t correct with regard to IBM’s profits. I said that it was a FUD article because it talks about how IBM is having problems with its new plant because it hasn’t seen a profit from the plant yet. THE PLANT JUST OPENED UP! The article is FUD because there is no way for a plant of that size to make a profit so close after opening its doors. It makes me wonder which company sponsored the article?
“For god’s sake, you’ve posted far more than the so-called PC trolls , who’ve only posted three times.”
The so-called PC trolls are PC trolls in deed. And there is nothing wrong with posting freequently.
“You’ve even gotten yourself moderated!”
And yet not for a legetimate reason. My modded down comment was pointing out the fact that this is a FUD article and that OS News shouldn’t have posted it anyways. If OS News likes to post FUD articles it seems only logical that they would mod-down my post that points this out now doesn’t it.
“I’d have to reply to some zealot x86 high school kid about how powerpc prices are not going to drop to x86 prices”
I’m assuming you’re talking about the whole computer and not the chip. If this is the case, Apple’s hardware are ALREADY at the same price as PC hardware and have been for quite some time. The only instance where this wan’t the case was with Apple’s professional towers. Now, Apple towers are actually far less than the x86 towers.
Its important to remember that now that the G5s are here, the entire lineup of Apple’s computers have prices that are either onlsy slightly more expensive, the same price, slightly less expensive or significantly less expensive when comparing the exact same hardware and software configuration (or as close as possible) to that of a PC.
If you’re talking about the chips themselves, then you’re wrong here too. The PowerPC chip has always been a significantly less expensive chip to manufacturer than that of x86. Its only now with the economies of scale that Intel and AMD can offer x86 chips at the same rate as that of PowerPC chips.
“So why don’t we as a community stop arguing with those guys, and let them post whatever the heck they want.”
The reason why we can’t (and won’t) is because posts such as those you mentioned are of the type that are typically also intertwined with so much other inacuracies that resulting post is little more than mere FUD. A large part of Apple’s current market position is due to the spread of FUD which which much of the public believed thus resulting in a self fulfilling profecy resulting in Apple’s temporary decline. Apple has emerged from that era, and you can bet that that FUD wont be left uncorrected now.
“It’ll save you a lot of anger and time, because they are millions of those trolls being generated from god knows how many wintel bridges.”
And there are an equally gowing number of individuals willing to circumvent their efforts.
“from my opinion, it’s probably better that apple charges an arm and a leg for workstation class computers”
Why would they do that? That would be idiocy.
Thankfully, their prices are right on par with those of comperable PCs… often times far less.
Your post itself can be labeled as a thinly veiled attempt at FUD.
“If IBM are having problems it could affect Apple badly.”
Thankfully they are not. The current situation is one that is planned for any new fab plant.
“IBM may lose interest in the G5 like Motorola did with the G3/G4.”
Motorola didn’t produce the G3… that was IBMs… and Motorola didn’t lose interest in the G4, the lost the means to invest resources in it because they were hemoranging money from so many other areas.
“Alternately low yields may affect supplies, slow the roadmap or limit future cost reductions.”
Thankfully this hasn’t happened and is unlikely to be a factor because Fishkil is a start of the art wafer fab plant. The current situation at Fishkill is one that is to be expected from any new fab. This article is very misleading.
“Regardless of the merits of the G5, Intel can produce vast numbers of very cheap x86 processors.”
As can IBM produce vast numbers of very cheap PowerPC processors. For that matter, Motorola can continue to produce vast numbers of G4 processors at an equally low price. Remember, the G4 is a significantly less expensive processor to produce. its only with the economies of scale that x86 is able to match the low prices of PowerPC chips.
Osnews must of need some traffic, well it got it with this story!
“Why should I buy a G5 64 bit, when I can get an AMD 64 bit for 1/10th the price?”
Give me prices fanboy, and I might just buy one. Does this AMD 64 have a OS and software to run on it? Can I run photoshop, after effects and Maya on it?
On a G5 which is a true 64 bit I can.
I’ve been reading these posts so far and it appears that we have a large and vocal portion of Apple fanboys who “just don’t get it”(tm).
They spout rubbish regarding issues that were valid 6 years ago, but have been addressed and no longer an issue in x86 based computers sold today.
Plug ‘n Pray was a nightmare, however, it has now been replaced by ACPI which combines the two functions, Power Management and Automatic Configuration into one. If you purchase a quality computer you will have a quality BIOS with a quality ACPI implementation.
For example, my brother owns an IBM Netvista PC, runs Windows 2000 (with service pack 4) and hasn’t suffered any of the so-called “issues” which the pro-Mac side of the computing fence try to promote. I am running a Dell XPS T550 and have yet to have one hardware conflict with any device I have added.
Regarding to system stability, again, this all comes down to buying a quality system. If you buy a system off some dodgy dealer, what do you expect? if you purchase a system of a reputable dealer you won’t suffer from system stability.
Regarding system speed. Yes, you can say, “well, do I really need 3Ghz?”, the question is not whether you need it NOW but whether you may need it in 2-3 years time. A 1Ghz Apple iMac/eMac may sound great on paper, however, what happens when the latest MacOS is released and it only runs acceptiably on your configuration?
With Apples continuous focus on the pro range, they are gradually losing touch with their grass roots customers who don’t feel inclined to spend another AUS$1000 just so they can get a machine that has a reasonable life span. Lets look at the iMac/eMac, when were they last upgraded? why didn’t Apple update their line up when they released the G5?
As for the operating system itself. I hear the same red herring being thrown into the ring everytime someone mentions it. Things like:
“Or Macs can be for people who are tired of the clogged Windows Registry, who want an OS that doesn’t have to be reinstalled every other month, who want total ease of use and never want to see a command console (that throws out Linux), or users who just want to play around with something different.”
from http://www.neowin.net/comments.php?category=main&id=12710#comment12…
trying to make out that for some reason, through some “act of god”, their how computer gradually slows down over time, which is a crock of crap. The ONLY programme I have come across which causes problems is Easy CD Creator. Anyone who runs it btw, should suffer every problem it causes. People who do complain install tonnes and tonnes and tonnes of shareware, registry tweakers, “system stability improvers” and other pointless pathetic crap which do 0 to improve the system. In the 8 years I have owned a PC I have NEVER felt the desire to go out and buy a copy of Norton Utilities.
I’ve had this copy of Windows 2000 running for 4 months, and again, I have not experienced any of this BS “slow down”. Defrag your HDD and don’t install crappy substandard software and you won’t have problems.
“Does this AMD 64 have a OS and software to run on it?
Windows XP Pro can run on it… however that isn’t a consumer OS
” Can I run photoshop”
Yes. he would be able to
“after effects”
Yes he would.
“and Maya on it?”
Yes he would
[i]”On a G5 which is a true 64 bit I can.”
Yes, you can, but make no mistake… AMD64 can also. However, this doesn’t justify his argument that an equivilent AMD 64 bit can be had for 1/10th the price… as that is false.
The object of this chip foundry was to produce custom chips for outside customers, as well as to supply their own needs. The issue is that they’re currently having problems getting a sufficient yeild on their chips to sell them profitably. It has nothing to do with how well Apple’s computers are selling. In fact, within IBM, the relationship with Apple is being touted as an example of the type of relationship IBM should be trying to establish with it’s other customers. It isn’t losing money because of Apple.
Also consider, it only lost $45 million on an investment of $3 billion. While not as good as a profit, that’s pretty much peanuts in context. Custom services is a new business for the Microeloctronics division, and this is the first year that foundry has been operational. As soon as they work the bugs out of their production processes, I have no doubt the new foundry will be contributing nicely to their bottom line.
