>>>> I think you do not realize the value for Imagination Technologies to have an
>>>> efficient performant and ultra-low latency operating system supplied to them.
>>> I guess they don't and won't either
>> I am not so sure, It is worth gold.
> Pipe dreams.
Let's see, why don't you say what would be the price of such a SoC? If it is less than 1 million, an excellent operating system is definitively an adequate compensation. If it is a lot more, some investors can be found to fund startup costs.
This doesn't change the fact that a smartphone or tablet would still be a poor substitute for a convergence box. The input output capabilities do not correspond to the needs (no OTA tuners, several missing peripherals, etc.). To me, the fact that Apple added an application store to its AppleTV is proof that the market is slowly evolving towards the convergence box.
> I consider it strange to call a SoC intended for NAS devices a "NAS design"
Well it is a whole design if you consider that it was accompanied by a PCB reference design which NAS manufacturers could use and have used as a starting point for their NAS boards. For example, boards in the PPC NAS models from Synology, as well as some other manufacturers are almost unchanged from the reference design. If it was only a SoC (and perhaps a test board but no board design which customers can reuse for their products), it would not be a NAS design but this isn't the case.
To all those who think that only ARM and X86 will survive, this is unlikely. The MIPS market is growing, give them 10 years and they will be used quite a bit. MorphOS doesn't need to chose a platform which is already popular, it needs to use a processor which will be popular, possiblly contributing (but not alone) to the said popularity.
> "I skate to where the puck is going to be, not where it has been."
said by - Wayne Gretzky (famous Canadian hockey player)
Moreover, in the file I said:
> The very reason why the mobile market, with tablets and smartphones is dominated by the
> ARM processor is quite simple, in the mobile market power consumption is of utmost
> importance and x86 simply cannot comptete.
This means that x86 will not become very popular in low power devices.
> Interestingly, the very problem which the x86 architecture suffers when trying to penetrate
> the BIG category [servers] is the exact same than in the SMALL category [mobile]:
> power consumption. There is a limit to the possible TDP of a silicon dye without the
> processor frying. When two processors are made at this limit, the most power efficient
> will also be the most powerfull.
And I also said:
> Moreover, datacenter operators are complaining about their energy bills.
> This is the very reason why everybody is trying (but failing) to build ARM servers.
> The problem with these is that once the clockspeed and number of cores in an ARM
> platform is scaled to server levels, its perfermance per watt is only comparable to x86
> but not [much] better.
The performance per watt metric will soon replace the performance per dollar as the choosing factor for datacenter operators. As soon as someone comes with an alternative to x86 which is more power efficient and easely available, it will become popular. In other words there will be another popular architecture along the x86 and the ARM since ARM has an adequate power efficiency in the mobility context but not the datacenter context. Finally, no one know for sure what that architecture will be but it will be but MIPS is a likely candidate, especially since its market share is growing and the interaptiv and Warrior-i are the processors of choice of cheap chinese tablets/smartphones which means it will have a significant market share very soon.
As a conclusion, there WILL be an alternative to the ARM/x86 duopoly.