Apple ready to embrace AMD?

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
No Chad you still failed quite badly.

Intel only expects 27% of all their chips to be 32nm by years end. If they have 3 more 32nm fabs coming online before then, that just shows how little an advantage 32nm is.

Intel cannot supply the real consumer market with 32nm, it's only ludicrous Clarkdales and i7 980x's that get made on their 32nm.

AMD on the other hand, will start off giving us Llano - the first APU and first true Fusion product, followed soon after with Bulldozer. Being 'AMD faithful' is good. :)
 


Hey could you make it a little more obvious that you are nit picking at my posts to purposely provoke me? If you had any 'comprehension skills' then maybe you could have understood that i was looking at that dekasev (was that the username?) guys post when i posted that reply, and not that article which i wasn't aware of at the time. But regardless, good job. Anything else of value to post? :sarcastic:
 

Jenny,
Literally nothing you just said makes ANY SENSE in the context of the conversation.

The point which whooshed over your head isn't the number of 32nm fabs right now, it is the number of leading edge fabs at any given point.

By the end of this year, Intel will have FOUR 32nm fabs and AMD will have ZERO, by the end of next year, Intel will probably still have 4 or maybe Five 32nm fabs because that is all they need(and one 22nm fab) and AMD will have ONE 32nm fab up and running to near full capacity.

12 months after that, AMD will still only have ONE 32nm fab.

Did you not understand what I meant about AMD always being capacity constrained?

Everything you have posted since I mentioned it, points to you being clueless, again.
 
You are a little coky ther chad..I can see that. Not that it is a bad thing, I just noticed it.
Everyone can look cocky when they are arguing against Jenny, her sillyness makes her opponents look like superstars by comparison. :lol:
 


No chad, sadly it is you who truly doesn't get it.

Very few fabs are built from the ground up these days. Intel needed more fabs, thats why they built the new ones. They weren't going to build them on an older process were they?

GF on the other hand has more fabs overall, and can upgrade most of them to the most recent technology.

GLOBALFOUNDRIES currently has five 200 mm fabs and one 300 mm fab in Singapore, as well as one leading-edge 300 mm fab complex in Dresden, Germany. To meet the demands of a growing customer base, the company has an aggressive capacity build-out plan, including expansion of Fab 1 in Dresden and Fab 7 in Singapore, as well as construction of a new leading-edge 300mm facility in Saratoga County, New York. The New York facility, which will be renamed as Fab 8, is on track to begin ramping initial production in 2012.

If you just take those 300mm ones and upgrade them, AMD would have 4x the current 300mm capacity than they do now. Saratoga Cty also has room for another two fabs.


AMD only has Dresden right now, and that is about to be *doubled* in capacity which simply means AMD has double capacity. Singapore will follow soon, then NY.

So right now, AMD can only get 30,000 wafers a month. With the dresden expansion that will be 60,000. Singapore needs retooling and expansion, so that's 120,000 by the end of 2010. AMD can now make ~5x more cpu's monthly than they have ever been able to.

Saratoga adds another 30,000 with expansion for another 60,000.

And that is just 300mm. :)
 

Don't get distracted or try to distract others over the building of fabs vs upgrading existing ones, whilst there is a time to market difference, the cost of an upgrade is also very, very high and GF have not indicated that at any point soon, they will have two full fabs running at the latest process required to supply enough of the market to seriously challenge Intel's market supremacy.

That is the downside of being capacity constrained.

GF on the other hand has more fabs overall, and can upgrade most of them to the most recent technology.
Upgrading fabs are still very costly and GF have not committed to upgrading enough of their fabs to the latest processes to seriously challenge Intel for marketshare.

AMD only has Dresden right now, and that is about to be *doubled* in capacity which simply means AMD has double capacity. Singapore will follow soon, then NY.