“I’ve been reading these posts so far and it appears that we have a large and vocal portion of Apple fanboys who “just don’t get it”(tm).”
I havn’t seen any “fanboys” yet there have been several posters who “just don’t get it”(tm).”
“They spout rubbish regarding issues that were valid 6 years ago, but have been addressed and no longer an issue in x86 based computers sold today.
Plug ‘n Pray was a nightmare, however, it has now been replaced by ACPI which combines the two functions, Power Management and Automatic Configuration into one.”
I don’t recall seeing one post about Plug and Play. However, plug and play on a PC still has room for improvement.
“my brother owns an IBM Netvista PC, runs Windows 2000 (with service pack 4) and hasn’t suffered any of the so-called “issues” which the pro-Mac side of the computing fence try to promote.”
What issues would those be?
“I am running a Dell XPS T550 and have yet to have one hardware conflict with any device I have added.”
Congratulations
“Regarding to system stability, again, this all comes down to buying a quality system. If you buy a system off some dodgy dealer, what do you expect? if you purchase a system of a reputable dealer you won’t suffer from system stability.”
it would seem that the better buy is had when you buy quality parts in the first place… something that Apple ships with every machine and at prices which are comparable to those on PCs
“Regarding system speed. Yes, you can say, “well, do I really need 3Ghz?”, the question is not whether you need it NOW but whether you may need it in 2-3 years time. A 1Ghz Apple iMac/eMac may sound great on paper, however, what happens when the latest MacOS is released and it only runs acceptiably on your configuration?”
One of the great things about the Mac OS is that with each successive release, the OS is FASTER. Granted, this wasn’t the case with OS X which did in fact require newer hardware… but once you graduated to the hardware that runs OS X, each successive release runs more efficiently… not less so. This is the way it was on Apple’s Pre OS X OSes as well.
As you eluded to, this isn’t the case with Windows which makes your argument an understandable one to adopt.
“With Apples continuous focus on the pro range, they are gradually losing touch with their grass roots customers who don’t feel inclined to spend another AUS$1000 just so they can get a machine that has a reasonable life span.”
Apple focuses on Pro and consumer lines. This shows no sign of stopping. Besides, macs are known for having a much longer life span that PCs anyways… so your argument is moot on both fronts.
“Lets look at the iMac/eMac, when were they last upgraded? why didn’t Apple update their line up when they released the G5?”
Because they were announcing a lot of hardware already at that trade show. Additionally, that trade show was for developers. It wouldn’t have made sense to upgrade a consumer computer at a gathering intended for professionals. Apple will likely update the iMacs mcuh the same way they have been doing for the past few months… upgrade them without a trade show and only a press release.
“As for the operating system itself. I hear the same red herring being thrown into the ring everytime someone mentions it. Things like:
“Or Macs can be for people who are tired of the clogged Windows Registry”
Seems like a logical selling point to me.
<i.”who want an OS that doesn’t have to be reinstalled every other month”[/i]
Whomever said that was wrong.
“who want total ease of use and never want to see a command console (that throws out Linux)”
OS X is arguably the easiest to use operating system on the planet today… and yet its users don’t have to use a command console, however, they can if they so choose. Tell me, how does that throw out Linux? Linux users have GUIs too. Many of them in fact.
“or users who just want to play around with something different.”
And that’s an incorrect argument because….?
“I’ve had this copy of Windows 2000 running for 4 months, and again, I have not experienced any of this BS “slow down”. Defrag your HDD and don’t install crappy substandard software and you won’t have problems.”
Sounds like good advice to me. : )
CooCooCaChoo, nobody was making the arguments that you were countering… lets not turn this thread into a Mac vs. PC debate.
Obviously you skip the whole thread and post some cheap rhetoric at the end. How about READ the thread rather than doing what some nimrod did in the pervious Mac related thread and moaned about video cards on the Mac even though I had actually covered it more than once within the same bloody thread.
READ –> THINK –> THINK –> WRITE –> POST
“Obviously you skip the whole thread and post some cheap rhetoric at the end.”
I read the whole thread and nothing about my post at the end was “rhetoric”. Do you even know what that means (without looking it up?)
“How about READ the thread rather than doing what some nimrod did in the pervious Mac related thread and moaned about video cards on the Mac even though I had actually covered it more than once within the same bloody thread.”
If you have issues with an individual response, break down their post line by line as I did with yours.
Your post eluded to a large voval group of Mac users making these comments. This of course in’t the case.
When you make broad generalizations such as you did, you can expect to get put in your place… as you were.
“the question is not do you need 3GHz now but 2-3 years from now.”
ROTFLMAO
you realise that an X86 PC in 3 years will be replaced in most folks homes who will actualy demand the full load of the processor in that time frame.
so in 3 years Intel will have 5-6 GHz and IBM will have 3-4 GHz.
if in that time frame their is a need for 3 GHz, then the Apple computer will be an appropriate buy.
thank-you for re-enforcing the point that no one is even close to needing the top speed proc.
For me, I have owned this current computer for 4 years. I see no reason why I should be forced to upgrade every 2 years before Bill Gates gets his creative juices into overdrive and expecting all and sundry to upgrade for his “latest and greatest” creation.
What I am trying to find in these discussions are reasons why Apples share isn’t increasing. I am trying to put arguments out here on the forum to try to rationalise why people haven’t migrated. For me, I’ve already made my mind up and now I am trying to decide between either an eMac or save up a little longer for a PowerMac G5, so the issue isn’t about me, the issue is about why the market works the it is working.
This is just the kind of thing that the politicians in the IBM corporate machine will use to prove that IBM should get out of the PPC business once and for all.
No indication yet it will go that way, but we’ve seen this kind of crap happen before. Someone with a cooler and wiser head on their shoulder prevails and gets a valuable product to be continued while the opponents of it ensure that the product fails in such a spectacular way that they can cry foul with “We told you so!” and fire the people who are “responsible” (namely, the cooler and more wiser heads).
you want the reasoning. it is realy simple.
you are trying to come up with some argument as if there is some problem with the product. there is no problem with the product.
the price is the same as comparable PC hardware. the OS is (depending on what you like) one of the best ever built. the software that comes with it is more than you can get from a basic window licence and better than you get from an x86 hardware OEM. al the best games are on the platform. all the best software is on the platform, some of which is not available in x86.
the week point for Apple, as have been sited is marketing. nothing else.
if Apple spent the cash they had on marketing and mindshare and had an advertisement every other comercial break on most TV stations out there and the comercials showed the types of things the Dell comercials show, apple would begin moving up the ranks.
perhaps they will begin such a campagn when there is an apple store in most cities….perhaps it will take some young exec to point out to steve that they need to move in that direction in marketing…who knows….
but the problem is marketing pure and simple.
excuse me? IBM makes most of their profits from Power and PPC lines.
hell, he 970 and the 980 will be a base for not only the Apple computers, but also network hardware, lower end workstations that run Linux, and countless other jobs.
IBMs Power/PPC devision is 10 times larger than their x86 devision. there is no way they are getting out.
“What I am trying to find in these discussions are reasons why Apples share isn’t increasing.”
They are increasing in every area where they are significantly better. Apple’s laptops are WAY ahead of the competition in sales. They are remaining about the same with their consumer computers and they took a massive decline in tower sales because the past year and a half, their towers were inferior. Now that the G5 has been introduced, you will see Apple gaining market share very gradually every quarter.
“I am trying to put arguments out here on the forum to try to rationalise why people haven’t migrated.”
If the “more expensive than PCs” argument were true, then Apple wouldn’t be so far ahead in sales in portables.
