[Info-vax] BASIC compiler in the hobbyist distribution

[seasoned_geek]

On Sunday, May 31, 2015 at 5:45:05 AM UTC-5, li... at openmailbox.org wrote:
> There is a difference but it's not big enough to:
> 
> 1) Make Intel a premium platform. Intel remains a crapware platform with
> high and low-ends.

Yugo came to America with a fully vaccumm controlled base model AND a turbo. BOTH were banned from the Tollway system because they could not maintain the posted speed limit.

> 
> 2) Stop the inevitable comparison between VMS on Intel and other commodity
> crapware OS and software running on Intel. And this is as a practical
> matter even worse than 1) above.

It does not matter how sweet the rose, once dipped in the septic tank, it will never be sweet again.

Once ported to x86 VMS will have a market lifespan of 15 minutes.

> 
> There are reasons IBM runs iOS on i boxes and AIX on POWER and none of it
> has to do with IBM's inability to write an OS or port anything anywhere
> they want. OS and software written specifically to target (good) bespoke
> hardware platforms work better, more efficiently, and get much more
> throughput than generic, portable OS and software even with some or much
> code optimization in the latter.
> 
> All the best loved platforms are and were owned end-to-end by one vendor.
> They were all developed to work together as greater than the sum of the
> parts. And they did.
> 

It never ceases to amaze me just how many people have bought into the marketing fraud that a game controller is good enough to run the world.

To achieve ultimate reliability and performance an OS must be written to utilize the strengths of their ecosystem. (This pretty much explains Windows since its ecosystem has no strengths.)

This IBM+NVIDIA venture will produce an ecosystem which will be the high end industry standard. People ignored AMD after they waxed INTELs arse with a 64-bit x86 INTEL could not build. They've been waxing their arse ever since, but people by into marketing fraud/hype/incompetence and horrible decisions get made. Having "dine & dive" Bob on their board didn't help AMD any either. So they bought NVIDIA and kept putting new designs on the video cards and now CUDA rules the land.

Very few of the projects for BOINC bother supporting Intel GPUs because they just don't measure up, and they don't support OpenCL. Even with a custom driver they still suck.

The IBM+NVIDIA native ecosystem will yield an exponential leap forward in processor power without any significant changes in either the CPU or GPU. This will be an "open" solution while anything INTEL does will be proprietary and and begin to be shunned internationally.

High end ARM isn't going to pan out. Too difficult to target designs for the near zero power consumption embedded market and the suck-all-power-from-the-planet server market.