And that is just 300mm.
The link from Dresden which talks about doubling the capacity makes it very clear that it is not a doubling of AMD's latest process technology for cutting edge CPU design, the doubling is largely occurring because of Dresden becoming a foundry for non-AMD customers and/or GPU's.

So let's get some things clear for you and possibly others.

Yes an existing fab is better than no fab at all, because you can convert a fab quicker than building one from scratch, however even when you convert, it still takes a long while and costs a fortune.

GF has not committed to investing in fabs to the extent where they can supply over 1/3 of the x86 market, let alone actually toppling Intel.
 
Nobody said it doesn't take a while and doesn't cost a lot. Neither is GF stupid enough to commit to investing in x86 just because AMD said they could sell more cpu's.

It's a gradual process, but it's gradually swinging towards AMD and away from intel.

I'll just get one thing clear for you and others who probably don't quite understand what this means.

AMD can scrape a profit with only 30k wafers a month - they are still making over $1bn in revenues. They won't hesistate to offer cpu's at 1/2 price, while still scraping a tiny profit.

Imagine the lowest end Thuban at $75, the highest at $150. AMD can still break even at those prices but can intel take the loss of mindshare?

There are factors at work here that you just couldn't comprehend Chad. Suffice it to say, AMD will be the only chip maker giving great prices for the forseeable future - and that includes Llano and Bulldozer.
 
Well who says they will have to supply a third of the x86 chips? You have to take into account that market share includes old computers too so it's impossible to get a third of the world to upgrade to the new chips.

Plus, if AMD can't produce chips fast enough I'm sure Intel will lend them a hand rofl (was kidding btw)
 

So how will they ever know when to invest big to try and take on Intel, or do they give up on ever doing that?

Imagine if they had bet the farm after the K8 came out? By the time that extra capacity was coming online, Conroe would either have been out, or close to it, and then AMD would have been facing massive fab underutilisation costs for years.

I'll just get one thing clear for you and others who probably don't quite understand what this means.

AMD can scrape a profit with only 30k wafers a month - they are still making over $1bn in revenues.
Where are you getting these figures from? It looks like fuzzy math Dubya would be proud of.

They won't hesistate to offer cpu's at 1/2 price, while still scraping a tiny profit.
They are making a tiny loss(once you remove the GF once off item) today with 1.5Billion in revenue, how are they going to make a profit with 33% less revenue and significantly lower asking prices? :pt1cable:

Imagine the lowest end Thuban at $75, the highest at $150. AMD can still break even at those prices
Actually I don't think they can, so your point doesn't make a lot of sense.

There are factors at work here that you just couldn't comprehend Chad.
The only thing here I can't comprehend is the inner workings of your twisted mind.

You are waving your hands around to distract people whilst making vague nonsense claims, and then pretending it is because you are all so advanced that no one can keep up with you.

Your problem is that you got caught out speaking bullsheet again, as you have serious trouble separating fantasy from reality.
 

The conversation has only been about marketshare of ongoing future sales.
 


TOTALLY AND COMPLETEY AGREE.

Only a complete and clueless moron could possibly think that anyone that prefers AMD would automatically somehow adopt and/or prefer Microsoft over Linux or FreeBSD.

But it is humorous seeing some posters getting desperate enough to try to make that kind of baseless claim.

Sadly Microsoft currently "invents" and adds things to their operating system that were already available in the early 90's in other operating systems such as OS/2.
 


I skimmed through most of the post's I admit but where did anybody say that if somebody prefers AMD they would adopt and/or prefer MS over linux FreeBSD. Anyways, regardless of what chip the user is running or prefers, that person will not likely adopt/prefer ms over a *nix system period. Which is what I believe you where getting at anyways.

I cant see anybody that isn't a complete moron adopt/prefer OSX/macs over the pc with Windows/*nix due to the fact that Apple is using AMD chips. YaY apple now has AMD inside, let me pay that 500 bucks apple tax. Giving that OSX does fall under *nix you know what I'm getting at. Only a fool would buy apple just because AMD provides the chips.