What this shows is that people ARE willing to pay for added benefit if it is significant even though some machines can be had for less.
The G5 towers (like Apple’s laptops) are an example of consumers seeing the significantly added benefit and buying them… even though Apple doesn;t sell bare-bones-laptops that can be had for less.
<i.”For me, I’ve already made my mind up and now I am trying to decide between either an eMac or save up a little longer for a PowerMac G5″[/i]
I’d go for the G5 if I were you… although the eMac is a nice machine in its own right.
“This is just the kind of thing that the politicians in the IBM corporate machine will use to prove that IBM should get out of the PPC business once and for all.”
No it isn’t. This shortcoming is one thats anticipated as a new wafer fab of Fishkills calaber is produced. PowerPC has proven itself time and time again. There is no reason why anyone at IBm would feel compelled to abandon it.
“No indication yet it will go that way, but we’ve seen this kind of crap happen before.”
I’m assuming you’re talking about motorola.
motorola didn’t “abandon PPC because it doesn’t sell, or that it doesn;t have a good road map or any similar reasons. (As a matter of fact, motorola hasn’t abandoned PPC at all. They were however hemoraging money and had to rerout their assets to the most profitable sectionsof their business… which caused their G4 R&D to languish for a couple years.
Can you lot renember the P3 1.3Ghz chips.
They had teathing problems and lost intel money.
Give the G5 time.
This is just the kind of thing that the politicians in the IBM corporate machine will use to prove that IBM should get out of the PPC business once and for all.
No indication yet it will go that way, but we’ve seen this kind of crap happen before. Someone with a cooler and wiser head on their shoulder prevails and gets a valuable product to be continued while the opponents of it ensure that the product fails in such a spectacular way that they can cry foul with “We told you so!” and fire the people who are “responsible” (namely, the cooler and more wiser heads).
LOL, you’ve really made my day. I was looking for a good piece of satire, and I’ve found it here.
The UNIX market is approximately $20billion dollars. That is a fair amount of lose change in anyones book. Do you really think IBM is suddenly going to roll over and simply GIVE the $20billion market to SUN Microsystems without a fight?
IBM has finally sprung back into life and in the future, you’ll see more of AIX, Linux and PowerPC. Yes, they have mentioned Itanium, however, they have also mentioned the fact that the PowerPC platform has alot of life left in it and they will not dump it until it becomes completely and utterly unviable. Considering that they are at the bleeding edge of atomic and dna computing, in 10 years time, for all we know they may have an atomic computer proto type ready for the market place.
“so in 3 years Intel will have 5-6 GHz and IBM will have 3-4 GHz.”
Uhmmm… You dont know much about the PowerPCs 970, 980, 990 and 9900 do you?
You might want to rethink your clock speed estime for 3 years from now, with IBM in the lead.
“The UNIX market is approximately $20billion dollars. That is a fair amount of lose change in anyones book. Do you really think IBM is suddenly going to roll over and simply GIVE the $20billion market to SUN Microsystems without a fight?”
Tell me, why are you laughing… You quote me, but you reference Jace. I didn’t say that, nor did I even imply that? What are you talking about?
Uhmmm… You dont know much about the PowerPCs 970, 980, 990 and 9900 do you?
…and neither you do. They’re only making theorical claims for the moment. Remember when the G4 was supposed to be the chip crushing any Intel offering? How many chips/OSes were theorically perfect but were killed when people tried to implement them pratically?
You might want to rethink your clock speed estime for 3 years from now, with IBM in the lead.
Remember when the P4 came out at a blazing 1.5GHz but was constantly beaten by the Athlon Tbird at a puny 1GHz? Clock speed doesn’t mean anything.
What are you crapping on about? the text that is italic is the text I quoted from Jace’s post, not yours.
“The UNIX market is approximately $20billion dollars. That is a fair amount of lose change in anyones book. Do you really think IBM is suddenly going to roll over and simply GIVE the $20billion market to SUN Microsystems without a fight?” <– Is my text.
italic = quoted
non-italic = my text
Well $45million loss is actually about a $345 million economic loss if you are looking for a 10% ROI.
Still not that bad though.
“Remember when the G4 was supposed to be the chip crushing any Intel offering?”
As we’ve learned with what we’ve seen with IBM, the PPC architecture is certinly capeable of it. Just because Motorola came across hard times and couldn’t bring the G4 to its fullest potential doesn’t mean that PPC is inferior.
“How many chips/OSes were theorically perfect but were killed when people tried to implement them pratically?”
Im sure there are some, but there are none that come to mind.
“You might want to rethink your clock speed estime for 3 years from now, with IBM in the lead.”
His argument is much more likely than you give it credit. Intel only managed to increase the Pentium’s speed by extending the pipeline… not by making the chip itself any faster. In reality, a P3 at the same MHz could achieve the almost identical speeds as the P4..
“Remember when the P4 came out at a blazing 1.5GHz but was constantly beaten by the Athlon Tbird at a puny 1GHz? Clock speed doesn’t mean anything.”
It CAN increase the speed, but it doesn’t nesserally mean that it will (as is the case with Intel’s latest chips. However, typically increases in chip architecture occur in paralyl to that of MHz on the PPC side, so it is far more likely that you will see dramatic speed improvements when MHz are added on the PPC side.
To put it simply, PPC MHz probably more indicitive of a computers overall speed than what is the case with x86 chips.
“What are you crapping on about? the text that is italic is the text I quoted from Jace’s post, not yours.”
Then why did you quote my text?
The text I quoted isn’t yours. THis is the original post:
Lovely…
By Jace (IP: 204.186.253.—) – Posted on 2003-07-19 16:51:40
This is just the kind of thing that the politicians in the IBM corporate machine will use to prove that IBM should get out of the PPC business once and for all.
No indication yet it will go that way, but we’ve seen this kind of crap happen before. Someone with a cooler and wiser head on their shoulder prevails and gets a valuable product to be continued while the opponents of it ensure that the product fails in such a spectacular way that they can cry foul with “We told you so!” and fire the people who are “responsible” (namely, the cooler and more wiser heads).
Get a bloody clue. Again, that is NOT your text. Read and repeat.
I mean the consumer PPC business. I really fluffed that comment by not specifying that. We saw, for some time, how the consumer PPC business was questionable (basically about only one acceptable buyer: Apple). The PPC for servers market isn’t the concern and I am aware of how healthy it is. Sorry for the mistake in my post. The rest of it, I still feel is valid.
As we’ve learned with what we’ve seen with IBM, the PPC architecture is certinly capeable of it. Just because Motorola came across hard times and couldn’t bring the G4 to its fullest potential doesn’t mean that PPC is inferior.
Oh, but I didn’t said that. I’m just saying that they were bragging that the G4 was going to be THE x86-killer and it didn’t killed anything. The same thing could happen with PPC 970, 980 & 990 as Intel is also busy to evolve their chips. They might look good on paper, but we still have to see them in action. Theorically, they are much better than x86s, but it doesn’t mean that they’ll pratically be. We’ll see in the following years.
Im sure there are some, but there are none that come to mind.
AFAIK, the Mach kernel was a great idea but was killed because it was too hard or something like that (someone correct me if I’m wrong!). I’m sure there are even more that we never heard of.
His argument is much more likely than you give it credit. Intel only managed to increase the Pentium’s speed by extending the pipeline… not by making the chip itself any faster. In reality, a P3 at the same MHz could achieve the almost identical speeds as the P4..
