Login
Username:

Password:

Remember me



Lost Password?

Register now!

Sections

Who's Online
34 user(s) are online (28 user(s) are browsing Forums)

Members: 0
Guests: 34

more...

Headlines

 
  Register To Post

TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
This may really make the fur fly, but here goes....

The NatAmi60 is fast, very fast, very, very, very fast!!!

With the proposed 68070 FPGA without an MMU, it'll be faster still!

So, what are the reasons that Hyperion could not recompile AOS4.x to 68K?


I know some of it is in PPC assembler, that was a HUGE amount of work, so that couldn't immediately be translated.



P.S. If I had the cash, I would buy Rogue a NatAmi60....
Whether he wanted it or not.

P.P.S. Any AOS4.x developers getting a NatAmi60?

P.P.P.S. I'm still entrenched in the A1 AOS4.x bunker. My A1 stays with me!

Support Amiga Fantasy cases!!!
How to program: 1. Start with lots and lots of 0's. 10. Add 1's, liberally.
"Details for OS 5 will be made public in the fourth quarter of 2007, ..." - Bill McEwen
Whoah!!! He spoke, a bit late.
Go to top
Re: TRULY "Too Hot To Handle"?
Not too shy to talk
Not too shy to talk


See User information
Quote:

The NatAmi60 is fast, very fast, very, very, very fast!!!


Is it? Have you actually used one?

Quote:

With the proposed 68070 FPGA without an MMU, it'll be faster still!


Wow, they're cloning the 68070? The 68070 was a pile of crap! It was around the same performance as the 68000

Quote:

So, what are the reasons that Hyperion could not recompile AOS4.x to 68K?


The NatAmi will not be as fast as a PC, and PCs are faaaaaaaar less obscure and more plentiful, so if there WAS a switch away from PPC, that would be the obvious choice, even taking the NatAmi into account.


Edited by xeron on 2008/6/9 10:22:22
Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@Atheist

It's not about recompiling. The kernel for example needs specific adaptions to a new CPU architecture. If I were to port to an entirely new CPU, I would certainly go for x86. Regardless of the speed of a 68k, the x86 will beat it, and the hardware is cheap and readily available.

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Supreme Council
Supreme Council


See User information
@Rogue

Bit OT but anyway, would you go x86-32 or x86-64Bit? Would the kernel be able to handle multi cores?

Vacca foeda. Sum, ergo edo

Mr Bobo Cornwater
Go to top
Re: TRULY "Too Hot To Handle"?
Just can't stay away
Just can't stay away


See User information
@Atheist

Quote:
The NatAmi60 is fast, very fast, very, very, very fast!!!
No, even compared to running UEA on the cheapest x86 PC you can buy it will be very slow.

Quote:
With the proposed 68070 FPGA without an MMU, it'll be faster still!
Are they gone completely mad now? The 68070 was a special m68k CPU used in the CD-i, IIRC it was a modified 68010, i.e. not only a MMU but a FPU is missing as well.

Quote:
So, what are the reasons that Hyperion could not recompile AOS4.x to 68K?
OS4 can't work without MMU
Even a 500 MHz 68060 would be way too slow for a PPC emulator, and without the possibility to run OS4 PPC software a m68k version of OS4 would be next to useless.


Quote:
P.P.S. Any AOS4.x developers getting a NatAmi60?
No.

Go to top
Re: TRULY "Too Hot To Handle"?
Just popping in
Just popping in


See User information
Some things are maybe a bit misleading. The planned 68070 is NOT a clone, it is a completely new design, a successor to the 68060.

For the thing with UAE, see this thread:

http://natami.net/knowledge.php?b=2¬e=641

QUOTE: "If Sysinfo makes you think that your PC is 100 times faster than a real AMIGA - don't believe Sysinfo.
A "cheating" 68k-JIT will look 10 times faster than it really is in Sysinfo."

BTW: Hi Atheist, I know you from the Natami board

Go to top
Re: TRULY "Too Hot To Handle"?
Supreme Council
Supreme Council


See User information
@acido

Successors was always using even numbers 000,020,040,060.

010 was only an evolution of the 000
030 was only an evolution of the 020

If sticking with the numbers then they should have used 080. Though not using the old motorola designations at all would have been better. Besides, 070 is already used.

