The horizontal scroll value takes almost no time at all to change, it takes even less time than setting the LCD output coordinates for a single sprite Wink
KermMartian wrote:
PCSEball runs at 120 frames per second, or 1208 16x16 sprites per second (at 15MHz).


Yikes; without having seen a 84+CSE in person, I was already biased from what I've heard and assumed the screenshot was an accurate representation of PCSEball's refresh rate (roughly 5–7 fps, it appeared to me), even though I should have known better.

With all this talk and speculation, I think it would be a good idea to have more calibrated screenshots/videos posted of various programs and operations so that the rest of us can gain a more accurate idea of the true graphical performance.
I just noticed this thread. I read through the whole thing and I am very surpised that people are even considering buying this thing. The idea of putting a z80 processor and paring it up with a 320x240 screen is the dumbest idea I have heard of in a very long time. Why can't they just throw in an arm processor like everyone else does? People seem to forget that a faster processer can reduce development time and hence reduce R&D cost. TI makes very nice arm processors and since they are the same company they can get them for low cost. $45 gets me a beaglebone black which has 512mb of ram and a 1gbz arm processor. By the way it was designed by texas instruments. The profit margin is lower however. The business model of the beaglebone black appears to be to give programmers experince with the processor in hopes that they use it in a design. TI also does this with the launchpad series at one point I could buy an msp430 launchpad for $4.30 and that includes shipping. I bought two before the price went up. I understand that the business model is different but that is no excuss for the high price. They have people perfectlly capable of desinging good low cost capable hardware so their is no excuss for using a z80 processor when the same compay designs CPUs and produces arm processors. Also right now I can go on ebay and buy a nice 3.2 inch tft screen (320x240) on ebay for around $10. That is in single quanities and of course in bulk that will be even less per unit. I'd estimate that a $90 dollar calculator would allow for a nice caclulator with a decent cpu and flash memory while having an acceptable profit margin. When recomending a calculator I would say get an HP prime or a casio prizm. Also the argument for buying a TI just because your teacher uses one is quite silly. (At least for the prizm) you would have to be quite the idiot to not be able to figure out how to get a basic understanding of its usage. For the more complex things read the manual carfully or do a search message boards that is how I figured out the extra minor details for some aspects of the prizm. Also I do not yet own a prime but I may get it one day. The hardware is nice.
Some fair points but from a programming perspective a good challenge can be quite nice ... and the 84C does present one when talking performance.

Still some nice software can still be made on it, which is good Smile.
Yes I do like challenges but there becomes a certian point where there is no real solutions. I guess the challenge is working around the hardware not trying to do what you really want to do. For example I wanted to challenge myself by getting camera modules on the arduino uno and mega. I did that but what I really wanted is fast frame-rate however there is a maximum limit on how many bytes per second can be transmitted. You can see that even though it is programmed to work it is slow https://www.youtube.com/watch?v=sEsZeDC6co0 One problem is that PCLK (pixel clock this lets me know when pixels are outputted) speed changes based on brightness I could make it go faster by manually counting clock cycles instead of a loop this would allow me to program the ov7670 to output data faster. The loop has to read from the port check the bit and branch based on that. NOP only takes one cycle. So in this instance I would have to make a compromise for speed. I have had no interest in trying to get a constant PCLK speed because I believe this would involve the use of a fixed exposure. See sometimes the extreme optimizations result in a loss of a feature that would be well missed. The reason I made the camera module project was for time lapse purposes and to not have auto exposure would be terrible. To write and test different pixel reading code for all exposure values would take a very long time.
I must agree with ProgrammerNerd on how sometimes working around the hardware isn't worth it, but the pcse is a piece of cake compared to the ti 81 Razz Not that any comparison between the two should ever be made, mind you.

ProgrammerNerd wrote:
I would say get an HP prime

And the HP prime is amazing.
ProgrammerNerd wrote:
I just noticed this thread. I read through the whole thing and I am very surpised that people are even considering buying this thing. The idea of putting a z80 processor and paring it up with a 320x240 screen is the dumbest idea I have heard of in a very long time. Why can't they just throw in an arm processor like everyone else does? People seem to forget that a faster processer can reduce development time and hence reduce R&D cost. TI makes very nice arm processors and since they are the same company they can get them for low cost. $45 gets me a beaglebone black which has 512mb of ram and a 1gbz arm processor. By the way it was designed by texas instruments. The profit margin is lower however. The business model of the beaglebone black appears to be to give programmers experince with the processor in hopes that they use it in a design. TI also does this with the launchpad series at one point I could buy an msp430 launchpad for $4.30 and that includes shipping. I bought two before the price went up. I understand that the business model is different but that is no excuss for the high price. They have people perfectlly capable of desinging good low cost capable hardware so their is no excuss for using a z80 processor when the same compay designs CPUs and produces arm processors. Also right now I can go on ebay and buy a nice 3.2 inch tft screen (320x240) on ebay for around $10. That is in single quanities and of course in bulk that will be even less per unit. I'd estimate that a $90 dollar calculator would allow for a nice caclulator with a decent cpu and flash memory while having an acceptable profit margin. When recomending a calculator I would say get an HP prime or a casio prizm. Also the argument for buying a TI just because your teacher uses one is quite silly. (At least for the prizm) you would have to be quite the idiot to not be able to figure out how to get a basic understanding of its usage. For the more complex things read the manual carfully or do a search message boards that is how I figured out the extra minor details for some aspects of the prizm. Also I do not yet own a prime but I may get it one day. The hardware is nice.