Well, a P3 at the same MHz was FASTER than a first-gen P4 at an identical speed (some people underclocked ‘Intel Confidential’ P4s). Intel just decided to take another approach. The P4 wasn’t made to be efficient. It was made to have a high clock speed and to be efficient only with optimised instructions (SSE/SSE2). I doubt that a P3 would be a match for the second-generation P4 (Northwood w/HT & 512KB cache) even at the same MHz. The Prescott looks even better (1MB cache, improved HT, new instructions).
It CAN increase the speed, but it doesn’t nesserally mean that it will (as is the case with Intel’s latest chips. However, typically increases in chip architecture occur in paralyl to that of MHz on the PPC side, so it is far more likely that you will see dramatic speed improvements when MHz are added on the PPC side.
I agree, but my point is that MHz doesn’t matter. I could build a chip that could run at 1 bazillion Hz but that would only be faster than a underclocked 386SX. You can’t compare fairly two different architectures by their clock frequency.
To put it simply, PPC MHz probably more indicitive of a computers overall speed than what is the case with x86 chips.
What you say is “my clock frequency is more indicative than yours”. IMO, clock frequency is NOT the way to measure performance at all. You can’t compare a 486 with a 68040 or even a G4 with a P3 with the clock speed.
[i]I mean the consumer PPC business. I really fluffed that comment by not specifying that. We saw, for some time, how the consumer PPC business was questionable (basically about only one acceptable buyer: Apple). The PPC for servers market isn’t the concern and I am aware of how healthy it is. Sorry for the mistake in my post. The rest of it, I still feel is valid.[i]
The important point here is that the PPC 970 is aimed both at the consumer market and the server market.
The PPC 970 is intended for IBM’s upcoming range of blade servers and lower end pSeries servers that are designed to compete with Intel systems. The growing importance of blade servers emphasises just how important the PPC 970 is to IBM’s overall strategy and suggests that it is as important to them as it is to Apple.
The addition of AltiVec or VMX in IBM terminology, while it is there to support Apple’s multimedia requirements and is normally of no use in a server, does however give it some advantage to IBM in the scientific supercomputer area. Hardware vector processing is useful for vectorizable numerically intensive compute. This is important in current life sciences supercomputing (ie. genomics and proteomics and also of course quantum chemistry). They will be able to use this in the large supercomputing clusters that they have been selling to the pharmaceutical industry, universities and major government labs. Without it Intel chips would have an advantage. IBM is a big player in this supercomputer area. So just one chip that’s important for notebooks, blade servers and supercomputer clusters because its low power consumption allows one to pack a lot of them in a small space.
Reposted – sorry I really do know how to complete a tag 😉
I mean the consumer PPC business. I really fluffed that comment by not specifying that. We saw, for some time, how the consumer PPC business was questionable (basically about only one acceptable buyer: Apple). The PPC for servers market isn’t the concern and I am aware of how healthy it is. Sorry for the mistake in my post. The rest of it, I still feel is valid.
The important point here is that the PPC 970 is aimed both at the consumer market and the server market.
The PPC 970 is intended for IBM’s upcoming range of blade servers and lower end pSeries servers that are designed to compete with Intel systems. The growing importance of blade servers emphasises just how important the PPC 970 is to IBM’s overall strategy and suggests that it is as important to them as it is to Apple.
The addition of AltiVec or VMX in IBM terminology, while it is there to support Apple’s multimedia requirements and is normally of no use in a server, does however give it some advantage to IBM in the scientific supercomputer area. Hardware vector processing is useful for vectorizable numerically intensive compute. This is important in current life sciences supercomputing (ie. genomics and proteomics and also of course quantum chemistry). They will be able to use this in the large supercomputing clusters that they have been selling to the pharmaceutical industry, universities and major government labs. Without it Intel chips would have an advantage. IBM is a big player in this supercomputer area. So just one chip that’s important for notebooks, blade servers and supercomputer clusters because its low power consumption allows one to pack a lot of them in a small space.
“I’m just saying that they were bragging that the G4 was going to be THE x86-killer and it didn’t killed anything.”
I don’t think anyone had any illusions that the G4 was going to “kill” the P4, as in the sense, people wouldn’t use the P4 anymore. Rather, if anyone used the word, “killed” it was used in the context that the G4 would significantly outperform the PIII… The G4 did out perform the PIII and the PIV (at the beginning of its life). Hod Motorola stuck to its roadmap, it would have “killed” (outperformed) the PIV, but because Motorola was hemorrhaging money, they had to allocate their resources elsewhere and the G4 didn’t live up to its capabilities. Make no mistake about it though, the reason why the G4 didn’t kill (outperform) the Pentium was NOT because of the PPC’s supposedly inferior design as many have suggested..
“The same thing could happen with PPC 970, 980 & 990 as Intel is also busy to evolve their chips.”
As long as Intel continues using x86 as the foundation for its technology, it is highly unlikely. Intel’s other chip technologies which aren’t x86-based, leave much to be desired. So, unless market dynamics change significantly in the way of IBM’s relationship with Apple or IBM experiences massive financial problems, it is unlikely. PowerPC will likely be both the MHz/GHz champion as well as the speed champion.
“They might look good on paper, but we still have to see them in action.”
No, we say them in action and the results shown by several different application developers, a Nasa scientist and an independent research company all agree that the G5 is in fact faster.
“Theorically, they are much better than x86s, but it doesn’t mean that they’ll pratically be.”
“practically” speaking, the chip is:
fully 32bit and 64bit native (no emulation) = Practical
Very cost effective = practical
Very scalable = practical
The list goes on.
lack of practicality is not a factor here.
“We’ll see in the following years.”
I suppose we will.
“AFAIK, the Mach kernel was a great idea but was killed because it was too hard or something like that (someone correct me if I’m wrong!). I’m sure there are even more that we never heard of.”
What? The Mach kernel was never killed, nor is it “too hard”
“A P3 at the same MHz was FASTER than a first-gen P4 at an identical speed (some people underclocked ‘Intel Confidential’ P4s). Intel just decided to take another approach.”
The approach was to sell people on MHz and GHz rather than actual speed. Surprisingly, it worked.
<i.”The P4 wasn’t made to be efficient. It was made to have a high clock speed and to be efficient only with optimised instructions (SSE/SSE2).”[/i]
Efficient is a relative term that can only be truly gauge in relation to its competitors. For example, instructions optimized for the G4 with AltiVec were and have always been more efficient than a a comparable x86 chip.
“I doubt that a P3 would be a match for the second-generation P4 (Northwood w/HT & 512KB cache) even at the same MHz. The Prescott looks even better (1MB cache, improved HT, new instructions).”
I agree, but the speed increase that will be had from these chips is not relative to the MHz that one might have experienced with previous versions of Intel’s chips simply because x86 doesn’t have the same potential for growth as does PPC.
“I agree, but my point is that MHz doesn’t matter.”
MHz doesn’t matter when a computer’s speed doesn’t run in parallel with its MHz. For example… unless a PIV @ 4GHz is 2x the speed of a PIV @ 2GHz then MHz doesn’t matter.
However, MHz DOES matter if the rate of speed moves in parallel with the MHz. Then, it would be a much closer gauge for the average consumer to determine a chip’s overall speed.
IBM has said that the Power5 is 400% faster than the Power4.What this means is that where Intel increased MHz without speed, IBM will be increasing MHz with speed. This means that the Power series of chips allow for MHz/GHz to be a much more accurate way to gauge a processors than is the case with x86 chips.
“I could build a chip that could run at 1 bazillion Hz but that would only be faster than a underclocked 386SX.”
Actually, on x86 chips there is a wall at which increasing the Pipeline stages to increase MHz will actually DECREASE the speed of the chip. This is the point where Intel is at now.
“You can’t compare fairly two different architectures by their clock frequency.”
