Having worked for a medium scale company using MIPS SOCs for thin clients, I regretfully vouch for this as correct to the bone. The engineering involved in designing heat flow and other non software aspects is labor of distinct quality and of design-build-trash-design-rebuild iterations.
Finding the common ground between energy consumption and computing strength is a royal pain of months on months, and when you realize you need to redesign the board and cover....then we go back to the mentioned above. Even designing and boot strapping the production line is a royal pain. Finding a trustworthe quality manufacturer is close to impossible in other than mega volumes, and forget about integrating your production line qa procedures if you are not big enough. And we did not discuss defect rate of production line and.. and... In short, yes, there is a very long way from POC to actual product. -Sivan On Sun, Feb 13, 2011 at 1:28 AM, Jean-Christian de Rivaz <j...@eclis.ch> wrote: > How many product do you have build and produce in a substantial quantity to > over simplify the reality like you do ? > > The first thing you need to know in the electronic component business is > that if you are not a big company that sell *million* of devices per year, > you don't ever have access to the minimal documentation needed to uses > advanced products. Unlikely selling opportunities only have access to the > "industrial" range of products that is usually one generation old. For > example Texas Instrument clearly split the OMAP products for big OEM, and > Sitara products for industrial quantity. And this example is easy. Things > get far more complicated for the 3G components (and not only for the silicon > parts). > > Without those components, you can't build a handset at all. Your suggestion > is only likely to produce a technical experiment to show that a set of > interconnected modules provides similar capabilities compared to a handset. > The transition from this experiment to something you could decently call a > handset is the very hard part of the work that only big companies can > possibly archives. I will not be surprised if the conformity testing of a > such complex device excess the cost of designing it. > > Building a product in any substantial quantity is not only a question of the > product itself, but many more in how you will solve all the problems in > making the logistic and the tools to manufacture the product. This is the > very hard part. If this was easy, then the market would be saturated by a > big numbers of small players. The reality is that only a few companies can > play at this level of the game. > > This is why so many peoples contributing to Meego, or interesting about it, > are actually so stressed about any announcement concerning Meego products. > The community itself can do something on the software, but are completely > left out of control about the hardware. Without big companies providing > Meego hardware, there is unlikely that Meego will have any uses in the > market it was designed for. > > Best Regards, > > Jean-Christian de Rivaz > > Le 11. 02. 11 23:46, Rudolf Streif a écrit : >> >> What do you need to build a handset? >> a) a processor board; start with the PandaBoard (dual-core OMAP4, BT, >> WiFi, HDMI, and much more) >> b) a wireless network radio (use your favorite search engine and search >> for "GSM radio module" or similar) >> c) a GPS module (same as b)) >> d) a display with capacitive touch (same as b)) >> e) a g-sensor/accelerometer (same as b)) >> f) engineering ingenuity > > _______________________________________________ > MeeGo-community mailing list > MeeGo-community@meego.com > http://lists.meego.com/listinfo/meego-community > _______________________________________________ MeeGo-community mailing list MeeGo-community@meego.com http://lists.meego.com/listinfo/meego-community
