You should not compare them at the same clock rate, the max rated frequency of the 6502 is 1 MHz, and the max rated frequency of the Z80A is 4 MHz. The processors demand different clocks because their internal pipeline is structured differently, but the end result is similar levels of processing power at each their rated frequency. I’ll leave claims on who has the edge to others, as long as the basis for comparison is correct Smile.
Exactly. Clock speed alone does not make anything.
A similar issue was also valid for PC in the back where AMD processors ( '386 era) were faster than Intel counterparts, but this was partly due to the fact that they were able to ran at a higher speeds.
There are also a lot of considerations that one need to do. For example, dram refresh. on z80 you get it out of the box, on 6502 you need an extra circuitry to do this, impacting processor performances.
Not to mention other factors like the video circuit slowdowns if the memory is shared with cpu.
Performances should be measured in microseconds needed to perform a specific operation (for example a store) at maximum clock rate of each processor also taking in account ram speed and other factors, not by merely comparing clock speeds or cycles .
There are a lot of legendary factors on 6502. Some people claims that it is a RISC processor. But those are bullshits.
RISC means reduced instruction set computing. For reduced we mean reduced in "complexity" . not in the number of instructions on the set. this is a common mistake and the source of the RISC claim about 6502.
The real thing is that, both processor were CISC, and to be honest, hyronically , it is the 6502 that has the most CISC (complex) istructions like the pre-post indexed operations that z80 lack (it only had IX / IY) registers.
Everyone knows it was Yamaha that killed the MSX. Yamaha was always late in providing its VDPs.
I find it really hard to believe that Yamaha was the culprit. It may be the escape goat. If a company is always late in completing the chip designs it gets replaced. You cannot mess up two generations and expect to be awarded another design. I would question that if that was the case, choosing Yamaha again and again was the mistake. There is more to it.
Exactly i agree. And to be honest, Yamaha was allowed to evolve the MSX. remove all branded Yamaha chips from a msx and you get a MSX1 with a lot of limits and a very standard and quite limited architecture cloned from an already obsolete machine (the spectravideo/coleco) . Without Yamaha we can forget MSX2, MSX-MUSIC/Audio/Opl etc.
Then, my two cents about the story of nishi about V9938. He claimed that should have been on MSX1. To me it's a liar.
The real thing is that he too late realized that the most weakest part of MSX was the TMS and told this story because of obvious reasons. the real thing is that msx was built by borrowing an architecture from a reference hw (the spectravideo) without worrying to much if this architecture was already obsolete.
Then, my two cents about the story of nishi about V9938. He claimed that should have been on MSX1. To me it's a liar.
The real thing is that he too late realized that the most weakest part of MSX was the TMS and told this story because of obvious reasons. the real thing is that msx was built by borrowing an architecture from a reference hw (the spectravideo) without worrying to much if this architecture was already obsolete.
That's what you say. One does not preclude the other since the V9938 is based on the TMS.
In addition, the detail that tends to come up that Nishi doesn't lie is that the SCREEN7/8 seems to have an added layer. Then a v9938 with 64kB didn't make sense anymore at the time of the MSX2. I'm sure this VDP was initially planned with 64 kB but the rest was added for the MSX2. They couldn't add more because the interval between the MSX1 and MSX2 was short.
PS: I remind you that the MSX-Music was added to the standard because Yamaha did not want to lower the price of the MSX-Audio chips.
I think the magic Y-values is a pretty cool feature
On the TMS9918 Y=208 is not so much an issue although perhaps an unnecessary annoyance since you can’t just place off-screen sprites just anywhere in the border area (if may need a check). On the V9938 it is more problematic since the sprites scroll along with the display offset, so when the screen is scrolled halfway down sprites with Y=208 (or Y=216 in 212 line mode) are in the middle of the visible area, causing a "forbidden line" that needs to be checked and causes a small jump in the sprite position after correction.
I see, I didn't know the sprites move along with the background when scrolling vertically on the V9938/V9958, that's like the opposite from all other video hardware that I know of where sprites are entirely unaffected by scrolling. Kinda feels opposing to the philosophy of the TMS nomenclature regarding "sprites", where they are thought of as spirits that moves around in their own world unaffected by the background layer.
It's also kinda weird that V9958 horizontal scrolling doesn't follow the same pattern as vertical scrolling.
Clock speed alone does not make anything.
Didn't say it does, and I didn't mean it that way you understood it. I said the Z80 needs to be clocked faster to be comparable, but I was probably wrong in saying that is for compensating a slower MPU, when it's more like part of how the two architectures work as Grauw said. I could swear that the 6502 is typically considered faster in general (clocked at their typical speeds), but I can't find any sources that backs such a claim.
Both architectures are definitely CISC, RISC would be more like the classic 8-bit PIC or ARM, MIPS, RISC-V etc, so yeah simplified more bare-bones instructions that must be combined to do more complex tasks is what defines it if I understood it right.
