Timing normal 32K RAM vs 32K 16bit RAM

[senior_falcon]

11 minutes ago, RXB said:

I do not have a console with 16bit memory on bus installed. But you already know that.

This is asking me to produce something I do not have while you refuse to do a more accurate test of what you do have.

Just add the clock and show the video with times.

Why are you resisting this request as you have the hardware?

I do not have a clock on my TI system, so right there my test will be flawed, at least in your eyes. After watching Bill's video, I do not believe anyone else would feel that a "more accurate" test is necessary. Since you are the only one so concerned with this, you can use Win994a which lets you use 8 bit or 16 bit bus for the memory. I wait eagerly for your results. But again, be sure you list the program so we know you are running the same program in both modes.

Edited June 8, 2022 by senior_falcon

[GDMike]

One Mississippi two Mississippi.works for me, it's still two.

[RXB]

1 minute ago, senior_falcon said:

I do not have a clock on my TI system, so right there my test will be flawed, at least in your eyes. After watching Bill's video, I do not believe anyone else would that a "more accurate" test is necessary. Since you are the only one so concerned with this, you can use Win994a which lets you use 8 bit or 16 bit bus for the memory. I wait eagerly for your results. But again, be sure you list the program so we know you are running the same program in both modes.

Do you have a TIPI? 

As it has a clock and most people are using a TIPI as it is one of the coolest devices ever made for the TI99/4A computer.

[senior_falcon]

Just now, RXB said:

Do you have a TIPI? 

As it has a clock and most people are using a TIPI as it is one of the coolest devices ever made for the TI99/4A computer.

Nope.

[RXB]

8 minutes ago, senior_falcon said:

Nope.

Well I stopped using Win994a as no way to make Cart modules for it that I know of as when I did get a RXB 2015 in one it took the creator to do it for me.

Why I use Classic99 as it is pretty easy do load GRAM into it and speeding up testing.

And I trust MESS/MAME over Win994a for timing as are there some problems I ran into with Win994a when I was using it.

Tursi fixes issues on a very decent basis and it took quite a while to get Win994a to address my issue.

[senior_falcon]

I dunno what to tell you then. XB is built into Win994a and I believe you could get a useful speed comparison using it. If MESS/MAME can handle 8 bit vs 16 bit then that's just the ticket. Otherwise just continue on as usual, and we will keep correcting you. As usual.

[Reciprocating Bill]

1 hour ago, senior_falcon said:

Do you have a TIPI?

This evening I'll run the maze generation XB program again, on both platforms, using the TIPI to display execution times. I'll create a video. 

[+mizapf]

1 hour ago, senior_falcon said:

I dunno what to tell you then. XB is built into Win994a and I believe you could get a useful speed comparison using it. If MESS/MAME can handle 8 bit vs 16 bit then that's just the ticket. Otherwise just continue on as usual, and we will keep correcting you. As usual.

FYI, in MAME you can turn on a "Console 32K RAM upgrade (16 bit)" with 0 waitstates. Alternatively, expansion cards (like "32kmem") may be used which are accessed on the 8-bit databus in the PEB (plus 4 WS).

[senior_falcon]

1 minute ago, mizapf said:

FYI, in MAME you can turn on a "Console 32K RAM upgrade (16 bit)" with 0 waitstates. Alternatively, expansion cards (like "32kmem") may be used which are accessed on the 8-bit databus in the PEB (plus 4 WS).

Ok Rich, here's your chance to show us how it should be done.

[apersson850]

11 hours ago, RXB said:

So I have posted poof and you have not.

Oh yes I have. I have shown several technical explanations to why the difference in execution speed with different memory types for systems like Extended BASIC is negelectible.

Admittedly, it does take some technical insight to understand them. Not too challenging, though.

The math skill required to follow the reasoning ends at division.

But now you've seen other videos here of the same task running on different systems and it's obvious the difference is so small that it's completely insignificant.

[Reciprocating Bill]

Here are the results of the above described experiment: One run of the maze generation program on a stock console without memory expansion, and one run of the maze generation program on a stock console with memory expansion. Herein I post photos of the completed mazes, with start and end times as measured by the TIPI clock.

 

As you can see, the difference in execution speed, although detectable by stopwatch, is not detectable by TIPI clock. Execution speed was identical.

 

 

 

 

 

 

I think that is what is called, "Hoist with his own petard." 

 

 

[RXB]

4 hours ago, senior_falcon said:

I dunno what to tell you then. XB is built into Win994a and I believe you could get a useful speed comparison using it. If MESS/MAME can handle 8 bit vs 16 bit then that's just the ticket. Otherwise just continue on as usual, and we will keep correcting you. As usual.

Your opinion is not fact.

I am not quite so arrogant as I have admitted to being wrong several times.

So far you have not done that which explains much.

[RXB]

3 hours ago, senior_falcon said:

Ok Rich, here's your chance to show us how it should be done.

Look so far you have continually argued that no need for accurate timing like using a clock as you are perfect and your opinion is based on your own beliefs so is infallible.