This is very true, and yet the 68030, 601, 603, 604, G3 and G4 were all incorrectly judged in this way.
The ironic things is that unlike unlike the G4 and P4, the G5 and all subsequent revisions to that chip architecture will increase in MHz and Speed. This means that a 4GHz G5 will not only be double the speed of a 2GHz G5, IBM’s roadmap suggests that it such a chip would actually be 16x as fast! (Remember, the IBm roadmap tells us that future versions of the g5 (980?) will be 400% faster at the same MHz. When the MHz is doubled at 400 percent then you get a 1600% speed increase!
“What you say is “my clock frequency is more indicative than yours”. IMO, clock frequency is NOT the way to measure performance at all. You can’t compare a 486 with a 68040 or even a G4 with a P3 with the clock speed.”
For those chips you are correct, but for the G5 and all subsequent revisions to the chip, it certainly is.
you missed the forest through the trees my friend.
I was not making a my cpu is better than your cpu argument.
the statment said that a 3GHz CPU today will be needed in 3 years.
hence my argument that since 3GHz will be needed in 3 years when intel will have 5 or 6 GHz, why buy power that you don’t need? or why care if you are paying for just the power that you do need (in the case of a Mac)
I was pointing out that the argument that “the speed is not needed” was supported by what coocoocachoo had said in an earlier post in which coo was trying to refute said claim.
no, the rest of your comment is not still valid. IBM is going to use the 970 and all the future iterations as low end Linux servers, so Apple just rides along for the fun of it and orders CPUs that IBM is developing for internal use anyway.
no, the rest of your comment is not still valid. IBM is going to use the 970 and all the future iterations as low end Linux servers, so Apple just rides along for the fun of it and orders CPUs that IBM is developing for internal use anyway.
IBM has already announced its forthcoming range of PPC 970 based blade servers – blade servers are not just low end servers they are aimed higher up in the enterprise.
As for my comments on supercomputing clusters – it was an IBM spokesmans comment that AltiVec being useful for life science computing that drew my attention to the advantages of the PPC 970 for scientific supercomputing. I am willing to bet that in a year or two IBM will be shipping supercomputing clusters with about a thousand processors.
The whole point of my post was to emphasise how important the new Power 4 line is to IBM and especially the PPC 970. The East Fishkill plant is pretty important to IBM’s own internal server stategy and there is no question that they will stop pushing ahead with the architecture .
mean the consumer PPC business. I really fluffed that comment by not specifying that. We saw, for some time, how the consumer PPC business was questionable (basically about only one acceptable buyer: Apple). The PPC for servers market isn’t the concern and I am aware of how healthy it is. Sorry for the mistake in my post. The rest of it, I still feel is valid.
so you are Jace?
I was commenting to Jace not to you. I agree with you.
” Plug ‘n Pray was a nightmare, however, it has now been replaced by ACPI which combines the two functions, Power Management and Automatic Configuration into one. If you purchase a quality computer you will have a quality BIOS with a quality ACPI implementation. ”
Utter, unadulterated misinformation. WHQL certification, or the forcing of manufacturers to turn the MS plug and pray model into something useable, has been what has made driver installation on Windows even remotely bearable. Prepackaged machines don’t count. The Windows driver installation systems is by far the biggest, most broken system of all time. It works because people spend megahours making it work. Any system that requires the installation of the device before the installation of the driver is _inherently_broken_. Some driver writers violate this spec for good out of the box installation, but at some point in the machine’s life, this model will quit working, mysteriously (no programs changed, etc.).
On Mac, you drop the driver on the system, and the system decides to load it or not based on device matching data in a XML list. Within this model, the user can plug and unplug the device anytime he likes. The system takes care of starting and stopping the module that it is targeted for. In the MS model, unplugging and replugging a device can be frought with danger. Even WHQL’d drivers are sometimes not reloaded properly, and Windows requests that you reinsert the driver disk. When I was reliant on Windows laptops, I always toted a stack of CDs just so I could be prepared when this happened.
” I’ve had this copy of Windows 2000 running for 4 months, and again, I have not experienced any of this BS “slow down”. Defrag your HDD and don’t install crappy substandard software and you won’t have problems”
You speak truth, and reveal the inherent weakness of the MS application installation model. On Macs, we don’t have this issue. Some more XML trickery and a well thought out executeable packaging model, and I can install any piece of junk software at any time and not suffer from it. In the MacOSX model, the app and the libs are one (even drivers can be loaded from the app package at runtime, and those drivers can be in the app package itself) as a single object, but also separate, in that the system will use already loaded or system libs if they are good enough, and packaged libs if the systems ones are not good enough. A similar mechanism exists on Windows, but it is not nearly as oppurtunistic nor refined, and even worse, no one uses it because its registry based, no obvious, and in general, looks like some idea MS came up with in 2 days to workaround some issue in their own shipping apps.
So please come again — I don’t understand why my system should have to remain static to remain stable. My MacOSX systems have no problems. It’s not the case with my windows boxes (I still keep a coupla PCs and a PC laptop around for hotmail, etc.).
ChipSlush
Rather, if anyone used the word, “killed” it was used in the context that the G4 would significantly outperform the PIII…
Yep, it’s in that way I use “killed”.
Make no mistake about it though, the reason why the G4 didn’t kill (outperform) the Pentium was NOT because of the PPC’s supposedly inferior design as many have suggested..
Don’t worry, I know that.
As long as Intel continues using x86 as the foundation for its technology, it is highly unlikely. Intel’s other chip technologies which aren’t x86-based, leave much to be desired.
Well, the Itanium2 isn’t that bad. The first one sucked though, and they still have a lot of work to do with the second one. I don’t think the PPC was a killer chip at its beginning. It evolved to become one.
So, unless market dynamics change significantly in the way of IBM’s relationship with Apple or IBM experiences massive financial problems, it is unlikely.
I guess it depend also on how many PPCs they sell. I don’t know if they’ll continue to develop them en masse if they can’t sell enough.
No, we say them in action and the results shown by several different application developers, a Nasa scientist and an independent research company all agree that the G5 is in fact faster.
I’m talking of the PPC chips that ain’t out yet, i.e. the 980 and 990. You can’t predict their performance even if they look good on paper.
“practically” speaking, the chip is:
fully 32bit and 64bit native (no emulation) = Practical
Very cost effective = practical
Very scalable = practical
The list goes on.
lack of practicality is not a factor here.
Again, I was talking of the PPC 980 & 990. Sorry if I wasn’t clear enough.
What? The Mach kernel was never killed, nor is it “too hard”
I wasn’t sure.
The approach was to sell people on MHz and GHz rather than actual speed. Surprisingly, it worked.
And why are you surprised? I’m not surprised at all. They’ve sold computers based on the clock frequency of their CPU for more than 20 years. Why do you think AMD bring our their “Performance Rating”? It’s not really different from what Cyrix(?) did some years ago.
Efficient is a relative term that can only be truly gauge in relation to its competitors. For example, instructions optimized for the G4 with AltiVec were and have always been more efficient than a a comparable x86 chip.
It’s not because AltiVec is more efficient that SSE/SSE2 that they ain’t efficient. I only said that SSE/SSE2 makes the P4 faster/more efficient than without them.
I agree, but the speed increase that will be had from these chips is not relative to the MHz that one might have experienced with previous versions of Intel’s chips simply because x86 doesn’t have the same potential for growth as does PPC.
Well, I’m not so surprised. The PPC is a quite new design (1993 I believe). The x86 design is so old that Noah carried some CPUs based on it in his Ark. If it wasn’t for backward compatibility, Intel would have dumped the x86 design and all the legacy architecture (VGA, COM ports, ISA… Yes, ISA is still in the architecture!).
