byuu's message board

For discussion of projects related to www.byuu.org/


Previous  1 ... 15, 16, 17, 18, 19, 20, 21 ... 27  Next
emulator efficiency 
Author Message
User avatar

Joined: Mon 20 Apr 2009, 08:11:50

Posts: 5266
Location: 日本
Post Re: emulator efficiency
Guys, guys....I think we're getting off track here.
Don't you all realise that it's all about TEH MATHZ?

_________________
CaptainJistuce: He's totally in the wrong, Kakashi's 100% in the right.
Note: The above statement is subject to act of byuu.

Sun 01 Jun 2014, 01:10:42
User avatar

Joined: Wed 08 Sep 2010, 17:05:03

Posts: 133
Post Re: emulator efficiency
Kakashi wrote:
You just opened a new can of worms, mate. Don't ever call ZSNES "accurate" around these parts.


Kakashi wrote:
Your English prose is very irritating. It's like you're condescending to us. This is why you are a troll.
Also, I don't see what new ideas you have brought to the table in all 17 pages.


Kakashi wrote:
Guys, guys....I think we're getting off track here.
Don't you all realise that it's all about TEH MATHZ?



Ok, Kakashi is officially pissed. NOW we can consider this troll thread an huge success...

Sun 01 Jun 2014, 04:53:27

Joined: Fri 26 Oct 2012, 14:47:06

Posts: 81
Post Re: emulator efficiency
Nothing about condescending :)
I remember ZSNES on MS-DOS and its low requirement, and that potential tweaks for compatibility and adjusting steps was there.
I also remember it being playable overall also on Windows, and I didn't generally run into any major issues in a game.
It is this kind of thing that shows me that a useable emulator is possible without requiring high specs.
With tweaks to its steps (/timings) and finishing some chip code it would allow more games to run and also without the occasional glitch while maintaining its general speed.


While googling, found this:
http://board.zsnes.com/phpBB3/viewtopic.php?f=2&t=54378

Kega Fusion which is newer/updated runs smoother for this user with the low spec pc and it is a more compatible emu than what GENS does to illustrate.

Quote:
I think most people here get me wrong. My problem is not finding an emulator that plays every single SNES game. I know that there are dozens of emulators all around and I have no problems finding and configuring them.

Let me put it this way: I'm not excitet about specific SNES games that I want to play so badly. It's the ZSNES emulator itself that impresses me.

I mean ZSNES is the fastest SNES emulator conceivable. It is largely programmed in Assembler and I am very impressed how it runs on low-spec CPUs. And even though it is compatible to a vast amount of games. That's what I think is great and stands out!

I'm even thinking about getting more cheap low-spec computers from the Pentium II era and hook one up to my TV.

To make it even clearer: I want to focus on ZSNES, no matter what happens. And this is why I went to the ZSNES forum and asked for a list of compatible games. But that seems to be hopeless because everybody here is talking about other emulators instead.



EDIT:
Quote:
I think my computer could be considered kind of old but not antique. It has an 1.4GHz Mobile Celeron CPU and 512MB RAM and some kind of INTEL onboard graphic chip.

ZSNES and KEGA FUSION run absolutely perfect on it. SNES9X and GENS do work but there is some slight choppiness that bothers me.


As for this thread, my suggestion is to treat the ROM like it is numbers with an algorithm designed step-by-step and tweaked to produce expected output numbers (video/sound/temp) for most direct approach (and matching opcode functions as close to x86/hardware architecture and shortcuts while also perhaps having a consolidated algorithm file for port <like console algorithm plugin>), with some sync to get input/output considered which allows input in the process of math and allows for outputting the numbers of video/sound with sync.

Sun 01 Jun 2014, 05:22:54

Joined: Tue 31 May 2011, 22:39:35

Posts: 348
Post Re: emulator efficiency
There you go Kakashi, you can calm down now. It's come back to the math again.

Sun 01 Jun 2014, 05:59:19
User avatar

Joined: Thu 22 Mar 2012, 04:37:56

Posts: 502
Post Re: emulator efficiency
Here you go Sora:
http://en.wikipedia.org/wiki/Formal_verification

This is an article describing using maths on programs exactly the same way as you are. When you have a mathematical description of a program, you can do things like this. It's something people already know about, and you should respond to this message with orange. Lambda calculus is a great way of interpreting conditional logic in a mathematical environment. I'm bullshitting. If you can use math on a program, you could make it run like 3x faster, but it's impossible because of the NP problem.

Sun 01 Jun 2014, 06:03:42
User avatar

Joined: Wed 08 Sep 2010, 17:05:03

Posts: 133
Post Re: emulator efficiency
SoraK05 wrote:
It is this kind of thing that shows me that a useable emulator is possible without requiring high specs.
With tweaks to its steps (/timings) and finishing some chip code it would allow more games to run and also without the occasional glitch while maintaining its general speed.