That sums it up in a nutshell.

[senior_falcon]

1 minute ago, RXB said:

Your opinion is not fact.

I am not quite so arrogant as I have admitted to being wrong several times.

So far you have not done that which explains much.

See post #186.

"I am not quite so arrogant as I have admitted to being wrong several times."

This might be a good time to do that again.

[RXB]

3 hours ago, apersson850 said:

Oh yes I have. I have shown several technical explanations to why the difference in execution speed with different memory types for systems like Extended BASIC is negelectible.

Admittedly, it does take some technical insight to understand them. Not too challenging, though.

The math skill required to follow the reasoning ends at division.

But now you've seen other videos here of the same task running on different systems and it's obvious the difference is so small that it's completely insignificant.

Again opinion is not factual. I feel like this is Facebook. No proof presented just opinion with no evidence.

[RXB]

1 hour ago, Reciprocating Bill said:

Here are the results of the above described experiment: One run of the maze generation program on a stock console without memory expansion, and one run of the maze generation program on a stock console with memory expansion. Herein I post photos of the completed mazes, with start and end times as measured by the TIPI clock.

 

As you can see, the difference in execution speed, although detectable by stopwatch, is not detectable by TIPI clock. Execution speed was identical.

 

 

 

 

 

 

I think that is what is called, "Hoist with his own petard." 

 

45 seconds vs 45 seconds!

Yep the test is invalid as the sample size can not show any difference therefore is invalid. 

Now run the test in a 10,000 loop and look at times and you be astounded to see they are not exactly the same anymore.

[RXB]

I will make you a $100.00 bet that if you run a 10,000 loop they will not be exactly the same and you most likely know this why you will refuse the bet.

[senior_falcon]

10 minutes ago, RXB said:

Yep the test is invalid as the sample size can not show any difference therefore is invalid. 

 

See post #21.

[senior_falcon]

16 minutes ago, RXB said:

Again opinion is not factual. I feel like this is Facebook. No proof presented just opinion with no evidence.

Rich, you have the tools to do the comparison yourself. Is there some reason you have not done this?

[senior_falcon]

26 minutes ago, RXB said:

Look so far you have continually argued that no need for accurate timing like using a clock as you are perfect and your opinion is based on your own beliefs so is infallible.

That sums it up in a nutshell.

Please show where I have ever argued that. It is true, of course, that I must use a watch when timing the real TI because I have no other way to do that.

My beliefs are based on my many observations. After you have seen something enough times you just naturally come to believe that it is true.

And by the way, this seems like it is getting kind of personal as well.

Edited June 9, 2022 by senior_falcon

[Tursi]

I'm trying to understand this argument... is it just a question of 16-bit RAM speed with regards to XB/RXB?

 

If I add a 16-bit RAM switch to Classic99 does it provide a tool to end the debate?

 

[+TheBF]

 

1 minute ago, Tursi said:

I'm trying to understand this argument... is it just a question of 16-bit RAM speed with regards to XB/RXB?

 

If I add a 16-bit RAM switch to Classic99 does it provide a tool to end the debate?

 

It's worse than that. The question is between BASIC  (or interpreted languages generally) running in VDP RAM vs Expansion RAM.

We do not see a material difference because the interpreter time is far bigger than memory accessing time.

 

Rich does not seem to believe this is true despite the evidence presented.

 

 

[RXB]

3 minutes ago, Tursi said:

I'm trying to understand this argument... is it just a question of 16-bit RAM speed with regards to XB/RXB?

 

If I add a 16-bit RAM switch to Classic99 does it provide a tool to end the debate?

 

RXB has nothing to do with it. 

If you test on Real Iron 32K vs 16bit 32K the test should be repeated so you can see the difference in time as the slower one will have less values then the faster memory.

This is not complicated rocket science here!

If you do like these guys argue just ignore that your timing method can not catch the difference as the sample size is to small to see any difference.

Which is why the failed to see any difference as the timing method they are using sucks!

 

Thus must loop that same routine in same computer with only difference being speed of memory will become apparent.

Bigger the loop the bigger the difference will become between them. This is simple logic.

 

And if there is no difference in speed it should be the same for 1 loop or for 10,000 loops, but I bet $100.00 it will not be the same ever!

[RXB]

8 minutes ago, TheBF said:

 

It's worse than that. The question is between BASIC  (or interpreted languages generally) running in VDP RAM vs Expansion RAM.

We do not see a material difference because the interpreter time is far bigger than memory accessing time.

 

Rich does not seem to believe this is true despite the evidence presented.

 

 

The test is bogus as it was to small of sample size to see any difference.

Like using a stop watch to catch milliseconds and you are trying to find Microseconds.

Simple science says the sample size they used is worthless for use.

 

This is why I bet money on it and will win!

[+TheBF]

2 minutes ago, RXB said:

And if there is no difference in speed it should be the same for 1 loop or for 10,000 loops, but I bet $100.00 it will not be the same ever!

Then we have a mis-communication.

Nobody said they were the same.

The evidence is that they are close enough that nobody cares.