In the beginning, everything was separate. Processors, Coprocessors, Video, Sound, Chipsets, Network, Input/Output ports were all on separate boards and cards. There was a profusion of manufacturers and variety was the norm of the day. Weitek, Number 9, Adlib and Hercules were the names that were popular amongst the enthusiasts.
Then slowly, convergence occurred. Processors merged with Coprocessors, Sound, Network and IO ports went on the motherboard. Chipsets and Video cards were next, giving birth to the infamous integrated chipsets.
This evolution did not go without pain. Hundreds of companies have disappeared in the wake of this shift even if they pioneered disruptive technology. Think about 3DFx, Hayes or Adlib.
Canadian-based ATI just joined them, after being swallowed by Chip maker AMD. Now I am uneasy with the whole big-name merger thing. The last time it happened, some people badly got burnt.
HP “merged” with Compaq in 2001 which, in turn purchased Tandem Computers and Digital Equipment Corporation a few years earlier and the entire acquisition process was marred by numerous management problems. This ended in Tandem, DEC and Compaq unique selling features and strength being almost entirely diluted in the larger HP Corporation.
Unfortunately, I fear that the same thing is going to happen to ATI, except that I believe that this has been prepared beforehand.
It might seem very much far-fetched but I am convinced that Nvidia, ATI and AMD have come to the conclusion that they MUST unite their forces together to be able to fight off Intel.
It is well-known that Intel has the largest percentage of the graphic units/modules/cards in the x86 market and it is only a matter of time before it set their sights on the lucrative discrete graphics card market and swallows the whole ecosystem.
The first step in that direction was achieved back in the late 90’s with the release of the i740 graphic chipset which was quite popular. Since then, Intel has regularly updated its core integrated graphic modules, but has yet to launch a powerful mid-range solution.
But there’s more to it. Intel has also licensed Power VR’s Imagination MBX technology and has stopped short of purchasing the firm behind Kyro, probably because of what authorities and competitors may say. The Licence agreement explicitly mentions that Intel will use it on its Integrated Circuits.
So ATI, AMD and Nvidia all know that Intel is going to come upon them and that they won’t be able to make it if they stand separate. However, if all three came together, they logically stand a better chance of competing with Intel.
However, ATI and Nvidia cannot merge together because regulators will say that they represent a monopoly which is not conducive to competition. But the same authorities would probably close their eyes on AMD buying ATI.
The next “obvious” step would be for AMD to wait for a few years, letting Nvidia getting away with crime and then just merge with it to form the ultimate green troll.
Does it make sense? I will leave it to you. I’ll just point you to some interesting observations. Five years ago, AMD and Nvidia formed a strategic alliance called SNAP (Strategic Nvidia AMD partnership), the result of which was the nForce series of chipset which buoyed AMD’s platform. SNAP has yet to snap, even after ATI was bought by AMD.
I also find Nvidia particularly composed and calm, especially after its biggest partner purchased its biggest competitor, which would probably have sent the share prices tumbling. After the initial shock, the share price has been climbing, showing to some extent that Nvidia is serene.
The turning point probably occurred when the move to multi core processing (MCP) became inevitable. MCP allows for some graphics features to be taken over by the processor, especially when combined with other innovations AMD is bringing onboard. To get the CPU/GPU duo to work as seamlessly as FPU and CPU did, is something that both AMD and Intel are looking forward to.
Well, let’s see for example if the competition in the graphics arena will be as hotly contested as it was in the last few years. I suspect that AMD will ask ATI to slow down the development cycle and allow Nvidia to save money and dedicate resources to segments where AMD and Nvidia can take on Intel together; I am thinking particularly of the lucrative server market where ATI is almost inexistent and of the mobile market where much remains to be done.
Looking back at what happened a few years back, I can see how Intel ***might have *** influenced HP to buy out Compaq and eliminate promising technologies like DEC’s Alpha, Tandem’s Himalaya and its own PA-RISC architecture to the profit of Intel’s Itanium. Did the same happen to ATI?