for a couple million dollars

DeltaV · Post by **DeltaV** » Fri Oct 28, 2011 3:28 pm

From the colorForth link on that site:

Rationale

Current software is shameful. Giant operating systems linger from the 1970's. Applications are team-produced with built-in obsolescence. User interfaces feature puzzle-solving.

With the huge RAM of modern computers, an operating system is no longer necessary, if it ever was. colorForth includes multi-tasking, and drivers for essential devices. But that is hardly an operating system in the style of Windows or Linux.

Megabytes of software may be required for historical compatibility, but sometimes we need a fresh start. ColorForth provides the necessary capability with kilobytes of code. At boot, it copies disk into RAM. Then compiles the macros that emulate the stack machine that Forth expects. As applications are requested, they are compiled.

A Forth application can take 1% the code employed by the same application written in C.

I like! Toss 'em a couple million!

seedload · Post by **seedload** » Fri Oct 28, 2011 7:16 pm

GIThruster wrote:I'm wondering about a friend who would make a good CA candidate, and whom might embrace energy independence and the need for things like LFTR as part of his platform. I've told him before he needs to run and one nice thing is that his wife and her family have worked around the country for more than 40 years. They're all quite knowledgeable from a practical standpoint. Can't help but wonder if people here really want to see nukes that burn waste, etc. Any candidate proposing something like LFTR would need a panel of experts in support of that position.

Somewhere like here would be a better place to go for opinions and options.

http://www.energyfromthorium.com/forum/

hanelyp · Post by **hanelyp** » Sat Oct 29, 2011 1:24 am

Rationale
With the huge RAM of modern computers, an operating system is no longer necessary, if it ever was. colorForth includes multi-tasking, and drivers for essential devices. But that is hardly an operating system in the style of Windows or Linux.

With the immense RAM and other multi-tasking capabilities of modern hardware, an operating system is needed now as much as ever. Specifically to mediate access to resources by applications which are blind to each other's expectations, and many of which are less than well behaved. It sounds like colorForth intends to take over a subset of essential OS functions.

MSimon · Post by **MSimon** » Sun Oct 30, 2011 9:52 am

hanelyp wrote:
Rationale
With the huge RAM of modern computers, an operating system is no longer necessary, if it ever was. colorForth includes multi-tasking, and drivers for essential devices. But that is hardly an operating system in the style of Windows or Linux.
With the immense RAM and other multi-tasking capabilities of modern hardware, an operating system is needed now as much as ever. Specifically to mediate access to resources by applications which are blind to each other's expectations, and many of which are less than well behaved. It sounds like colorForth intends to take over a subset of essential OS functions.

Well no. You isolate the various functions. i.e. the tasks are run in parallel.

If we got rid of long pipelines, out of sequence processing etc. and made more smaller cores on a chip the tasks could be isolated and much more parallelized. We do that to some extent with display processors in a PC. And the display processors are starting to run to multicore architectures.

In any case the GA 144 is more useful for embedded because of its slow core speed (~ 650 MHz) limited RAM (about 1 MB IIRC) and limited cores - 64 words of RAM (up to 4 instructions per word), 64 words of preprogrammed ROM (can be customized).

But for embedded it is one of the slickest chips I have ever seen. Power management is automatic and at idle it uses about 15 uA @ 1.8V. That is 100nA per core.

MSimon · Post by **MSimon** » Sun Oct 30, 2011 9:58 am

The point is that with new ways of doing things we are sticking with legacy instead of just emulating it.

Skipjack · Post by **Skipjack** » Sun Oct 30, 2011 1:08 pm

Msimon, the problem with the GPU- cores is that their speed degrades exponentially with complexity of the code.
An additional operation that would act as an optimization on a CPU will make a GPU slower.
They favor fast code over "smart" code, at least in my experience.

MSimon · Post by **MSimon** » Mon Oct 31, 2011 4:57 am

Skipjack wrote:Msimon, the problem with the GPU- cores is that their speed degrades exponentially with complexity of the code.
An additional operation that would act as an optimization on a CPU will make a GPU slower.
They favor fast code over "smart" code, at least in my experience.

True. But the way I look at these chips is not as software, it is as software designed hardware. Basically they replace HCMOS. i.e. stuff that runs at about 1 to 10 MHz. So what is the slowest the chip will run? About 2 to 3 MHz if it executes 256 (or so) instructions.

Now what are the advantages? Board respins are not required if there is an error or improvement. You change the code.

The ultimate price is about $2 a pop. Not a bad price for 20 or 50 CMOS chips.

I'm working on a project using one of these. I'll report on it once I get done.