New Kickstart 3.9.1 68k on the way

[biggun]

Easy people.
Please please be easy. :-)

Lets phrase it this way:

There are few people left in the scene which spend time to improve the system still.


Whether these people develop new CPU cards,
or develop new chipsets, or try to write or port games,
or develop a new OS alternative
or whether they try to patch old OS.

The intention to help and keep the platform living as shared by all of us.

So lets please try not to fight but be happy that people have the will to improve our system.



I see great motivation in the work that Cosmos does.

I also Thomas point here that if this effort of Cosmos would be coordinated more with the OS4 guys or with the AROS guys the end result might even be better.

[biggun]

Updating the Amiga operating system is a tough task to begin with, and if it is to be taken beyond a hobby project, the work will have to involve a plan as to what goals should be achieved, a development work schedule, documentation, testing, collecting feedback and a release schedule. This calls for a coordinated team, and a one-man-army approach is likely to be overwhelmed by the workload.

You are certainly right here.
And developing OS upgrades in ASM is certainly not an easy task too.


On the other hand I can see areas where an ASM Hero could do great stuff on AMIGA.

For example the Phoenix Core we play around already includes
a number of new features and we will include more to strengthen it.

Options to code memory reads 100% latency free!
Options to write branches with will never mispredict!
And SIMD instructions.

An ASM hero who uses these well and is willing to count clockcycles for optimization
could certainly write stuff like new JPEG databtypes with this
which will run 10 times faster than the fastest 68060 does run today.

I know its a bold statement but I'm sure a good coder could do stuff like this.
I also think that good video playback will be possible with this.

Of course we won't compete with latest x86 cores.
But these would be "isolated" coding areas where an ASM expert could really change something.

[biggun]

There are many things that shouldn't really belong into kickstart. In fact, the kickstart should be as small as possible to ease upgrading and replacing components. Let's be realistic: Assume that we have a hard disk controller which can read at 1.5MB/sec (slow) behind a filing system, so you probably get 500k/second. That means that the entire kickstart could be loaded in a second. Give it maybe two seconds. Is two seconds boot time worth the effort of placing stuff in ROM?

This makes sense.
The ROM read physically limited to 3.5 MB/sec on classic
To fastmem on a CPU card the read speed would be over a magnitude faster.
This means for perfromance reasons we want to put the Kick in fastmem anyway.

So A1000 design to have a small bootrom and the KICK then placed into WOM-Ram is ideally for this.

[biggun]

If you want a faster Kickstart for 68000 base machines than why not remap it in to 32 bit Fast RAM?

This is what I'm saying.
In 68K CPU card you can read over 100 Mb/sec from memory.
This means you want your Kick in fastmem and not in ROM in the end.