 


LOL - I know, just a bad day at work - got stuck doing hiring interviews on top of my employee performance evals.

Well at least we're hiring once again, so maybe another supervisor to help out and I can get down to a more reasonable # of employees to mind 😛.
 
I might as well add to the speculation here...

Apple would be stupid not to keep an eye on the market to see what is available. They are being smart and hedging their bets. Especially now with their dedication to OpenCl, it would make sense to keep their CPU and GPU options open.

It would not be unheard of if Apple were to roll out next month with machines running on AMD hardware. If you look at the OSX86 project, you will see that there have been many successful OS X installs on AMD hardware, and that is without official Apple sanction. On top of that, AMD architecture makes use of several open standards, and all their parts are fairly interchangeable as long as the part fits the socket, and sometimes there is even a socket adapter.
 


Fuddy's post contradicts what Intel was saying about achieving 32nm crossover sometime this summer (50% 32nm production). I suspect the truth lies somewhere between the two, but Fuddy's track record for accuracy is also not the most sterling one around :kaola:

BTW, one investment firm had a sell recommendation on AMD, saying the stock will drop to $5 by year's end. Could be based on the IDC report that global PC shipments went up by 24% in Q1, which makes one wonder why AMD's products failed to produce a profit, whereas Intel had a blowout quarter.
 
AMD's products produced a profit of $182m so I'm not sure where you got that idea from fazers. With gpu supplies still constrained, it's been a pretty good quarter considering.
 


Well if you consider the deconsolidation of the last piece of their fab interests, a "product" (which nobody else does), then I guess that's true 😛. However the stuff they sell to OEMs, such as CPUs & GPUs, earned them a negative 68M. According to Reuters:

Net income rose to $257 million, or 35 cents per share, in its fourth quarter ended March 27, after a loss of $416 million in the year-ago period.

The company's results also included a one-time non-cash gain of $325 million from AMD's stake in GlobalFoundries.

AMD also has low expectations for the current quarter, which means that for the first time we'll get to see exactly how profitable their product sales are, unobscured by the GF deal since AMD is out of pieces to sell.
 


It was about 8 posts down on the third page of posts that somebody attempted to claim that Microsoft was somehow a savior for AMD.

ALSO: Many people do not realize that the OSX is based on UNIX/BSD. When Apple switched from their proprietary operating system to using an open source BSD for their kernel even though they maintained a proprietary GUI they gained a lot of respect from people that prefer UNIX. Interestingly that demographic includes many developers that predominantly work on non-Apple UNIX boxes. Also interesting to note is that demographic generally does not include PC "enthusiasts" who predominantly seem to be anti-Apple.

EDIT: From what I remember the main reason they went with Intel Core2 chips when they switched from PowerPC was for reasons of power use and battery run time in their laptops. (And not for many of the other reasons that people often like to quote.) For Apple to consider switching to AMD is not surprising or shocking in any way. What would be shocking would be if they ruled out AMD chips merely because their adoption would disappoint a bunch of Intel fans that are already anti-Apple.
 



Yes but what about the $183m equity loss related to GF we didn't hear about?


AMD paid $114m to GF for fab underutilisation so they have room to play with yet, but still made record first quarter revenues. It's going to be another quarter before they are both truly seperate on the books.

In no way can it be considered a bad quarter for AMD - record revenues, almost $2 billion in cash up from $1.75bn and margins increased by 4%.
 


Thanks, I will go back and look for the post after I write this own.

I agree with just about everything you said there. I will add some of my own thoughts to it though. Most people do not even know what UNIX/BSD even is period. Most of the people that do know about them also know about OSX being built of them.

The PC enthusiasts is always going to be anti-Apple. So put it simply the pc person will simply not pay for a machine with little to no hardware options + the apple tax.

Yes Apple went with Intel for reasons of power use/battery life in the laptops. But they also went with the more efficient chip. The Intel performance per watt was greater.