2006: AMD Acquires ATI: Future Is Fusion!
Although the takeover of ATI - a decision that may arguably be called the decision of the decade - saved AMD during hard times, its acquisition completely collapsed AMD's own plans, ruined a lot of prospects for ATI, but caused massive ramifications across the industries.
By the end of 2005 it became clear that GPU-accelerated computing had massive potential for various needs (most importantly, for high-performance computing [HPC] where AMD's chips were popular) and high-definition visualizations will prevail over simplistic user interfaces of the first half of the 2000s. AMD realized that in ten years from then it chips may not scale in performance as easily and as quickly as graphics processors and therefore it will have to share the HPC space with special-purpose GPU-based accelerators. Besides, given the fact that it did not have its own chipsets and graphics processors, eventually it would not be able to compete against Intel when it comes to platforms as it was clear that the graphics core was heading under the hood of microprocessors. So in late 2005 AMD started to negotiate with ATI Technologies considering merger or acquisition.
There was a great rationale for AMD to buy ATI at the time. AMD was growing both in terms of CPU sales and market share, AMD had massive manufacturing capacity expansion plans for 2010 and forward, AMD needed to have own computing platform with chipsets and embedded graphics cores, AMD needed to enter new markets, most importantly, markets of GPUs and consumer electronics. Furthermore, AMD knew it would lose performance crown to Intel in late 2006 and clearly needed profitable sources of revenue.
ATI was in a different situation. The company has always said that it is tailored for product portfolio expansion. In mid-2006, 77% of ATI's revenues came from various PC segments (GPUs, chipsets), while about 23% were the earnings of consumer electronics divisions. While the competition between ATI and Nvidia intensified back then, ATI had bright prospects on the DTV and ultra-mobile graphics markets. There was a threat that Intel enters the market of discrete graphics adapters and that would hurt ATI's sales, but this was certainly not the primary reason why ATI decided to become a part of AMD. Perhaps, the prospects to form an extremely competitive and extremely large corporation inspired the move, perhaps there were more incentives, but the companies officially became one in late October, 2006.
Unfortunately, the merger did not proceed smoothly. After ATI began to lose market share, sales volumes and failed to deliver DirectX 10 lineup on time, questions arose whether the merger was actually necessary. Moreover, AMD-ATI reconsidered the Fusion central processing unit (CPU) roadmap a number of times, which made the merge even more questionable. The year 2007 was a complete disaster for the “new” AMD. The company delayed release of its 65nm quad-core AMD Opteron “Barcelona” processor and subsequently AMD Phenom X4 “Agena” chip by around three quarters and when it released them it appeared that the design had an erratum, which essentially stalled shipments. Moreover, ATI Radeon HD 2000 failed to beat Nvidia GeForce 8000-series in terms of performance and then ATI Radeon HD 3000 did not manage to offer performance advantages too. The graphics business unit - ATI - was back on track already in 2008 with award-winning Radeon HD 4000-series, meanwhile, AMD still fails to deliver products that beat Intel's with no compromise.
But the worst things that happened to the combined AMD were the necessity to sell ultra-low-power GPU assets, DTV assets as well as manufacturing facilities. At present AMD simply cannot compete on the market of consumer electronics and it will take years till AMD develops new ultra low-power graphics processing technology for smartphones and tablets as well as adopt x86 micro-architecture for those devices. Moreover, without actual in-depth knowledge of Android or Windows Phone platforms, the company will not be able to compete against already established players like Nvidia, Qualcomm, Samsung or Texas Instruments. Finally, without own production facilities, the company cannot quickly tailor process technology in order to boost speeds and improve yields of its current CPUs.
While AMD continues to have bright ideas about the future of CPU and GPU technologies, it lacks many assets to be successful in the next generation of devices.