I still think you're giving too much weight to the operating losses from the first year. At the same time based on another one of your posts and an assumed near holiday release date, PS4 won't be out long enough to have that kind of impact on FY14. I say "that kind" when saying 10M sold and $1B loss as that's 2-2.5M sold per month before the end of FY14. They might not even have production up to that level let alone sell that many in that time frame.
PS3.5, Webkit done and HTML5 <video> + WebMAF + Playready DRM in all Sony platforms and 20X a year increase in Internet connected platforms. It took 6 years to get 80 million connected platforms and it should be 100+ million a year after 2012. Also h.256 is coming on line with 4K as well as more efficient IPTV. Late 2012 Sony will have their own streaming IPTV and an expanded PSN.
And there is
this patent from August 2010 which is a HSA compatible plugin (1PPU+4SPU) wafer for AMD SOC. PS4 could have been a 4 wafer 1PPC+4SPU or Sony has another use for the wafers.
Aug 2010 Sony files patent for 1PPU+4SPU building block
Aug 2010 Xbox 720 Draft document CPU is either ARM or X86 with Backward Compatability via 3 PPC CPUs to be removed after 3 years
Dec 2010 Sony publishes 1PPU+4SPU building block patent with Xbar (same as AMD Fusion SOC CPU building blocks) 1 to 4 building block designs. 2=PS3, 4=PS4 PS3 Cell design dead
March 2011 AMD states AMD Fusion APU could be perfect in a Game console
July 2011 Digitimes rumor of a PS4 coming late 2012 with Kinect Interface (Probably true if PS3.5)
July 2011 Microsoft publishes the domain names microsoft-sony.com and sony-microsoft.com
Dec 2011 Sony publishes a processor agnostic Distributed processing patent using the terms Cell and APU interchangeably= HSA Foundation, APU =? AMD Fusion APU
Three August 2010 dates, interesting, what are the odds?
Jan 2012 Rumors that PS4 is (from developer hardware) X86
March 2012 Rumors firm up AMD Fusion APUs mentioned for PS4 Really wild rumors for Xbox 720 (PPC, 16 CPUs)
Till today where most Xbox 720 rumors are found to be based on the 2010 Xbox 720 draft-1 PPT presentation to non-technical sales types.
Given the above I think by at least 2010 Microsoft had already decided on X86 and AMD Fusion SOC (AMD said they had been planning this for 5 years)
Microsoft-Sony, not make a console together, share the SOC hardware with their own OS and on line services. With AMD making/providing the same southbridge, I/O and GPU series with the same laws of physics as to what can be packed into a SOC, they will have identical designs or nearly identical. If nearly identical costs more than identical where does it make sense to try to be different?
Only if one releases much later than the other is there a chance that the designs will be significantly different. Both in 2013, how will the designs be different?
This is obvious......but we are not able to put this together because we are assuming competition keeps it from happening. The services (on-line) offered are enough of a differentiation and this includes the OS, Free features etc.
The Xbox361 and PS3.5 started planning before August 2010 (we have proof of the Xbox361 in the Xbox 720 powerpoint). It must be in a SOC and must use a modern hardware design to support low power modes. Both would have to use the same everything except CPU. If the cost to include 6 SPUs is less than the cost of having two different SOCs then making the same SOC for both is economically the best decision. Supporting this is microsoft-sony.com and August 2010 for both Sony and Microsoft, the cite for start Nextgen Sony job posting and my cites for 1PPU4SPU patent and the release date for the Xbox 720 powerpoint.
Xbox361 and PS3.5 are next generation and play a key role as a bridge between older and newer more powerful both in the short term and long term. I suspect the Xbox361 (And PS3.5) SOC will find it's way into many CE platforms......that makes sense also.
OnLive model and serving handhelds are Next generation and I'd guess PS3.5 and Xbox 365.