Video: TI-99/4A BASIC Performance, Games and Comparison to Other 8 Bit Systems

[llopis]

Hi everybody,

 

I just released the last video in the TI-99/4A series (for now). This one focusing on using it, checking out some of the BASIC performance issues, and trying out a few games.

 

I'm hoping to do a followup sometime with the RAM expansion, speech synthesizer, and more games. Hope you enjoy it!

 

[Omega-TI]

I truly would have loved to have seen the Dragon's Lair cartridge featured near the end of this video.  Since the video addresses 'extras' built-into cartridges, I'm of the opinion that there is no finer example of a basic (non expanded console) pushed to the edge of the envelope with music, sounds, video and speech than Tursi's cartridge.

[llopis]

Just now, Omega-TI said:

I truly would have loved to have seen the Dragon's Lair cartridge featured near the end of this video.  Since the video addresses 'extras' built-into cartridges, I'm of the opinion that there is no finer example of a basic (non expanded console) pushed to the edge of the envelope with music, sounds, video and speech than Tursi's cartridge.

Me too but... I only found out about it right after I released the video!!! I will definitely cover it in a future video.

I'm bummed that I completely missed the boat and I can't get a Dragon's Lair cartridge today, but I actually spent an hour this morning going over the presentation of how it was done. Really, really amazing stuff!

 

[Omega-TI]

This kinda makes me wonder, if I could still run GWBASIC on my Windows 10 computer how fast the program would execute.

[llopis]

3 minutes ago, Omega-TI said:

This kinda makes me wonder, if I could still run GWBASIC on my Windows 10 computer how fast the program would execute.

People tried it with GW-BASIC on an IBM clone at 4.77Mhz and it was 34 seconds ?

https://docs.google.com/spreadsheets/d/1bfWSR2Ngy1RPedS6j-M607eeAhsd40-nhAfswILzzS8

Someone reported making that with Javascript and being something like 0.1 seconds. I suspect 99% of that was probably loading runtime libraries. Haha.

[Omega-TI]

Not bad.  Truthfully I miss GWBASIC...

 

[speccery]

Since ZX Spectrum Next is on the list I need to put my icy99 TI-99/4A FPGA implementation to the list as well. It currently runs the test in 20 seconds (CPU core clock at 25MHz, but the cycles are not directly comparable with TMS9900) both using TI Basic and Extended Basic. Obviously I need to push that below the 12 seconds the ZX Spectrum Next is doing here. 

[speccery]

2 hours ago, llopis said:

I'm hoping to do a followup sometime with the RAM expansion, speech synthesizer, and more games. Hope you enjoy it!

 

Thanks Noel for taking the time to make these interesting videos!

[Omega-TI]

In Classic 99 set to system maximum I get about 19-20 seconds as well...

 

 

 

[RXB]

4 hours ago, Omega-TI said:

In Classic 99 set to system maximum I get about 19-20 seconds as well...

 

 

 

Ok this is just weird!

Maximum with Classic99 is 15 seconds.

But on Overdrive it took 9 seconds?

So Overdrive is way faster then Maximum?

 

Edited December 11, 2020 by RXB
missing text

[HOME AUTOMATION]

The only way to go faster...
...is to slingshot around the Sun.?

[Omega-TI]

4 hours ago, RXB said:

Ok this is just weird!

Maximum with Classic99 is 15 seconds.

But on Overdrive it took 9 seconds?

So Overdrive is way faster then Maximum?

 

 

You are right, that IS weird.  I tried it on Overdrive and got 8 seconds.  I'm assuming Classic 99 only uses one CPU and thread?  Imagine what it could do with all six cores and twelve threads.

 

 

 

[Asmusr]

 

[Omega-TI]

Wow, that is one old message!  It does give a new meaning to the word, "eventually" though! ?

[RXB]

3 hours ago, Omega-TI said:

 

You are right, that IS weird.  I tried it on Overdrive and got 8 seconds.  I'm assuming Classic 99 only uses one CPU and thread?  Imagine what it could do with all six cores and twelve threads.

 

 

 

I have had 14 Classic99 all running at same time on Overdrive and it only hit 60% of CPU use overall.

It was using 12 CPU AMD 3900 and some of  them some only 6% but a couple at 24% so where it got the 60% load is not clear.

Only one of them was at 80% and then again I was also streaming video and using OBS Studio and 4 PDF open and Browser with 3 tabs open.

[Papalapa]

Good job llopis, well done...

[+mizapf]

13 hours ago, Omega-TI said:

 

You are right, that IS weird.  I tried it on Overdrive and got 8 seconds.  I'm assuming Classic 99 only uses one CPU and thread?  Imagine what it could do with all six cores and twelve threads.

 

Parallel cores and threads are only useful when the task can be parallelized. As for MAME, the core gurus have not managed to get this working or even conceived after 20+ years.

[Retrospect]

I watched the video and towards the back end, on the games list I saw "Lunar Mission" as in the first one i did, not "2" , and i thought "oh please god no don't choose that!" (you couldn't you didnt have 32K) but i'm saying that because for some reason i released the slowest ever compiled TI game and i am so grateful you didn't show it on the video.  

[+TheBF]

I'm not sure but there might be something wrong with your interpreters.

 

( Apologies: I couldn't resist trying this one)

 

I did have to resort to 32bit arithmetic in Forth do what BASIC does easily in order to have a sum greater than 65535 but the video is Classic99 running normal speed.

 

Spoiler

\ Uses 32 bit arithmetic

DECIMAL
: #0.0    0 0 ;      ( returns 32 bit 0 )
CREATE S  0 , 0 ,    ( 32bit variable )

: RUN
   CR
   11 0
   DO
     #0.0 S 2!
     1001 0
     DO
        S 2@ I  M+  S 2!
     LOOP
     ." ."
   LOOP
   SPACE S 2@ DU.
;

 

 

And to honest I wrote a direct translation of the BASIC which you would never do in Forth.

Instead of using a variable S you would normally accumulate the total on the data stack and then print it.

Spoiler

: FORTHWAY
   CR
   11 0
   DO
     #0.0   \ accumulate total on the data stack
     1001 0
     DO
       I M+ ( -- total)
     LOOP
     ." ."
   LOOP
   SPACE S 2@ DU.
;