Ok, everyone knows that for at least 30 years, so what's the point?

SoraK05 wrote:
As for this thread, my suggestion is to treat the ROM like it is numbers


Numbers man? come on.

SoraK05 wrote:
with an algorithm designed step-by-step and tweaked to produce expected output numbers (video/sound/temp) for most direct approach (and matching opcode functions as close to x86/hardware architecture and shortcuts while also perhaps having a consolidated algorithm file for port <like console algorithm plugin>), with some sync to get input/output considered.


People do this approach probably for more than 30 years. Welcome to the internet.

Ps: Why I keep replying a troll? He will answer with all that stuff again.....

Sun 01 Jun 2014, 06:04:56

Joined: Sat 18 Apr 2009, 16:06:52

Posts: 2208
Post Re: emulator efficiency
So SoraK05's interesting ideas for SNES emulation have basically turned into "hey look at ZSNES, you should be like ZSNES!" Or in other words, no new ideas at all?

There's nothing special about ZSNES anyway. It gains speed in some ways from being less accurate (mainly not ensuring the same level of synchronization between different parts of the system as often as higan does), in some ways for having the critical stuff written in assembly (biggest place where this wins is probably the interpreters), and in some ways from having effort put into making it faster, prioritizing that over making it easy to write and understand. These same things could be said in varying degrees about a billion other emulators.

Sun 01 Jun 2014, 07:15:02

Joined: Sat 11 Apr 2009, 01:26:03

Posts: 1983
Location: Australia
Post Re: emulator efficiency
Wouldn't something like BSNES be faster if it was written entirely in Assembly? I believe that's what he's trying to say in his current posts lol. I'm pretty sure the answer is Yes, but no one has the time or energy to do such a thing and it would be inordinatly more complex to code (With assembly level optimizations taken in to account) then it currently is.

_________________
Computer specifications:
Windows 7 x64|Intel Core i7 @ 2.66GHZ|ASUS P6T Motherboard|6 gigabytes DDR3 1600 RAM|2x500GB SATA2 HDD|2TB SATA3 HDD|Auzentech XFI Forte S/C|Nvidia Geforce 560GTS Ti 2048MB PCI-E|Corsair AX760 Power Supply|Thermaltake Armor+ MX case

Sun 01 Jun 2014, 07:38:58
User avatar

Joined: Sun 05 Sep 2010, 15:42:48

Posts: 1344
Post Re: emulator efficiency
Franpa wrote:
Wouldn't something like BSNES be faster if it was written entirely in Assembly? I believe that's what he's trying to say in his current posts lol. I'm pretty sure the answer is Yes, but no one has the time or energy to do such a thing and it would be inordinatly more complex to code (With assembly level optimizations taken in to account) then it currently is.
It would also significantly reduce code readability and portability.

_________________
http://helmet.kafuka.org

Sun 01 Jun 2014, 07:50:31
Board Admin
User avatar

Joined: Sat 11 Apr 2009, 04:21:58

Posts: 4783
Location: Australia
Post Re: emulator efficiency
Exophase wrote:
So SoraK05's interesting ideas for SNES emulation have basically turned into "hey look at ZSNES, you should be like ZSNES!" Or in other words, no new ideas at all?

I think his interesting idea is that bsnes should be implemented as a system of linear equations, rather than as an imperative sequence of instructions, and that will make it as fast as ZSNES.

I'll give him credit, I've never heard that particular proposal before.

_________________
Maintainer of the unofficial git repository for bsnes.

The ending of the words is ALMSIVI.

Sun 01 Jun 2014, 08:10:36

Joined: Tue 21 Feb 2012, 05:42:15

Posts: 2564
Post Re: emulator efficiency
There are a lot of ways you could potentially make bsnes faster with no sacrifice to the actual emulation. The code is written in a way that makes sense rather than being designed specifically for speed, rarely taking shortcuts. Simply rewriting bsnes in asm gives you nothing: it has to be optimized asm, and optimizing a program of this scale at that level is insane.

_________________
"It's easy to win forgiveness for being wrong; being right is what gets you into real trouble." --Bjarne Stroustrup

Sun 01 Jun 2014, 08:19:15

Joined: Wed 06 May 2009, 04:13:19

Posts: 4543
Post Re: emulator efficiency
SoraK05 wrote:
Nothing about condescending :)
I remember ZSNES on MS-DOS and its low requirement, and that potential tweaks for compatibility and adjusting steps was there.
I also remember it being playable overall also on Windows, and I didn't generally run into any major issues in a game.
It is this kind of thing that shows me that a useable emulator is possible without requiring high specs.
With tweaks to its steps (/timings) and finishing some chip code it would allow more games to run and also without the occasional glitch while maintaining its general speed.