Though I seriously doubt that they will be able to create a superscalar processor with execution pipelines, branch predictions, parallel executions and whatnot. It's more likely that it will just handle the 060 instruction set with some extras but not actually be very advanced in terms of a cpu design, with the clock speed being the only thing going for it.


Edited by orgin on 2008/6/27 12:05:26
Vacca foeda. Sum, ergo edo

Mr Bobo Cornwater
Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@orgin

Quote:
Bit OT but anyway, would you go x86-32 or x86-64Bit?


In the beginning at least 32 bits. And switch in CPU would be enough work already. I don't really think that the current API is suitable for 64 bits either.

Quote:
Would the kernel be able to handle multi cores?


Yes. We have a pretty good idea on how to achieve that, the only reason it's not in yet is the fact that we lack a multicore platform to test on.

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Supreme Council
Supreme Council


See User information
@Rogue

Okey. I suppose that it's safe to say that you or someone else on the team haven't used some spare time trying to port some of the core components just for the fun of it? :)

Vacca foeda. Sum, ergo edo

Mr Bobo Cornwater
Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@orgin

This is going to be a major step.

You'd have to rework and in fact rewrite the whole of the system startup, low-level handler code, interrupts, assembly parts, and the MMU code. Add to that that I don't have the slightest idea about x86 (last time I did x86 assembly was on a Siemens PC running CCPM/86 back in the 1980's), it is not likely to happen just for fun.

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@Atheist

Quote:
With the proposed 68070 FPGA without an MMU, it'll be faster still!

So, what are the reasons that Hyperion could not recompile AOS4.x to 68K?


I completely overlooked the first part...

AmigaOS 4.0 requires an MMU. It never worked without one, and will never work without one.

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Supreme Council
Supreme Council


See User information
@Rogue

Darn, bit too early to create x86.os4depot.net then :P

Vacca foeda. Sum, ergo edo

Mr Bobo Cornwater
Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@orgin

Quote:

orgin wrote:
@Rogue

Darn, bit too early to create x86.os4depot.net then :P


Yes ;)

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Just can't stay away
Just can't stay away


See User information
@Rogue

Quote:
we lack a multicore platform to test on

I thought you had a dual CPU AmigaOne ?

Rock lobster bit me, so excuse me.
X1000 + AmigaOS 4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." - Seymour Cray
Go to top
Re: TRULY "Too Hot To Handle"?
Not too shy to talk
Not too shy to talk


See User information
@orgin

Dont, my hart was thumping at a 100mph when you mentioned x86.os4depot.net that would be bloody amazing!! Without bringing up the old issue of ppc vs x86, I just can not imagine how many users there would be with an x86 version of os4! Ok, im going back down to my rabbit hole where I belong.

Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@TSK

Yeah, still have the module, a dual 7410 card. Unfortunately, there is no info on the thing, I don't know the voltage levels, or how the thing works internally. And since they're not going to go into production anymore, there isn't much point.

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Not too shy to talk
Not too shy to talk


See User information
@Rogue
would one of the Mac dual cpu modules be able to work with A1 XE if it could be physically made to fit? it has the same connector, not sure about pinouts though. Cpu's are the same 7455 as A1.

A1XEG4 PPC 7457 1.3GHz
CDTV!!
Go to top
Re: TRULY "Too Hot To Handle"?
Quite a regular
Quite a regular


See User information
@Swisso

I have no idea.

Seriously, if you do want to contact me write me a mail. You're more likely to get a reply then.
Go to top
Re: TRULY "Too Hot To Handle"?
Not too shy to talk
Not too shy to talk


See User information
@Rogue

Quote:

Rogue wrote:
@orgin


Quote:
Would the kernel be able to handle multi cores?


Yes. We have a pretty good idea on how to achieve that, the only reason it's not in yet is the fact that we lack a multicore platform to test on.


Would a PS3 fill the bill? :)

Valiant@Camelot
AmigaOne XE, 800Mhz, 1GB, 9250 Radeon, OS4.1u7
Sam440ep, 666Mhz, 512Mb, 9250 Radeon, OS4.1u6
A1-X1000, 1.8Ghz, 1GB, 9250 Radeon, OS4.1x
A1-X5000/40 2.2Ghz, 2GB, Radeon HD 7700, OS4.1 FE ud 2
Go to top

  Register To Post

 




Currently Active Users Viewing This Thread: 1 ( 0 members and 1 Anonymous Users )




Powered by XOOPS 2.0 © 2001-2016 The XOOPS Project