Sinatech: ATI/AMD Aquisition Agreement Reached: ATI facing a big shake-up

Status
Not open for further replies.
One thing I find interesting is that we seem to be reaching a feature saturation point in 3D graphics. ATI went two generations on the R300 architecture, NVidia went two with NV4x, and I could easily see 3 or 4 with DX10 hardware. Software has a lot of room for improvement with current hardware, let alone the future.

Could longer GPU design cycles be more feasible? The insanely parallel nature of graphics makes it quite amenable to copy and paste for the vast majority of the die. Get one scalable unit right, and you're good for a long time. I think 2-3 years could be enough to tweak the logic for a shader array to run at 3 GHz+, especially with full access to a fab of AMD's calibre.

The possibilities...
 
Mintmaster said:
I think 2-3 years could be enough to tweak the logic for a shader array to run at 3 GHz+, especially with full access to a fab of AMD's calibre.
I believe such an ALU would be idle quite often..;)
 
_xxx_ said:
Why scary?

Nobody wants to see an already small market become even more small or in any way bias. I'm sure a key worry would be not only a halt of CrossFire support and Chipset support for Intel CPUs under ATI but also that AMD may want ATI to make their cards somehow require AMD parts to function far down the line. I personally dont want to see ATI synonymous with primarily AMD based computers nor do i want to see Nvidia reap the benefits as the only performance graphics option for either Intel/AMD computers. You know they (AMD) have something like this in mind for the distant future if AMD feels so strongly that they should completely take over ATI. Buying a design or hiring on ATI to aid them in making central processing graphics processors a reality is one thing, but the complete take over of the company is entirely different. In that case its even possible that AMD would choose to dissolve ATIs part in the enthusiast and mainstream graphics market all together in order to have the entire company concentrating on what they want. Now wouldnt that suck? It wouldnt be the end of the world if it happened but i personally dont want my options cut down to literally 1 choice. Not to mention AMD is going to be right back in the stone age as far as OEM deals if their 2008 Core design fails against Intels new Core (new intel core architecture every 2 years) which will be harmful to ATI since they would be linked as part of the same company. Intel vs AMD or Intel vs AMD/ATI, i'd bet my money on Intel. Investors and OEMs problably feel the same way.
 
Last edited by a moderator:
Jawed said:
Chartered is producing A64:

http://www.amdboard.com/chartered.html

and is obviously one of the foundaries able to produce GPUs for ATI/NVidia, though it is only producing 130nm stuff for ATI (Imageon) as far as I can tell.

Jawed

Nvidia CEO Huang Jen-hsun has described TSMC as its closest production partner, with the foundry making 90% of its products. The other 10% are manufactured by UMC and Chartered Semiconductor.
source: http://www.digitimes.com/mobos/a20060630A1002.html

I think they are already manufacturing something for Nvidia, but i can't tell what it is for sure.
 
SugarCoat: noone's that stupid to directly fight Intel that way, AMD is still far away from that place financially. That would essentially be a suicide.

The takeover might still leave ATI as an independently operating unit but just sharing resources/foundries. You know how desperate AMD is to get out a good chipset of their own, this would be the opportunity to make it. There are surely lots of things about power saving, bus systems, memory interfaces etc. to be shared to benefit both. It doesn't have to be a close-minded "specializing", could as well be a strategic move for purely financial/IP reasons.

And as said, that would leave more place for the new competitors to catch up, since if nV ruled the market, they'd surely tune down the pace of development to increase margins.
 
Well, no one of repute has stepped in to add fuel to the Chinese fire today. Big deals often get announced after stock market hours to let people digest it without roiling the market. Friday after hours is an even better choice for that reason.

So if we get past tomorrow afternoon without a deal announced or another significant source suggesting there's a "there" there, ATI might have survived another round of rumor-monging.
 
geo said:
Well, no one of repute has stepped in to add fuel to the Chinese fire today. Big deals often get announced after stock market hours to let people digest it without roiling the market. Friday after hours is an even better choice for that reason.

So if we get past tomorrow afternoon without a deal announced or another significant source suggesting there's a "there" there, ATI might have survived another round of rumor-monging.

something I cant say explicitly.:p
 
_xxx_ said:
SugarCoat: noone's that stupid to directly fight Intel that way, AMD is still far away from that place financially. That would essentially be a suicide.

The takeover might still leave ATI as an independently operating unit but just sharing resources/foundries. You know how desperate AMD is to get out a good chipset of their own, this would be the opportunity to make it. There are surely lots of things about power saving, bus systems, memory interfaces etc. to be shared to benefit both. It doesn't have to be a close-minded "specializing", could as well be a strategic move for purely financial/IP reasons.

And as said, that would leave more place for the new competitors to catch up, since if nV ruled the market, they'd surely tune down the pace of development to increase margins.