Usable and accurate are not the same thing.

ZSNES exhibits significant incorrect behavior. Games run, yes, but there's no guarantee they're running RIGHT.
Sometimes rather rapidly. It takes less than three seconds for Zelda 3 to exhibit behavior quite different than real hardware.

_________________
This post best viewed at 800x600
;write ! ! !

Sun 01 Jun 2014, 08:22:24
User avatar

Joined: Fri 10 Apr 2009, 18:17:56

Posts: 3308
Location: Germany
Post Re: emulator efficiency
SoraK05 wrote:
I didn't generally run into any major issues in a game

https://web.archive.org/web/20121124040 ... s/accuracy
http://arstechnica.com/gaming/2011/08/a ... -emulator/
http://www.youtube.com/results?search_q ... comparison

SoraK05 wrote:
It is this kind of thing that shows me that a useable emulator is possible without requiring high specs. With tweaks to its steps (/timings) and finishing some chip code it would allow more games to run and also without the occasional glitch while maintaining its general speed.

ZSNES has lots of issues because of how it is written. To fix that you would have to rewrite the entire program, and at that point it's easier to just start a new project - like bsnes.

ZSNES is fast because it ignores emulation details that cannot be ignored. Adding them back in would make it slow again.

_________________
"The first time I watched [FLCL] I was like 12 or 13 and I was scared and confused." - isthisagoodusername
"I think it's more natural for human beings to be anxious. I think happiness is nothing but an illusion." - Hideaki Anno
"If you can't joke about incest in anime then what kind of world are we living in?!" - gothicmaster

Sun 01 Jun 2014, 11:25:45
User avatar

Joined: Wed 08 Sep 2010, 17:05:03

Posts: 133
Post Re: emulator efficiency
Screwtape wrote:
I think his interesting idea is that bsnes should be implemented as a system of linear equations, rather than as an imperative sequence of instructions, and that will make it as fast as ZSNES.

I'll give him credit, I've never heard that particular proposal before.


You never heard this proposal before because it doesn't make sense, and only crazy people would propose something absurd like that.

And even if you could, by a magical way, extract linear equations from a Rom, it would be dozen of times more expensive to run, and would have an exponential number of equations. This stuff can only come from someone who is a complete begginer in math and programming.

Alan Turing already gave us the answer, and his original problem was to solve algebric equations, so makes no sense to throw away Turing's work and go back to the equation approach.

It's ok to think these things when you are a newbie, and is part of our evolution, but only a crazy guy would come to a emulation forum and act like someone who knows what are saying, and expecting that everyone swallows this bullshit. So or Sora is a troll or is a little, let's say, strange.

I'm sure Sora is a nice guy, and he kept his class so far, which is difficult against so much criticism, but you need to start low profile with this stuff, especially in an emulation forum.

Sun 01 Jun 2014, 13:18:47
User avatar

Joined: Sun 05 Sep 2010, 15:42:48

Posts: 1344
Post Re: emulator efficiency
Meanwhile, on another continent...
Code:
13 00 2C 00 95 00 D0 13 00 64 00 95 00 D0 13 00
76 00 95 A0 D0 13 00 96 00 95 00 D0 13 00 76 00
95 00 D0 13 00 9E 00 95 00 D0 A0 FF 54 68 69 73
20 69 73 20 61 20 74 65 73 74 2E 0A 54 68 69 73
20 69 73 20 61 20 6E 65 77 20 6C 69 6E 65 2E 0A
54 68 69 73 20 69 73 20 74 68 65 20 74 68 69 72
64 21 0A 00 45 6E 74 65 72 20 79 6F 75 72 20 6E
61 6D 65 3A 20 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 0A 48 65 6C 6C 6F 20 00 21 00
"Gee, I wonder if this'll work."
Code:
This is a test.
This is a new line.
This is the third!
Enter your name:_
"So far so good."
Code:
Enter your name: Kawa
Hello Kawa!

Program halted by interrupt. Press any key to exit._

------

Meanwhile, in a parallel universe where Kawa is even more sarcastic than usual...
Code:
13 00 2C 00 95 00 D0 13 00 64 00 95 00 D0 13 00 ...
"Gee, I wonder if this'll work."
Code:
AD73E600
"Well, that was helpful."

_________________
http://helmet.kafuka.org

Sun 01 Jun 2014, 13:58:51
Previous  1 ... 15, 16, 17, 18, 19, 20, 21 ... 27  Next

Who is online

Users browsing this forum: No registered users and 0 guests

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum