Since these small SOC are BGA, a new mainboard would also have to be purchased to get the CPU/GPU.
But I can't imagine this being too expensive. $65- $75 for a new SOC/mainboard is a lot cheaper than paying $299 for another low quality netbook.
But, it's not just $65-75, even assuming the prices could be that low retail. There will still be different price tiers, based on what the market will bear. Zacate's prices are still strong not due to it being some kind of i3-killer, but because it's performance and feature balance is desirable to many. So, you might get a $75 Atom, but that upgrade would net you a generational IGP improvement only. For a real upgrade, consider it more like $100-150, and then you still might need to buy new RAM.
Then, if you are allowed to swap out other parts, like the display, expect very high RMA rates. Regular users, who don't treat their notebooks well anyway, are not going to have high success rates swapping out keyboards, displays, or AIBs.
Finally, what ports are you going to standardize on? External IO in a notebook or tablet is part of the body and frame. Not only is there no IO plate, but such a plate would necessarily add bulk, since now whatever space it takes up can not be used to support the structure of the notebook. If you decide on some set, then you can't add or remove them for good reaosns (space saving, tons of USBs, multiple Expresscards, VGA v. HDMI v. DVI v. DP, etc.). A desktop can differentiate itself through several means, which are not dependent upon the body, but portable computers, by and large, don't have that luxury. They
could come up with standards for replaceable GPUs (tie form factor to TDP), but attempts at that have failed for reasons I'm not sure of.
The netbook, thanks to market manipulation, is a rather special case, and other portable network devices, from unencumbered notebooks to smart phones, do not share the netbooks' lack of improvements over time. Some makers were able to differentiate theirs (Samsung had a many-USB model, and one with bumpers on the corners, off the top of my head), but they were still stuck in shackles, keeping them all very close to cookie-cutter.
The key factor is making the cheap hardware and software easy to swap.
When 1/4" of thickness is not the entire world, that works. But, 1/4"
is a world of difference in notebooks, and 1/8" is for tablets and phones. If you make the hardware cheap and easy to swap out, you can't make it light or thin. if you want to make it thin and light, but easy to swap out, it will still not be cheap. And, if you succeed, those who don't follow your standard, and who can make computers
that differentiate themselves from the competition, will instead get mass market appeal.
IMO, this is why using Microsoft Windows fails. Nobody is going to pay $100 for a single use OEM license or $180 for a multi-use Retail license and then install that OS on a cheap $65 SOC Mainboard.
Non-portable Atoms disprove this. People can and do spend money on an OS to run on such cheap hardware. They'd prefer cheaper, of course, but it was a market-manipulating move to make a cheaper-only OS option for cheaper hardware. The OS' cost should be orthogonal to the hardware's. Microsoft may very well need to change their pricing schemes in the future, however. It is also a good example of why we need AMD to be competitive: late and hot, Zacate still put Atom in its place, and took off where netbooks stagnated, even models like the C-50, which offer fairly little over Atom.
High end hardware would be the goal of someone spending that kind of money. The economics of the MS OS situation is probably one good reason we only saw larger size notebooks like ASUS C90 or OCZ "DIY Notebook" (with their heavy and difficult to remove coolers) billed as upgradeable.
Even those, however, had the exact same problem as I noted about my desktop: Nehalem changed sockets, and actually did real beneficial work through the socket change. Core 2s did need material changes to mobos, to take full advantage of their power management, before that. So, you would need to be able to upgrade the entire mobo ($200?) to get anything useful out of the upgrade.
It's like other electronics. Many people tend to either buy cheap and throw them away, want to have the latest and greatest, and/or feel that their technology spending is part of their self-worth. Those people are easiest to market to. If people demanded products good enough to keep for several years, then what you see sold on the shelves would reflect that. I especially believe this to be true since products do exist that fill those needs, in the form of configurable notebook bare-bones for small vendors, and business notebooks from big vendors. If people stopped buying the cheap crap, manufacturers could easily shift to making more of the higher end series, and then go from there. As it is, that market is just too small.
Facebook
Twitter