A collection of Atari BASIC benchmarks

[Mazzspeed]

3 minutes ago, Faicuai said:

No need for any of that since his choices have been already stated already on his GitHub.

 

Nothing else is more relevant at this point. We will follow them as closely as possible.

 

The rest only exists on your (tormented) toddler mind (not the Author´s or anyone else's here).

 

 

 

 

Sweet mate, no problem. Whatever rocks your little world, you go right ahead and believe it. ?

 

Edited June 9, 2021 by Mazzspeed

[Faicuai]

 

4 minutes ago, Mazzspeed said:

you go right ahead and believe it.

Yes, gotta give a bit of credence and at least believe he did what he says, on his site.

 

The rest is irrelevant (it does not matter if you like it or not, though).

 

 

[Mazzspeed]

1 minute ago, Faicuai said:

 

Yes, gotta give a bit of credence and at least believe he did what he says, on his site.

 

The rest is irrelevant (it does not matter if you like it or not, though).

 

 

Not interested in word mincing mate. The fact you were surprised that computers existed with no graphics chip speaks volumes.

Edited June 9, 2021 by Mazzspeed

[Faicuai]

21 minutes ago, Mazzspeed said:

Not interested in word mincing mate.

There is no word-mincing. What he did is what counts as common reference (nothing else).

 

And that is how we proceed: he runs a text-based terminal? we will do it. He runs a GPU/CPU model to show results? we will do it as well. Vis-a-vis, since his methodology is shot with holes.

 

There is no surprise (whatsoever) with a terminal-based model, as that is how we've run CP/m from Indius/GT (there is no GPU on it, of any kind). I already pointed to this reality eons ago on this thread.

 

Your fallacious argumentative style is useless, at this point. 

 

 

 

 

 

[Mazzspeed]

22 minutes ago, Faicuai said:

There is no word-mincing. What he did is what counts as common reference (nothing else).

Yeah, and lets forget everything else on that GitHub page in relation to testing limitations and methodology.

 

You believe whatever you want to believe, I'm over caring.

Edited June 9, 2021 by Mazzspeed

[Faicuai]

5 minutes ago, Mazzspeed said:

Yeah, and lets forget everything else on that GitHub page

What he DID is what counts. ACTIONS trump WORDS.

 

The rest is pure imaginary garbage.

 

Enjoy!

[Mazzspeed]

11 minutes ago, Faicuai said:

What he DID is what counts. ACTIONS trump WORDS.

 

The rest is pure imaginary garbage.

OK, just to quote yourself as according to you I'm the one being argumentative (I like the way you shamelessly chose to omit the part of my post regarding limitations and methodology when quoting my reply):

23 minutes ago, Faicuai said:

Your fallacious argumentative style is useless, at this point. 

Yet I'm stating believe whatever you want and you're the one that can't stop arguing, can you see the contradiction here? I don't care for your testing methodology, believe whatever you want.

Edited June 9, 2021 by Mazzspeed

[Faicuai]

Just now, Mazzspeed said:

Yet I'm stating believe whatever you want

What he DID is what counts.

 

This is irreductible and unquestionable. The rest belongs to your own delusion.

 

ACTIONS. Nothing else.

[Mazzspeed]

It's funny how it's not just my delusion.

[Faicuai]

5 hours ago, Mazzspeed said:

It's funny how it's not just my delusion.

You are right, I may have forgotten to include a buffoon here and there (my apologies for such inexcusable omission).

 

The truth of the matter is that you cannot honestly attempt to deconstruct factual-execution with guidelines-and-principles (and much less both stemming from the same person / author !) The former trumps the latter (not the other way around). 

 

To put it in perspective, you got a clear, detailed warning of this on post #74, and we are now in #160 !! After 86 posts, the author's actions have not changed a single bit (unless he updates the Github now)... but you still know better than himself... ??

 

You see now why you have been called intellectually dishonest, above?

[+Stephen]

On 6/2/2021 at 2:41 PM, Maury Markowitz said:

 

... thereby eliminating the CPU waiting on memory. Perhaps there is some other term than "wait state" I should use?

 

In any event, these numbers are not the same conditions, which explains the different numbers.

 

On 6/2/2021 at 2:51 PM, Maury Markowitz said:

 

Sure, but then it's no longer the same benchmark. The conditions of the test are just as important as the code.

 

Can you run the first set in ANTIC 0?

Just leaving these here for you again since this is the only point some of us are trying to get through your thick head.

[Mazzspeed]

49 minutes ago, Faicuai said:

You are right, I may have forgotten to include a buffoon here and there (my apologies for such inexcusable omission).

 

The truth of the matter is that you cannot honestly attempt to deconstruct factual-execution with guidelines-and-principles (and much less both stemming from the same person / author !) The former trumps the latter (not the other way around). 

 

To put it in perspective, you got a clear, detailed warning of this on post #74, and we are now in #160 !! After 86 posts, the author's actions have not changed a single bit (unless he updates the Github now)... but you still know better than himself... ??

 

You see now why you have been called intellectually dishonest, above?

Listen mate, I'm about to go to bed.

 

As stated, I'm intolerable of bullshit. I haven't read your waffle, I'm not interested in your waffle - But as stated, believe in the Easter Bunny if you want, I don't care. I certainly don't care for your convoluted excuse for testing methodology, and I'm not the only one.

 

You know what's funny? You drive a C63...one more digit and it'll be a C64! How's that for irony?

Edited June 9, 2021 by Mazzspeed

[Faicuai]

4 minutes ago, Mazzspeed said:

Listen mate, I'm about to go to bed.

Don't forget your plushy bear and you oreo cookies!

 

A c63 turned-into-a-c64! That would be cool! Finally a C64 that is not shit-slow! ?

 

??

 

 

[Faicuai]

6 hours ago, Stephen said:

Just leaving these here for you again

Since post #37,  we have now a better view of how the author actually conducted his own tests (which Maury did not see, at the beginning, either).

 

We are now in post #164. Having you stuck so far back on this discussion, is what holds you back out of any meaningful contribution... and precisely what has caused it to inflate it from a few simple posts, into a +160 posts NOP-sled.

 

 

[Mazzspeed]

4 minutes ago, Faicuai said:

Since post #37,  have now a better view of how the author actually conducted his own tests (which Maury did not see, at the beginning, either).

 

Having you stuck so far back on this discussion, is what holds you back out of any meaningful contribution... and precisely what has caused it to inflate it from a few simple posts, into a +160 posts NOP-sled.

 

 

I can't see your shameless manipulative bias changing anyone's mind, as it's clearly reduculious and well outside the requirements and regulations surrounding this particular benchmark.

 

But hey, believe what you want. I keep saying this and you keep returning with back against the wall posts that are no more than personal attacks. If anyone's dragging this thread off course, it's yourself.

 

[Faicuai]

21 hours ago, Mazzspeed said:

I can't see your shameless manipulative bias

I cannot see a single contribution of yours (on the entire 164+ posts thread) directly supporting Maury's repository building-efforts...

 

Not a single one. You know exactly why (and I know better, too).

 

Numbers (and ACTIONS) don't lie.

 

Sweet dreams! 

 

 

[drpeter]

On 6/4/2021 at 1:44 PM, drpeter said:

On 6/4/2021 at 10:31 AM, Mazzspeed said:

Bitcoin miner for the C64

Holy cow- you mean there is such a thing!?

 

How long would it likely take the entire residual world population of C64's to find a single bitcoin, and how much electricity would it consume, I wonder?

 

Seems I'm not the only one wondering- Bitcoin Mining With C64

 

TLDR: way too long & way too much money ?

 

Now the Atari 8-bit, that's a different matter.

 

By turning off DMA, you can die staring at a blank screen instead while awaiting your first bitcoin ?

Edited June 9, 2021 by drpeter

[Mazzspeed]

7 hours ago, drpeter said:

 

Seems I'm not the only one wondering- Bitcoin Mining With C64

 

TLDR: way too long & way too much money ?

 

Now the Atari 8-bit, that's a different matter.

 

By turning off DMA, you can die staring at a blank screen instead while awaiting your first bitcoin ?

Yeah, there's a bitcoin miner for the C64, I think The 8 Bit Guy did a video on it.

 

It's just basically a proof of concept thing, and a bit of a tongue in cheek laugh. I personally haven't tried it, I don't even mine bitcoin's on PC's. No one's actually going to use it, I hope.

Edited June 9, 2021 by Mazzspeed

[Maury Markowitz]

A seven page thread over whether or not turning off DMA is valid, complete with name-calling etc. Unexpected!

 

So let me simplify matters as the author of the repo in question:

 

The benchmarks should be run in Antic Mode 0 with the screen turned on. Any other display is considered an invalid test.

 

Changes to the math pack - not the rest of the ROM - are allowed, but have to be clearly identified.

[+Stephen]

42 minutes ago, Maury Markowitz said:

A seven page thread over whether or not turning off DMA is valid, complete with name-calling etc. Unexpected!

 

So let me simplify matters as the author of the repo in question:

 

The benchmarks should be run in Antic Mode 0 with the screen turned on. Any other display is considered an invalid test.

 

Changes to the math pack - not the rest of the ROM - are allowed, but have to be clearly identified.

Oh wow - look at that.  I guess I'm not the buffoon with the problem, and I was correct in calling out the DMA issue.  Thanks for the validation.

 

I put that in bold and big text so it can show up easily, even if we don't have a theater scaler in the way of our pure high bandwidth signal rendering chain.

[drpeter]

1 hour ago, Maury Markowitz said:

A seven page thread over whether or not turning off DMA is valid, complete with name-calling etc. Unexpected!

Really?

I take it you haven't followed similar 'comparison' threads...

 

1 hour ago, Maury Markowitz said:

Changes to the math pack - not the rest of the ROM - are allowed, but have to be clearly identified.

 

Based on prior experience, I fear you may need to be even more specific to avoid WW5 breaking out...  ?? e.g.

 

required- Standard BASIC Graphics 0 (Antic Mode 2, 40 columns by 24 lines), screen DMA and VBI enabled, PMG/sprite DMA disabled, all stock OS VBI code running

required- display via stock ANTIC & GTIA only

required- CPU 6502/6502C @ 1.79 MHz only

required- all declared variables and calculations must be floating-point only

state- NTSC @ 59.9227 Hz or PAL @49.8607 Hz only

state- stock OS ROM only- e.g. OS-A PAL/NTSC, OS-B PAL/NTSC, XL/XE Rev 1, 2,3, XL/XE/XEGS Rev 4, 1200XL

state- math pack & version e.g. stock Atari OS, Newell, Altirra Ver xxx, Turbo Basic Ver 1.5, MS Basic

state- real or emulated hardware version e.g. Atari 800, Atari 130 XE and emulator & version if relevant e.g. Altirra 3.90

state- BASIC type and version e.g. Turbo Basic XL 1.5, Atari Basic Rev C

state- if BASIC is compiled and if so, compiler and version used

 

all of the above must remain unmodified throughout running of the benchmark

 

I suspect you also need to define absolutely the sole BASIC listing that is valid for running the benchmark, or that will end up being endlessly 'optimised' too...

 

A tight statement of intent never hurts, such as ' The purpose of this benchmark is to run a standard and immutable BASIC listing in a variety of interpreted BASIC environments on real or emulated stock Atari 8-bit hardware and firmware within a defined standardised setup such that variation in the output benchmark is limited as far as possible to differing versions of BASIC and floating point mathpacks alone.'

Edited June 11, 2021 by drpeter

[drpeter]

6 minutes ago, Stephen said:

Antic Mode 0

He means Antic Mode 2 (BASIC Graphics mode 0)

[+Stephen]

1 minute ago, drpeter said:

8 minutes ago, Stephen said:

Antic Mode 0

He means Antic Mode 2 (BASIC Graphics mode 0)

The only relevant part of that is the fact he said DMA MUST BE ON.  That's been the bone of contention the entire time.

[drpeter]

17 minutes ago, Stephen said:

The only relevant part of that is the fact he said DMA MUST BE ON.  That's been the bone of contention the entire time.

In any sane world, obviously, but in a parallel universe where (for example) Covid vaccines make you magnetic and susceptible to 'uploads' via 5G,  'ANTIC Mode 0' might be seized upon as an invitation to have a display list of blank lines...

Edited June 11, 2021 by drpeter

[Faicuai]

52 minutes ago, drpeter said:

required- Standard BASIC Graphics 0 (Antic Mode 2, 40 columns by 24 lines), screen DMA and VBI enabled, PMG/sprite DMA disabled, all stock OS VBI code running

required- display via stock ANTIC & GTIA only

required- CPU 6502/6502C @ 1.79 MHz only

state- NTSC @ 59.9227 Hz or PAL @49.8607 Hz only

state- stock OS ROM only- e.g. OS-A PAL/NTSC, OS-B PAL/NTSC, XL/XE Rev 1, 2,3, XL/XE/XEGS Rev 4, 1200XL

state- math pack & version e.g. stock Atari OS, Newell, Altirra Ver xxx, Turbo Basic Ver 1.5, MS Basic

state- real or emulated hardware version e.g. Atari 800, Atari 130 XE and emulator & version if relevant e.g. Altirra 3.90

state- BASIC type and version e.g. Turbo Basic XL 1.5, Atari Basic Rev C

state- if BASIC is compiled and if so, compiler and version used

 

Exactly!

 

ALL of these variables fall under OUR choice-criteria, hence must be stated black-and-white, up-front. 

 

@Maury Markowitz

I would only add to the above list a clear or simple explanation of the reason for choosing a [GPU+CPU dominant] profile when, in fact, we can freely choose to run all of these tests in a [CPU-dominant] interactive environment (also generally available to any potential user, both in real and emulated domains).

 

This is important because, in the case of Bench64 (my specific contribution to the repo), all bets were thrown off as soon as its original author formally included the MinZ configuration and operating environment as a real contender (not to mention his plans to test on a PDP-8 minicomputer-compatible rig, etc.)

 

The author's actions speak for himself. We can perfectly model our A8 operating environment to suit the author's own.