Well, I have stated in numerous posts over the past several months that AMD really needs to kill of socket AM3+ because the need to scale back their expenses. Focusing on socket FM2 for the consumer market makes sense for AMD both in the short term and long term. Basically having both APU and CPU for the consumer market is not really viable since both processors basically competes with each other for financial resources within AMD. Additionally, there is some sales cannibalization with consumers who are basically deciding if they should go with either socket FM2 or AM3+ as opposed to socket FM2 / AM3+ vs. socket 1155 (Intel).
AMD fans will not like the idea of the demise of socket AM3+ since it more or less means the only choice will be the lower end performance of socket FM2, but it really is the time for AMD to make the tough decision even though it is at the expense of the consumers. After all which scenario would be worse:
1. AMD continues to produce both AM3+ and FM2, but eventually faces the dire situation of running out of money to develop either one any further. And that hasn't taken into consideration the Radeon graphic cards yet. The outcome may actually be bankruptcy and the liquidation of AMD's assets. I would assume at that point VIA technologies would be interested in acquiring some AMD assets and patents.
2. The actual end of production of socket AM3+. AMD will truly stop competing with Intel in the performance arena. However, they will likely survive as a smaller leaner company with the prospect of growing again. Yes, people looking for a performance CPU will not doubt buy Intel whether they like it or not. But AMD's APUs are still a viable product for both laptop and desktop. The important take away is that AMD may live to fight another day.
While AMD's APUs are rather lackluster in performance at the moment, if they sole focus on that single product line for the consumer market, they should be able to improve performance and power consumption in later generations. Unfortunately, the APUs are a bit power hungry compared to Intel's Ivy Bridge CPUs and recently I have been recommending Ivy Bridge over Trinity for HTPC builds simply cause in the long term Ivy Bridge does use less electricity than Trinity; especially the Core i3 CPUs. Of course in cities where the cost of KWH is relatively inexpensive the argument holds little value. But for those who pays a lot per KWH (like 2x the national average), then power efficiency can play a larger role.
Hopefully AMD will push forward with the development of ARM server chips. While they may not match the performance level of the Opteron CPUs, they should end up costing less to purchase and reduce power consumption. Those two points should make ARM server chips an attractive alternative for small / medium businesses who are willing to give up some performance for a lower purchasing price as well as lower operating costs.
Hopefully the development of ARM server chips will also mean that AMD can get a shot to crack into the world of mobile devices (tablets & smartphones) where ARM based processors known as SoC (System on Chip) rules. It's not going to be easy though since Qualcomm and nVidia are amongst the larger and well known players in the SoC segment. Others include Samsung and Apple. Texas Instruments exited out of the SoC market because they feel the overall revenue generated by SoC is too small for them to bother with. They are currently in negotiations with Amazon to sell their SoC division. How much? Dunno. $3 - $4 billion perhaps?
This report HERE, which quotes an AMD spokesman, says semiaccurate is, well, not accurate at all.
Until and unless AMD officially announces it, probably during their Q4 earnings conference call in about 2 months from now, they would want to deny it. The last thing their stock price needs right now is for this rumor to gain traction and spread. If that happened, their stock price would drop maybe 50 cents or more. While 50 cents doesn't seem like a whole lot, it would be a drop of over 25% of the current price of under $2 a share..
We saw the same rumor on OBR's website last summer, so this would be the 2nd go-round for it. Demerjan appears to have a number of AMD contacts inside the company so I wouldn't discredit the rumor just yet. With the background of the other rumors about AMD laying off another 15% of the workforce starting in January, and the announcement by AMD that they are "exploring options" with JP Morgan to alleviate their cashflow crunch, I think it is entirely plausible that AMD is considering doing this, only they haven't formally decided yet which is why they say they are still "on track".