First off, have you heard of the Nspires? Or of their PLT line? They have already made ARM-based calcs. Have you also heard of the m68k-based calcs? TI already has separate product lines.

You wonder why can't they just "throw in an arm processor like everyone else". You claim that it can reduce development time. The z80 line of TI calcs were released in 1990, just 5 years after Acorn made the ARM1 and 4 years after the production ARM2, and just 3 years after the first production ARM computer (Acorn Archimedes). The z80 has been out for 11 years by then. The z80 was and is popular for embedded uses, whereas ARM2 was most likely too new or not suitable for a low-power small embedded device.

You stated that "People seem to forget that a faster processer can reduce development time and hence reduce R&D cost." Have you forgotten about Moore's law? As the cost of computing power falls, the cost for R&D, manufacturing, and testing rise.

The z80 line of calcs has been going on for 24 years now. There has been a lot of work put into the OS and its math functionality as well as the hardware designed for it. If you want to "just throw in an arm processor", you just threw away 24 years of work because now you need to get software that was written in z80 asm to be replaced, a hardware team to create a new system around an ARM-based CPU, and to get on marketing and supporting a whole new line of calcs.

And you also complain about the price. Just as KermMartian asked Peter Balyta "So to summarize, when you're buying a TI calculator, you're not paying for the calculator, you're also paying for the complete solution behind it. " with the response "Absolutely.", there is a lot more going on than just the hardware cost.

You also compare the calc's cost to that of a beaglebone black. Are you not forgetting that is is costly to design an ASIC incorporating a Z80 core, USB core, RAM, and other peripherals? Then the cost to produce these chips for a less-wide audience than a generic microprocessor? If you go this route, you basically end up running around trying to morph into the Nspire line.

Putting in a color display IMO seemed like the appropriate thing for TI to do in order to keep their separate active calculator product lines together and not falling behind. Does the TI-84+CSE do math to the extent and beyond than the original mission for the early z80 calcs? Yes. Is it usable for educational environments with advantages over the TI-84+SE? Yes.
Edit just noticed Ahelpers response. I admit I did not take into account existing z80 code but I still do believe that more powerful hardware can reduce R&D due to the fact that it is easier to program. Also yes I have seen the Nspire and it has lots of potential the only problem with it is that TI locks out native programs. Otherwise it would be great calculator you could use ndless but will that work forever, no it takes lots of time to play cat and mouse with TI. Also Peter Balyta is a marketer it is his job to say stuff like that.
My below post was in response to jetlego
About it not being worth it to work around hardware. One thing people seem to be forgetting is the cost savings in terms of time of the programmer and also R&D cost go down as mentioned before. Programming for the arm is easy and the same with other 32bit cpus such as the sh series. A powerful open calculator allows for people to quickly develop programs for it. For example when I wrote my casio prizm mpeg2 player all I had to do was download a library called libmpeg2 change a few lines then compile it for the prizm's sh4 cpu. Then I just modified an example that uses libmpeg2 and I am done. It only took me a few hours worth of time to develop that program but it got the community excited as they got to quickly see the capabilities of the prizm. At-least in theory having nice hardware means more programs released that are of acceptable quality. My theory seems incorrect for the prizm as of now but I think that if I put together more practice examples of programming for the prizm and practical example of library usage (allowing people to reuse existing code) and making more cool programs that show what a better CPU can do, I think it will motivative the community and more people may take up prizm development then there will be a network effect (as in more people start programing for the prizm because many people are programming for the prizm). This is just my theory and I do not see this idea limited to the prizm, it is just an example I used because I am familiar with it.
You are clearly ignoring Moore's law. Just because something is easier to program doesn't make it less costly. If you had to take an 8080 or z80 and wanted to design (Note, design, not fab) a basic system that you could get on to writing applications for, you can get that working relatively easy as the systems were simple. Look at current i3/i5/i7 chips. The cost for designing a system to run applications skyrockets exponentially. The amount of complexity for designing the hardware is much higher and requires more planning and development time. Once you get that, you can't just easily plop libraries in and code willy-nilly. You have a complex system and with that comes the need for a complex underlying based operating system. You have to deal with a lot more hardware. You will spend a LOT of time working on designing and testing your system because the complexity is so much higher.

"I still do believe that more powerful hardware can reduce R&D due to the fact that it is easier to program " How could designing a new calc from near-scratch save you money vs. replacing the LCD + driver and modifying the existing OS?

And Peter Balyta isn't marketing anymore, he is the President of TI Education.
  
Register to Join the Conversation
Have your own thoughts to add to this or any other topic? Want to ask a question, offer a suggestion, share your own programs and projects, upload a file to the file archives, get help with calculator and computer programming, or simply chat with like-minded coders and tech and calculator enthusiasts via the site-wide AJAX SAX widget? Registration for a free Cemetech account only takes a minute.

» Go to Registration page
Page 4 of 4
» All times are UTC - 5 Hours
 
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
You cannot vote in polls in this forum

 

Advertisement