No pain, no gain, AMD rules
Hector's ATI move was the greatest strategy moves since Caeser's conquest of Gaul or the D-day of the allies.
Intel will be so obsolete by June 2007. AMD will simply frag Intel in multimedia, graphics and microprocessors.
The idea is simple. For every AMD GPU sold with an Intel processor, the price will be $35 higher. All of Intel's IGP business will be junked.
I saw discussion on AMD's past failure to capture more market with Opteron. The blunder was due AMD's timidness when dealing with Intel.
Had AMD sued Intel on Opteron launch, Intel would have been killed already. But, now it's not too late. K10 will officially finish Intel off. DELL will be 100% AMD on K10.
posted by Sharikou, Ph. D. @ 4/11/2007 09:17:00 AM
21 Comments:
...For every AMD GPU sold with an Intel processor, the price will be $35 higher
then why would anyone buy an AMD GPU then?
it appears that you never thought this trough... lol.
"As yet unlaunched, its DX10 entry-level chip, R(V)610 is already set to become the fastest-selling graphics chip of all time."
"It could happen that AMD ships more than 100 million 65nm GPUs by the end of 2007"
Intel's failure as a GPU manufacturer has left the door wide open for AMD. Less than 1 year after the acquisition of ATI, AMD is positioned to dominate Intel once again.
This would be a great opportunity for AMD to get back at Intel for years of crippling the competition with their compilers.
When the chip detects that an Intel product is running, it can slow down everything by say... 30%!
I mean after all, bugs like this happen all the time right?
It's not deliberate, it's an oversight, due to Intel processors not being tested with those chips.
Penix
For perspective, in 2006 there were 312.55 Million GPU's shipped.
Intel's market share of the GPU market is about 37%, which would equate to about 116 Million GPU's.
Nvidia would be about 88 Million GPU's.
AMD's market share was about 23%, which would equate to about 72 Million GPU's.
Discrete GPU's totaled about 109 Million GPU's, or about 34% of the GPU market.
Now if we factor in three quarters of sales, AMD for the rest of 2006 would have sales of about 54 Million GPU's.
That is well short of the 100 Million your referring to, so unless AMD plans on gaining significant market share, that 100 Million number is pulled out of a hat.
All number pertaining to total GPU's can be found searching Google with Jon Peddie Research.
Also, not forgetting the Skype / Intel alliance, it would be fair for AMD to 'also' limit features and functionality, no?
Also, not forgetting the Skype / Intel alliance, it would be fair for AMD to 'also' limit features and functionality, no?
Why not? Suppose AMD GPU is 20% faster than Nvidia's. Then AMD can lower its GPU performance by 19% on Intel platforms.
enumae said...
That is well short of the 100 Million your referring to, so unless AMD plans on gaining significant market share, that 100 Million number is pulled out of a hat.
Yes, that is well under 100 million. That is also a 2006 figure. We are talking about 2007 now. Once released, the R610 will be the fastest selling chip ever. AMD is going to step right over nVidia and stomp on Intel.
It is obvious that you did not read the article in Sharikou's post. Please do so before you comment further.
Grabbing ATI was probably a good move, Sharikou, but Hector would have been alot smarter had he done the deal 3-6 months earlier with the stock at $40 rather than $24.
Penix
Apparently I have missed something that you and Sharikou got.
The article is about Dell buying AMD's new DX10 GPU's. While also making the claims that you quoted.
Where Sharikou gets his $35 higher figure from I don't know, maybe you could explain this.
Now back to my point to you, I was showing you that the 100 Million GPU's statement, which you quoted from the article, is not likely to be realistic.
Now, I don't speak Penix, so please explain to me how in 9 months time AMD is going to increase its market share by about 85% in order to sell 100 Million GPU's?
Also please explain how much you feel the GPU market will grow in 2007 vs 2006.
PS like I said the other day, AMD good, Intel bad, thats all you see.
Who wants to bet that by the end of Q3 NV will have the fastest discrete GPU on the market?
Too bad AMD can't even get manufacturers that use Turdion cpus to use AMD graphics in their systems.
It's funny to see all the AMD cpu computers that come with Nvidia graphics. I wonder what the likes of HP, Lenovo, Acer, Sony, Gateway, and Toshiba know that you don't know Doctor?
http://www.shopping.hp.com/webapp/shopping/computer_can_series.do?storeName=computer_store&category=notebooks&series_name=tx1000z_series&tab_switch=true&a1=Usage&v1=Travel%2Fmobility&tab=specs
Buying ATI wasn't a brilliant move; it was AMD's only move.
This continuing talk about Intel going bankrupt or becoming obsolete is pure nonsense.
This is reality. If tomorrow Intel lost 33% of its market for processors then it has cash on hand and it could sell off a FAB or two without missing a payroll. Remember that Intel could lose 1/3rd of its production and still be twice as big as AMD. This would simply cause the current volume share to shift from 75% Intel/25% AMD to 66% Intel/33% AMD.