You say no one is that stupid to directly fight Intel directly, yet thats exactly what i would call it if AMD were to take out one of Intels key enthusiast chipset partners. Currently the bulk of ATIs concentration is not on Chipsets and of course isnt on making an Intel or AMD cpu contain some sort of 3D function. However if AMD were to aquire ATI, it would be for those reasons and could invariably switch the primary concerns of the company as a whole to making chipsets and low end graphics options for AMD. Why would AMD care about the relatively small amounts of money made from flagship or exotic card setups if they assimilated (we are the borg!) ATI? The only people that lose out are the consumers of medium to high end graphics options, yet AMD would gain their own chipset production and undoubtably a veteran team to work on solutions for use within AMD processors, things that AMD sees important. They wont care nearly as much about the Radeon700 core with DX10.2 and 96 shader processors, or what have you, when such products return relatively pitiful profits by comparison.

Point being we could see ATI drop out of the enthusiast graphics market a few years down the line if/when AMDs priorities switch under such a merger.

Personally i dont understand it. AMD shouldnt need to buy out ATI for their chipsets nor should they need such a high profile company to make discreet graphics within central processors. I question the move financially and if it wouldnt be safer and simpler for AMD to persue such a venture by taking over someone with less noteriety like Matrox.
 
Last edited by a moderator:
SugarCoat said:
Point being we could see ATI drop out of the enthusiast graphics market a few years down the line if/when AMDs priorities switch under such a merger.

Yeah, that's the nightmare. Certainly not for R600. Probably not even for R700. But I'd be really worried about R800 and later.

The thing is, if you're not competing at the top, then eventually it hurts you technologically in the middle and lower as well (aside from obvious market "bads" that come from not having healthy competition at the top end of the market).
 
I dont see any good to be honest. A binding partnership is a good thing. However what we're talking about sounds more like a hostile take over. Hostile in that AMDs priorities and goals for such a take over surely arent the same as what are currently ATIs goals and priorities. A merger really scares me because i just dont see it in AMDs best interest to keep ATIs R&D department spending millions of dollars and thousands of man hours researching enthusasit graphics products. AMDs board of directors would be law with Orton sitting to the side near the water bubbler and a leash around his neck.

If it turns out true i will look forward to hearing why ATI agreed. It might be good for the company but it leaves gamers and devs sort of out in the cold and in a constant state of worry in my opinion.
 
my guess is, acquisition/merger talks between AMD and ATI did happen, but probably broke down. it doesn't make sense for ATI. I'm hoping it didn't happen.
 
IF the market stays the same though, it's a plus for ATI, because merged with AMD the two become bigger with more resources to fight competitors, and that's what mergers are all about.

The wildcard though, is I guess then an Intel Nvidia merger might be in the offing, and that would be much stronger, mainly due to Intel's strength.

But that's a big assumption. It seems kinda odd that Nvidia and Intel would just merge, just because their competitors did, just like snapping fingers.

So basically if the ATI/AMD merger happened and the situation is AMD/ATI merged, Nvidia seperate, and Intel seperate, ATI/AMD is better off than before.

But damn, there are only four main players here, two in each market, that's bare competition as it is. I really hate to see it whittled down even further.

But still the plum here is Intel. Whichever IHV merges with them (if ever) is the one sitting pretty, because of their huge size.
 
I think it would be fun to see what happens. :) I don't believe it for a second but if you were going to speculate on it, it raises a lot of possibilities.

Let's see, ATI has in interesting memory controller, AMD is moving to DDR3, it will have a cpu port for co processors. Video card prices for the top of the line have been pretty constant at $500+. AMD has some interest in lower cost pcs. AMD may have more opportunity with DELL in desktops and DELL likes all in one solutions. Consoles and laptops, integration solutions needed. I can imagine a gpu connected directly to the cpu via the coprocessor link fed with DD3. Upgrade the video with just a chip not a complete board (with power restrictions on PCIE and buying expensive ram over and over again). Or you could have a smaller amount of fast ram on a coprocessor board and a lot higher bandwidth to main memory and the cpu. Downside, proprietary systems, not important with consoles or laptops however. Something has to give with the huge desktops with $$$ components, john doe isn't interested and you pretty much need the top of the line when it comes to video cards or you are just upgrading spec sheets that have no meaning for performance. I think a trend towards integration is a possible path the industry may follow, just don't see it happening any time soon.

So, there could be some synergy there, which would be interesting to see, total flop business wise I'm sure. :)
 
Orton brought this on himself with shitty gross margins in the GPU space, late products and overall poor execution. Orton has weakened ATI to the point that these rumors have legs. Nvidia gross margins are up over 40% and ATI's are finally back to just 30%...nuff said. Until ATI can improve gross margins in their core GPU segment, they are ripe for the picking.
 
NEW YORK, July 6 (Reuters) - Computer processor maker Advanced Micro Devices Inc. (AMD.N: Quote, Profile, Research) on Thursday slashed its forecast for sales in the second quarter, saying sales are expected to fall 9 percent from the first quarter.

It previously forecast that sales would be flat to slightly down from the prior quarter.

The company said it expected second-quarter sales of about $1.22 billion, up 52 percent from a year earlier, but below analysts' average forecast of $1.31 billion, according to Reuters Estimates.


Ouch...but not a complete meltdown.
 
Status
Not open for further replies.
Back
Top