MHz doesn’t matter when a computer’s speed doesn’t run in parallel with its MHz. For example… unless a PIV @ 4GHz is 2x the speed of a PIV @ 2GHz then MHz doesn’t matter.
However, MHz DOES matter if the rate of speed moves in parallel with the MHz. Then, it would be a much closer gauge for the average consumer to determine a chip’s overall speed.
IBM has said that the Power5 is 400% faster than the Power4.What this means is that where Intel increased MHz without speed, IBM will be increasing MHz with speed. This means that the Power series of chips allow for MHz/GHz to be a much more accurate way to gauge a processors than is the case with x86 chips.
I guess you didn’t understand me. MHz doesn’t matter at all to me because you can’t compare architectures directly. The G5 2GHz seems to be faster than a P4 3GHz even if the latter have a 50% higher clock frequency.
Actually, on x86 chips there is a wall at which increasing the Pipeline stages to increase MHz will actually DECREASE the speed of the chip. This is the point where Intel is at now.
I know, and it’s a feature. They wanted it that way. They claim they’ll be able to extend the clock speed (and the performance) up to 10GHz. However, that’s a claim, like IBM claims that the Power5 will be 400% faster.
This is very true, and yet the 68030, 601, 603, 604, G3 and G4 were all incorrectly judged in this way.
AFAIK, Apple was never really good with marketing either.
The ironic things is that unlike unlike the G4 and P4, the G5 and all subsequent revisions to that chip architecture will increase in MHz and Speed. This means that a 4GHz G5 will not only be double the speed of a 2GHz G5, IBM’s roadmap suggests that it such a chip would actually be 16x as fast! (Remember, the IBm roadmap tells us that future versions of the g5 (980?) will be 400% faster at the same MHz. When the MHz is doubled at 400 percent then you get a 1600% speed increase!
Aye, but those are claims… We will see what will really happens when it’ll be out.
Hmm, unless I’m affected by alzeimer, I don’t remember arguing with you for that.
hey guys. I have an idea. Go start your own foundry and tell me if you have any teething problems. After you try that if you don’t then come back and you can spread your trolling without any comments. BUT, spare us until you do.
Give them a chance. Its a brand new fab and its IBM, the masters of execution. REmember remember remember. IBM actually holds onto its experienced and older employees, something that Intel is not so good at.
>>>”As long as Intel continues using x86 as the foundation for its technology, it is highly unlikely. Intel’s other chip technologies which aren’t x86-based, leave much to be desired.”<<<
“Well, the Itanium2 isn’t that bad.”
And yet its not all that good either. Not a good foundation for competing against a juggernaut like IBM and the G5
“The first one sucked though, and they still have a lot of work to do with the second one.”
I think the foundation itself is flawed from the beginning.
“I don’t think the PPC was a killer chip at its beginning. It evolved to become one.”
Naturally, the PPC would evolve as it grew older as i would expect any chip to do, but what we’re seeing now with the G5 is really closer to what was anticipated with the G4… which was the “glory-days” intended direction that was envisioned from the start.
“I guess it depend also on how many PPCs they sell.”
Not necessarily… Unlike the G4 and P4 which is required to be at least self-sustaining for it to be a profitable venture, The G5 is a derivative of the Power-series of chips which means that IBM would be making them anyways. It’s important to remember that the G5 is a scaled down version of the Power4/5… a chip venture which IBM already has a worthy market anyways. Apple’s use of the 970 is only gravy on top of an already successful chip architecture.
“I don’t know if they’ll continue to develop them en masse if they can’t sell enough.”
That’s the beauty of it. While Apple seems poised to sell them ‘en mass anyways, they don’t have to because the chip is scaled down from an already successful chip architecture.
“I’m talking of the PPC chips that ain’t out yet, i.e. the 980 and 990. You can’t predict their performance even if they look good on paper.”
IBM has said that they already have working prototypes of the 980… or at least the Power5 to which they’ve boasted a 400% speed increase at the same MHz of the Power4. This isn’t something that’s just on paper… we’re talking about something that’s way beyond that point… something that we could theoretically see shipping in as little as 6 months… although more likely closer to 12 months (according to IBM’s roadmap).
“The approach was to sell people on MHz and GHz rather than actual speed. Surprisingly, it worked.
And why are you surprised?”
I’m surprise because consumers were duped into buying something that they were led to believe was twice as fast when in fact it wasn’t. In retrospect, I probably shouldn’t have been surprised considering how gullible the computing populace really is.
“I’m not surprised at all. They’ve sold computers based on the clock frequency of their CPU for more than 20 years.”
All the while misleading the public to believe that MHz and GHz were a way to gauge performance… I.E. a 3GHz processor being supposedly 2x as fast as a 1.5 GHz processor.
“Why do you think AMD bring our their “Performance Rating”?”
Because AMD (whose chips also were not an accurate gauge to measure performance relative to MHz and GHz they were a bit closer than Intel was. This was certainly something to boast about.)
“It’s not because AltiVec is more efficient that SSE/SSE2 that they ain’t efficient. I only said that SSE/SSE2 makes the P4 faster/more efficient than without them.”
You are right about that however efficient is a relative term considering the fact that a G4 G3 604 etc was far more efficient speed wise than the intel chip it has been compared to throughout the years.
“Well, I’m not so surprised. The PPC is a quite new design (1993 I believe). The x86 design is so old that Noah carried some CPUs based on it in his Ark. If it wasn’t for backward compatibility, Intel would have dumped the x86 design and all the legacy architecture (VGA, COM ports, ISA… Yes, ISA is still in the architecture!).”
I agree totally. Unfortunately it is this need to retain compatibly with x86 which will continue to hamper the architecture.
“MHz doesn’t matter at all to me because you can’t compare architectures directly.”
Agreed, but the only reason why you can’t compare architectures directly is because neither the G4 nor P4 (as well as all of their previous incarnations) had speed that was relative to its MHz/GHz.
What that meant was that although the P4’s clock speed was faster, the G4 (for the majority of its life) was either faster or at the same speed. (The chip fell behind for the latter part of its lifetime in comparisons against the P4)
The G5 and all its subsequent generations will be the only chip that will have a MHz/GHz rating that is relative to its speed increase so therefore you will be able to make more accurate cross platform comparisons to show how much faster the chip is. It wont be about MHz/GHz or speed as it is/has been the case when comparing the G4 and P4 (which can’t be down). Instead it will be about the G5’s speed AND MHz/GHz.
So now we can say, look at this 4 GHz P4 (which will probably be only 20% faster than a a 3GHz P4), the MHz are the same. Does that mean that are the same with regard to speed? Let’s check. Wow, the G5 is 700% faster!
“The G5 2GHz seems to be faster than a P4 3GHz even if the latter have a 50% higher clock frequency.”
The G4 was this way to for the majority of its lifetime.
>>>“Actually, on x86 chips there is a wall at which increasing the Pipeline stages to increase MHz will actually DECREASE the speed of the chip. This is the point where Intel is at now.”<<<
“I know, and it’s a feature. They wanted it that way. They claim they’ll be able to extend the clock speed (and the performance) up to 10GHz. However, that’s a claim, like IBM claims that the Power5 will be 400% faster.”
Here’s the beauty of that comparison… and the one that I’ve been trying to communicate all this time. Because the Pentium series of chips are against a wall where increasing MHz/GHz actually decreases speed, they wont be able to go any faster. As a matter of fact, they will go SLOWER.
On the other hand, future generations of the G5 will not only get faster in MHz/GHz but they will be 400% faster than the current G5s at the same MHz.