It is not conceivable that this could drop lower even with catastrophic circumstances. The fact is that AMD simply does not have the capacity to supply more of the market and there is no other supplier.
Worst possible case scenario for Intel is the loss of about 9% of their current volume. Best possible case scenario for AMD is a gain of about 15% of their current volume.
In other words, it would be impossible for AMD to convert more than 5% with current market volumes which would shift to 70% Intel/30% AMD. With a severe contraction in the market (loss of total market volume) this could be 66% Intel/33% AMD. This is as bad as it could possibly get for Intel. This is a long way from bankruptcy.
The real effect from R600 is not that it will put Intel out of business but that combined with the DTX motherboard standard it will secure AMD's position in the mid to low end desktop market.
bubba
"Too bad AMD can't even get manufacturers that use Turdion cpus to use AMD graphics in their systems.
It's funny to see all the AMD cpu computers that come with Nvidia graphics. I wonder what the likes of HP, Lenovo, Acer, Sony, Gateway, and Toshiba know that you don't know Doctor?"
This is before the release of R600. ATI should gain some share on nVidia with the R600 series and strengthen this further with the R700 series which should be released perhaps early 2008. There is no doubt that ATI was losing ground to nVidia before the purchase but that trend should reverse. Essentially, you should see ATI becoming much stronger in integrated graphics while nVidia remains strong in discrete graphics. I expect ATI to remain #2 behind nVidia in discrete.
Can you imagine buying an Intel c2d machine in Best Buy with an AMD sticker on it.
Why not? Suppose AMD GPU is 20% faster than Nvidia's.
You're dreaming. The Geforce 8800 GTX is twice as fast as anything Ati has. When they finally stumble R600 out the door (how many times has that been delayed now?!) they have the 8800 Ultra waiting.
When DAAMIT finally does ship R600, I suspect they'll advise all the tech sites to use a Core 2 Duo to do their reviewing on. DAAMIT does want, after all, the fastest CPUs for it's review.
Can you imagine buying an Intel c2d machine in Best Buy with an AMD sticker on it.
Damn interesting point!
But before that ever happens, you can bet your pennies that Intel will prevent its system builders from doing so.
In my opinion it would be kind of hypocritical since for years they forced the Centrino brand down their face!
Then again, AMD would pressure those same system builders to show the AMD stickers.
Do we have a new war on hand? Who would the system builders listen to? Who would have more power of influence?
Grabbing ATI was probably a good move, Sharikou, but Hector would have been alot smarter had he done the deal 3-6 months earlier with the stock at $40 rather than $24.
It's kind of irrelevant! First AMD paid mostly cash. And as someone has mentioned before, ATI was probably NOT interested in a stock deal. Without a doubt AMD put this deal together with a clear objection of eventually issuing shares against the ATI debt. Which would explain all the clauses attached to raising any new capital.
The reason I said it was irrelevant is to the fact that if AMD issues convertible bonds, the money goes to pay the debt. When the bonds are converted, the share value should have recovered to the $30 or $40 level.
The only negative aspect of all this is that AMD must pay a premium to the bond-holders in order to get them to buy them, thus a slight stockholder dilution.
As far as I'm concerned and as an AMD stock holder, they can issue all the stock they wish so long as they wipe the debt and raise a bit of cash. Then give Intel a war they will never soon forget.
"If tomorrow Intel lost 33% of its market for processors then it has cash on hand and it could sell off a FAB or two without missing a payroll."
Scientia, I don't agree with this. Intel's processor business will be losing money if it lost 33% revenue. However, unlike AMD, Intel's operations have been totally based on the fact that its processor business makes enough money to fund R&D, marketing, and other money-losing projects. Of course Intel can sell fabs to meet payroll but that's clearly not a sustainable act.
Intel is not likely to lose 33% processor share in at least 10 or 20 quarters (or probably never). The company will do anything to make sure of it. However, this doesn't mean it's okay for Intel to lose 33% processor share. Quite the contrary, in fact.
scientia
"There is no doubt that ATI was losing ground to nVidia before the purchase but that trend should reverse"
Perhaps on low end and gamers market but NV has professional market entirely for itself. It has been generating huge profit from selling its Quadros and ATI has almost nothing. Main problem is it's lack of support and not so good drivers.
I wonder what will AMD have to put against this
Ho Ho:"I wonder what will AMD have to put against this"
Against what? The 14 iterations of p6 uArch whose performance increase is flattened out and presented without an axis mark?
Notice that Intel's calling its future multi-core processors as "many core." An implicit meaning of it is the cores are many, but they are independent with each other. I.e., they do not work together. I fail to see how great is such an approach.
Post a Comment
<< Home