In terms of performance, we could find no fault with the system. AMD's newer Congo platform does deliver much better performance compared to the older Yukon platform, while handily beating the Intel Atom yet again. While it still loses out to Intel's CULV processor based platform, it's still a pretty decent effort on AMD's part. Battery life wise, the Acer Ferrari One managed to secure a respectable uptime of 175 minutes, though power consumption at (much higher than either of Intel's solutions) remained somewhat of a weakness of the platform.

The 28nm process should start with risk production in late 2010, it looks like Q4 2010 now, and high end complex graphics chips are the main target for this. We are quite sure that AMD & ATI are interested in 28nm graphics chips and it’s just a matter of time when Nvidia will start considering Globalfoundries as an alternative fab.

Also learned is that from CPU tape out to shipping takes about 1 full year.

CPU manufacture is much more complex and takes more time than making a GPU and if the new AMD CPU will ship in Q3 2011, it better be taped out by middle of next year. The new K10.5 quad and dual-core revision that should start shipping in Q2 2010 has most likely been taped out by the middle of 2009 and this new core revision can provide some higher clocks and lower TDPs.

Lastly he finds out that they are testing out 32nm production with 300mm wafers now.

This schedule is part of AMDs design process and has nothing to do with the Fab, but it is too early to talk about the state of the process and yields. The 32nm transition will be a tough one. Despite the fact that transistors and transistor pitch are getting smaller, AMD also needs to implement high metal gate, and as far as we understood, this is not a walk in the park.

AMD's 40nm ATI Radeon HD 5000 series and 55nm ATI Radeon HD 4000 series GPUs are currently both facing shortages; the former due low yields on Taiwan Semiconductor Manufacturing Company's (TSMC's) 40nm process and the later due to AMD's conservative attitude towards its 55nm GPU sales after the launch of 40nm products, the sources noted.

AMD released a new generation of Radeon HD5000 series of graphics cards, in addition to the first to provide support for DX11, there is the biggest highlight Eyefinity technology, for the first time graphics output interface from the two-headed extended to three or even six, AMD claims that Eyefinity technology, the output image has gone beyond the human eye resolution limit, can be for gamers or professionals to provide the perfect display of the program.

We just recently found out some rather good news, Ontario in its 40nm won’t be based on Phenom / Athlon K10.5 45nm shrunk core. Originally we were thinking that AMD will simply take its dual-core in 45nm and will shrink it to 40nm bulk process, but we found out that the dual-core in that product is going to be Bobcat, a low-power dual core.

If ATI plays it safe it will develop the chip for both TSMC and Globalfoundries and will benchmark which of the two gets the job done better. AMD is yet to announce that it will officially do its GPUs in Dresden bulk part of factory, but this is something that won’t surprise many people.

He said that GPU scaling gives 2X performance every 18 months and AMD is not slowing down in CPU perfrmance either. The GPU gives the best performance per watt per dollar, he said. “We’re totally focused on the platform now and we have all the essentials.”

During the press conference, New York prosecutors said Intel abused its monopoly power “as a central business strategy” rather than just in isolated incidents. In addition, the attorney general’s office claimed to have evidence linking Intel’s top executives to these abusive actions.

The New York Times claims that Cuomo's office examined millions of pages of evidence over a 23-month period.