This means that:
IBM will have its MHz/GHz go up while having its actual speed increase rise in multiples of four.
Intel willhave its MHz/GHz go up while having its actual speed DECREASE in multiples of ?
“AFAIK, Apple was never really good with marketing either.”
For some things they were, for others they were not.
>>>“…When the MHz is doubled at 400 percent then you get a 1600% speed increase!”<<<
“Aye, but those are claims… We will see what will really happens when it’ll be out.”
They are claims made by IBM for sample chips that they already have in existence… not just something on paper.
“Hmm, unless I’m affected by alzeimer, I don’t remember arguing with you for that.
“
I was initially responding to this:
http://www.osnews.com/comment.php?news_id=4071&offset=60&rows=75#12…
Utter, unadulterated misinformation. WHQL certification, or the forcing of manufacturers to turn the MS plug and pray model into something useable, has been what has made driver installation on Windows even remotely bearable. Prepackaged machines don’t count……..
If you buy quality hardware who create quality drivers you won’t have any problems. The problems occur when one has to resort to running drivers from previous versions on the latest version of Windows. That is when issues start occuring.
Also, the WDM model is very sound, it is unfortuate however we still have idiots unwilling to properly change their WHOLE driver over to it. You are wondering why that sort of thing happens? because we have charlie cheapskate management who demand that the piece of software be buzzword compliant NOW rather than testing it properly and ensuring it actually works.
Regarding WHQL, you know VERY little about. Lets remove Microsoft and replace it with SUN. SUN has a very similar scheme in which you yourself use software which simulate common hardware faults and issues so that the driver can compensate for these issues. Lets now put Microsoft back into the equation, WHQL is a stress testing assurance programme which enables the end users to know that if they purchase hardware and the drivers are using that logo, they are certified to have been stress tested with the version of Windows labelled on the box.
” I’ve had this copy of Windows 2000 running for 4 months, and again, I have not experienced any of this BS “slow down”. Defrag your HDD and don’t install crappy substandard software and you won’t have problems”
You speak truth, and reveal the inherent weakness of the MS application installation model. On Macs, we don’t have this issue. Some more XML trickery and a well thought out executeable packaging model, and I can install any piece of junk software at any time and not suffer from it. In the MacOSX model, the app and the libs are one (even drivers can be loaded from the app package at runtime, and those drivers can be in the app package itself) as a single object, but also separate, in that the system will use already loaded or system libs if they are good enough, and packaged libs if the systems ones are not good enough. A similar mechanism exists on Windows, but it is not nearly as oppurtunistic nor refined, and even worse, no one uses it because its registry based, no obvious, and in general, looks like some idea MS came up with in 2 days to workaround some issue in their own shipping apps.
So please come again — I don’t understand why my system should have to remain static to remain stable. My MacOSX systems have no problems. It’s not the case with my windows boxes (I still keep a coupla PCs and a PC laptop around for hotmail, etc.).
The package model is nice, and Microsoft produce MSI which is a very good piece of work, too bad we have lazy companies unwilling to actually USE the software out there developed and bundled with Microsofts development package.
I am not boasting or saying that MacOS X sucks and Windows works, all I am saying is that those who say that one or the other are superior lack any experience on either platform. I have used both and both operating systems are very good quality, and each have their quirks. There is no use holding your hands over you ears when you hear something you don’t like. The sooner you realise that your operating system, whether it be Linux, Windows or MacOS, ISN’T perfect, the sooner people will actually view IT for what it is, a tool to get work done and not some religious icon.
apple certainly would begin to compete and probably kick some serious ass. if g5 was available through other vendors and apple became an os and software vendor.
Actually, I know quite a bit of the pain of WHQL. It is a stress test, but to be stressed, it must load. Doesn’t mean anything to WHQL with no driver loaded, and there are test cases for plug and play (such as modems that plugin to the USB ports, etc.). For PCI cards, sure there may not be nearly as much stressing as the Plug part of the equation, but drivers for firewire and USB devices must survive extreme stressing of plug and unplug. And yet, this still doesn’t reveal real flaws that occur in the real world because of the overkill and complexity of the MS matching model.
And I do buy quality hardware — 3com PCMCIA cards and a thinkpad laptop. But I had this issue occur more than once… and just once is a bug.
I also disagree that the application support layer is properly designed on Windows. If the layer was properly designed, application developers would use it, but in fact, they don’t. If I remember correctly, I don’t think MS even uses it for many of their apps.
I will admit that XP is decent, but having the forthought of good engineering throughout every component, like MacOSX, it is not even close. Of course, after what, 7(?) years of development, NT had better be OK in terms of reliability. Thats alot of time for making even the buggiest, most badly designed piece of software, reliable.
I also disagree that the application support layer is properly designed on Windows. If the layer was properly designed, application developers would use it, but in fact, they don’t. If I remember correctly, I don’t think MS even uses it for many of their apps.
But these developers don’t use them. Just look at the number of half-assed 16bit to 32bit ports that happened the early days by two bit operations claiming that they’re the first to be 32bit.
Just look at the number of developers who are completely clueless about Windows technologies and continue to use API calls that have been CLEARLY labelled as depreciated. These developers are the idiots, not Microsoft.
What is Microsoft meant to do? send around an extortion squad and hold these developers at gun point to force them to follow the correct proceedures?
These proceedures are written and available to all and sundry. The people who don’t follow them up are the developers. Good developers will self educate themselves about the platform they are working on and follow the advice of the software company, stupid developers will continue to use bad techniques and when Microsoft drop support for something, such as an old API call been labelled depreciated for 5 years, we have this idiot of a developer going, “Ooooooooooooooooh, I didn’t know that!”. Well, you didn’t bloody know that because you were too worried about coding and less about clueing on what is happening around you.
This will get moderated down, but I’ll have fun posting it
And Eugenia Loli-Queru will not be able to curse me in my email either, like last time when I made her angry, she will just have to take it out on us the readers, or worse…
But here is reality:
Eugenia Loli-Queru is still ticked off that her pet OS, BeOS, didn’t get picked by Apple. Macs could have 12GHz processors, crack all the primes in the 2048 bit space during processor idle time, and operate by thought. Eugenia Loli-Queru would run stories about how Windows XP or BeOS would still be better. And the fact that this post will be modded down, and the fact she has personally cursed me via email before, is pretty much proof of what she is about (I won’t imply anything here, fill in your own adjective or noun
). She continues to promote the anti-Mac attitude among her peers at OSNEWS, or she hires or recruits like minded individuals. What can one expect from one whose heart was taken by a Frenchman, and crushed by a man who possesses The Reality Distortion Field? I leave this as a thought excercise.
The only thing I don’t like about moderation? It doesn’t seem to work on Eugenia Loli-Queru, maybe its because that is a stress too oft supplied, or just doesn’t work at the editorial level.
Also, I simply apple-C’s Eugenia Loli-Queru’s name in at every point I used it, so the spelling is just as correct as the website itself spells it (I was cursed for that too).
In addition, if I am to be moderated, please moderate this story. Your terms and conditions note OS-only discussions. OSX is pretty much processor agnostic. What G5 yields from a plant that isn’t quite a year old, and its relationship to any properly designed OS. I don’t see the connectivity.
One more thing: I realize that Eugenia Loli-Queru didn’t write this article, but a carefully selected person who thinks pretty much like her did.
Hi, my name is <deleted due to the Eugenia Loli-Queru protection program> and I have:
2 x86 PCs running Windows XP
1 x86 PC running Solaris 9
6 G4 towers running OSX
1 17″ Powerbook
1 12″ iBook
l8r
And yet its not all that good either. Not a good foundation for competing against a juggernaut like IBM and the G5
Intel is the largest chip manufacturer in the world and commands arguably the most advanced fab plants in the world.