That's what you say. One does not preclude the other since the V9938 is based on the TMS.
Nishi has already proved us to be a non reliable man. he have changed his mind too often.
Additionally i do not think making a relatively very simple vdp with only the useful things would have been so time expensive for yamaha. (like the SMS one, that is a relatively simple and improved VDP based on TMS. Why did not happen on msx? ).
It is extremely strange that if i need a video chip and challenge yamaha to do this and then yamaha delayed to much there are no rumours around. I never heard about a complain of being too late due to yamaha in the eighties.
The same apply to sega SMS. I've never heard about sega issued yamaha because they delayed too much the SMS vdp.
In my opinion there was no issue because there was simply no delay.
It would be interesting to ask some yamaha engineer involved on V9938 to see its point of view. I will not surprised at all if the story told to us is not the same that nishi say.
additionally, Yamaha proved to do well chips with the v9990. but the v9990 had no compatibility bottleneck around,
what IMHO slowed down the development of v9938 a lot.
In addition, the detail that tends to come up that Nishi doesn't lie is that the SCREEN7/8 seems to have an added layer. Then a v9938 with 64kB didn't make sense anymore at the time of the MSX2. I'm sure this VDP was initially planned with 64 kB but the rest was added for the MSX2. They couldn't add more because the interval between the MSX1 and MSX2 was short.
Even with only 64K of VRAM MSX2 would have been a decent machine. V9938 issues are others.
Developers complain about sprite engine, to much limited, and not well designed.
They also complain about VDP blitter speed and some other issues.
Having a 512x212@16 color or a 256 color mode is not what developers wanted on a 8 bit machine of '85.
they wanted decent sprites, full decent hw support for scrolling and decent color-clash free tiled mode.
(Curiously is what the SMS VDP had? ;-) )
Additionally, competitors had most 16 colors on screen with at least a 320x200 pixels resolution, often with limitations. So MSX2 would have been better even with 64K
maybe the screen 7/8 was added during development as another "BRILLIANT" nishi idea who had always wanted high resolution and color features and failed to see the most natural target of the machine (games)
PS: I remind you that the MSX-Music was added to the standard because Yamaha did not want to lower the price of the MSX-Audio chips.
I remind you that without yamaha sound chips, msx would have been with only the GI chip, barely enough...
Both architectures are definitely CISC, RISC would be more like the classic 8-bit PIC or ARM, MIPS, RISC-V etc, so yeah simplified more bare-bones instructions that must be combined to do more complex tasks is what defines it if I understood it right.
You are understanding well.
But the same people that say that 6502 is faster than a z80 also typically say that 6502 is RISC. Myths.
About sources that prove the fact that, at usual speed the z80 is more faster than a 6502, there are some around.
Most games developed on Spectrum (z80) and later ported to 6502 were almost rewritten from scracth due to impossibility to do a lighter convertion because of so different architectures. So games were written from scratch on 6502 machine code.
Nevertheless, games that relied only on cpu on c64, without the ability to use hw specific features, even optimized, ran 30% or more slower than on a ZX Spectrum.
Honestly it’s a bit of a shame to reject the points where MSX2/2+ excels the most among other 8-bit computers, that is in graphical fidelity (resolution, colours). And perhaps in processing speed too if you consider the turboR. In these aspects, MSX has qualities that are only matched by 16-bit systems.
Secondly, among some facts there is also a lot of opinion, speculation, and urban legends. Honestly what would be really awesome if some Japanese documentary maker made a deep dive documentary series about MSX, with interviews with all involved parties (not just Mr. Nishi). Engineers and managers from ASCII, Yamaha, Panasonic, Konami, etc. That would give an extremely interesting insight in the business world that created MSX, something which is quite obscured to us because back than Japan was quite closed due to communication barriers like distance, language and culture, and this applies to the Japanese business even more so.
Honestly what would be really awesome if some Japanese documentary maker made a deep dive documentary series about MSX, with interviews with all involved parties (not just Mr. Nishi). Engineers and managers from ASCII, Yamaha, Panasonic, Konami, etc.
That would be phenomenal.
Another clue that suggests that the TMS should not have been in the MSX1: It is the only component that is not made in Japan.
Now Mister Nishi announces the MSX0 before the MSX3 (it's not clear for the average user). And also the download system (Project Egg) very well.
On the other hand, the crowdfunding system is negative. I want to buy the complete Hardware MSX3 (MSX0) and all the Hardware that will be released in 2023 and after but not by paying weeks or months in advance (like MSXVR). If he hasn't found a Japanese manufacturer, it's a bad start. I fully trust Mr. Nishi but I sometimes fear a gap between theory and practice (sale of new hardware, complete system, download of old software, new games, etc.)
fingers crossed and so excited to buy new MSX Hardware in 2023 