I can purchase and have delivered right now both hardware and software that are fully 64-bit running on the Itanium2. Not next month or down the road. RIGHT NOW. Apple also dosen’t plan to offer a fully 64-bit OS on the G5. I can get one right now on the Itanium.
Apple’s use of the 970 is only gravy on top of an already successful chip architecture.
The G5 is GRAVY on top of another chip ? lol. Why is IBM complaining about a plant that makes these processors then ? Its just gravy right ?
That’s the beauty of it. While Apple seems poised to sell them ‘en mass anyways, they don’t have to because the chip is scaled down from an already successful chip architecture.
That dosen’t mean it was cheap to design the processor.
Because the Pentium series of chips are against a wall where increasing MHz/GHz actually decreases speed, they wont be able to go any faster. As a matter of fact, they will go SLOWER.
Intel willhave its MHz/GHz go up while having its actual speed DECREASE in multiples of ?
Intel has it made if the competition is as dim as you are.
Lets remove Microsoft and replace it with SUN. SUN has a very similar scheme in which you yourself use software which simulate common hardware faults and issues so that the driver can compensate for these issues.
I think the main advantages of Solaris are things like the scrubber which allow you to identify what bad hardware components caused a specific kernel panic and so forth.
“Logic Rocket v2.0”, I totally agree with you.
Eugenia Loli-Queru always picks news articles that are against Linux, against SUN, against IBM.
“apple certainly would begin to compete and probably kick some serious ass. if g5 was available through other vendors and apple became an os and software vendor.”
You mean, give up the lions share of sales to what would otherwise be Apple’s competitors and at the same time turn the organized configured parts that make up the Mac to something similar to the unorganized conglomeration of mismatched parts that resemples the PC?
“Intel is the largest chip manufacturer in the world and commands arguably the most advanced fab plants in the world.”
And yet it produced a less than optimal chip (Itanium/Itanium2) which is mean to compete with IBM. What I was getting at is that despite all Intel’s power, they created a mediocre high end chip. When you’re competing against a company like IBM, that’s not a good idea.
“I can purchase and have delivered right now both hardware and software that are fully 64-bit running on the Itanium2.”
And yet its 32bit applications will run far less than perfectly.
“Not next month or down the road. RIGHT NOW. Apple also dosen’t plan to offer a fully 64-bit OS on the G5. I can get one right now on the Itanium.”
That’s not true. Apple DOES plan to offer a fully 64-bit OS on the G5. That comment is pure FUD.
First of all, the only evidence that we have to suggest that we wont is a report published by TheRegister.
Second, we KNOW that the first OS (code named smegal) is a placeholder OS meant to ship with the G5 until Panther which will be shipped in November and is said to be fully 64 bit. The Register suggested that Panther won’t be 64 bit when what it probably meant was that smegal.
Third, even if The register is right, it only means that the Finder won’t be 64 bit. Every other aspect of the OS will be 64 bit, including the ability to run 64 bit applications.
What does this mean? It means that this is a non-issue.
“The G5 is GRAVY on top of another chip ? lol. Why is IBM complaining about a plant that makes these processors then ? Its just gravy right ?”
They aren’t complaining about the chips… they’re really not complaining at all. What they said, and what this story strongly misrepresents is that IBM just made a 3 billion dollar wafer fabrication plant at that they haven’t yet seen a return on that investment yet. That’s to be expected considering the fact that the chip hasn’t even gone into shipping computers yet and thus had a chance to even start getting a return on investment.
What I was saying is that the Poer4/5 is already a profitable chip and that the G5 doesn’t even have to be self sustaining to make it a profitable chip venture for IBM because it is a scalled back version of an already-sucessful chip.
>>>“That’s the beauty of it. While Apple seems poised to sell them ‘en mass anyways, they don’t have to because the chip is scaled down from an already successful chip architecture.”<<<
“That dosen’t mean it was cheap to design the processor.”
Again, the processor was already designed. The only engineering involved was in scaling back the chip to accomidate a consumer level processor as well as add technologies like hypertransport and AltiVec. (Trivial expenses for a chip of this calibar)
An editors job is to print stories that appeal to the pubs target audience. The apple, sun, IBM stories do that.
The simple fact is that the Itanium2 rapes any other 64-bit general purpose processor; the 1.5 GHz 6 MiB cache parts walk all over anything IBM has to offer.
PPC is playing catch-up.
“The simple fact is that the Itanium2 rapes any other 64-bit general purpose processor; the 1.5 GHz 6 MiB cache parts walk all over anything IBM has to offer.”
You got stats to back that up?
And yet it produced a less than optimal chip (Itanium/Itanium2) which is mean to compete with IBM. What I was getting at is that despite all Intel’s power, they created a mediocre high end chip. When you’re competing against a company like IBM, that’s not a good idea.
How can you say that it’s a mediocre high end chip? I can understand your point if you only base your fact on one specific benchmark where a 32-way Power4 beat a 64-way Itanium2. However, I’ve checked other results, and the 32-way Itanium2 1.5GHz is head-to-head to the 32-way Power4 1.7GHz. It’s even quite faster in SPEC (but I must admit that SPEC is far from being representative of real world performance).
And yet its 32bit applications will run far less than perfectly.
That’s because the 32-bit applications uses a completely different architecture. IA64 is very different from x86. 32-bit PPC is not really different from 64-bit PPC. IBM don’t have to deal with backward compatibility for an architecture that was made before the existence of dinosaurs. I must say that Intel could have done their emulation better (they’re the designers of the x86 set, after all), but I don’t think they did a bad job. I don’t think most Itanium buyers plan to run 32-bit applications on it, anyway.
“An editors job is to print stories that appeal to the pubs target audience. The apple, sun, IBM stories do that.”
Lets assume that a company gets 90% negative media attention and 10% positive media attention from the general technology news scene, all of it warrented and justified.
Now lets assume that a company gets 10% negative media attention and 90% positive media attention from the general technology news scene, all of it warrented and justified.
Eugenia has said that she covers each company equally, posting both negative and positive news for each company equally.
If she posts 50% positive news and 50% negative news about a company that is getting 90% positive and 10% negative (all of it warrented and justified), she is doing the technology community a dis-service.
On that same token, If she posts 50% positive news and 50% negative news about a company that is getting 10% positive and 90% negative (all of it warrented and justified), she (again) is doing the technology community a dis-service.
Eugenia DOES give more positive news coverage to companies who’s business model don’t deserve it and DOES give negative media coverage to companies who’s business model don’t deserve it.
While I would agree that it is an editors job is to print stories that appeal to the pubs target audience. The current string of apple, sun, IBM stories are building an audience of trolls and flame starters rather than addressing the already existant wide readers base.
Being a Mac user I’m more sensitive to the unbalance of Apple/Mac centric news articles that are posted to this site. For example, every day, I surf through a long list of news articles from MacSurfer, it seems that despite the endless list of pro news Apple/Mac pieces that are coming out of Apple (Mac Surfer collects Apple centric news be it good or bad), whenever I find something that casts Apple in a negative light, or could be misconstrued as casting Apple in a negative light, or simply something that just feeds the trolls on this forum, you can bet that it will appear on OS News.
I mean, I am able to summarize OS News’s Apple news articles just by collecting whatever negative press is being said about Apple on any given day.
To OS News’s credit, they don’t always post negative Apple-centric articles such as these, but then again some days there is just simply no negative Apple news floating around the web. However, whenever there is, you can bank on them appearing here.
Yes, thanks, I do have stats to back that up.