an article about z80 affecting MSX computers was published

Page 5/7
1 | 2 | 3 | 4 | | 6 | 7

By erpirao

Paladin (872)

erpirao's picture

16-12-2018, 13:21

I think the R800 is a modified Z800, with many mutilated / trimmed features, to save costs.
is this link and after doing a test to the cpu280 the z280 @ 12 is compared with a 68k @ 10.
Also, to put the R800 in perspective, your Ram is mounted @ 90ns (approx 11mhz), which gives it a significant performance advantage.

By mars2000you

Enlighted (5467)

mars2000you's picture

16-12-2018, 13:23

erpirao wrote:

bluemsx+msxturboR emulation:
using clkspeed.com
effective clockspeed: 17897700hz
standar (optional) : 3579545hz
relative performance: 499.9%

Maybe you should also check with openMSX and Turbo R emulation. Some other discussions suggest that the Turbo R 'average' speed in blueMSX is slightly faster than on the real machine while it is slightly slower in openMSX compared to the real machine. So the truth should be in the middle!

By Grauw

Enlighted (7967)

Grauw's picture

16-12-2018, 14:21

litwr wrote:

@Grauw You have almost convinced me that R800 is just simplified Z800 but the other MSX guys say that it can't be true...

Well we don’t know all that much about the R800, that’s the thing. And there’s some things about the R800 that can’t be explained yet (like the discrepancy in the duration of call + ret vs. call + nop + ret, the latter is two cycles longer). And realistically speaking, a CPU can’t “just” get the speed that it does, there must be some form of pipelining at least, and I’m wondering if it maybe has a higher internal clock speed as well. I find the R800 vs. Z800 / Z280 angle interesting and in need for more investigation. Especially because of all those hints at unused features in the R800 pinout.

mars2000you wrote:

Maybe you should also check with openMSX and Turbo R emulation. Some other discussions suggest that the Turbo R 'average' speed in blueMSX is slightly faster than on the real machine while it is slightly slower in openMSX compared to the real machine. So the truth should be in the middle!

Indeed openMSX is slightly too slow, so the emulation must not be perfect yet and there must be some details about the timing that are unknown. When I presented my call + ret finding to wouterv (or was it mth) a few years back he also wondered if it was something to do with some sort of pipelining.

By erpirao

Paladin (872)

erpirao's picture

16-12-2018, 14:54

mars2000you wrote:
erpirao wrote:

bluemsx+msxturboR emulation:
using clkspeed.com
effective clockspeed: 17897700hz
standar (optional) : 3579545hz
relative performance: 499.9%

Maybe you should also check with openMSX and Turbo R emulation. Some other discussions suggest that the Turbo R 'average' speed in blueMSX is slightly faster than on the real machine while it is slightly slower in openMSX compared to the real machine. So the truth should be in the middle!

your are rigth
width real a1gt:
using clkspeed.com
effective clockspeed: 17897700hz
standar (optional) : 3579545hz
relative performance: 499.9%

no change.

using login.com
cpu: z80280
mhz: 18.7536
almost 14mhz difference

By litwr

Rookie (30)

litwr's picture

21-12-2018, 15:21

So I have just added to the article

Quote:

There are opinions that the R800 is a simplified Z800, running at four times the frequency of the bus.

I also added

Quote:

Interestingly, Zilog products had special traps on chip for those who tried to make copies of them, for example, the base Z80 had 9 traps and they, according to reviews of those who did it, slowed down the copying process for almost a year.

So it is really interesting, have products like Z800, HD64180, Z280, Z180, Z380, ... the similar traps?

By erpirao

Paladin (872)

erpirao's picture

21-12-2018, 16:45

Looking back at the old magazines, I found this in the last Hnostar:
I'm sorry that it's in Spanish


It is very striking that the improvement of the R800 on the Z380 is not much higher.

By zett

Champion (511)

zett's picture

21-12-2018, 20:11

if you want to make a future msx you best gona work on the stuff of lpe z380

By erpirao

Paladin (872)

erpirao's picture

21-12-2018, 20:37

zett wrote:

if you want to make a future msx you best gona work on the stuff of lpe z380

A few months ago I was talking with Leonardo Padial about the z380, he explained that there was a beta tester that made a demo sending the calculations to the z380 and then the z80 only reflected the results were apparently very impressive.
to see if I can borrow a z380 and find that software.

By zett

Champion (511)

zett's picture

21-12-2018, 22:18

erpirao wrote:
zett wrote:

if you want to make a future msx you best gona work on the stuff of lpe z380

A few months ago I was talking with Leonardo Padial about the z380, he explained that there was a beta tester that made a demo sending the calculations to the z380 and then the z80 only reflected the results were apparently very impressive.
to see if I can borrow a z380 and find that software.

nice. yeah that padial made some nice hardware

By litwr

Rookie (30)

litwr's picture

22-12-2018, 07:18

erpirao wrote:

Looking back at the old magazines, I found this in the last Hnostar:
It is very striking that the improvement of the R800 on the Z380 is not much higher.

Thank you very much. Benchmarks for the Z8000, Z800/Z280, Z380 are quite rare. Sad They show that the R800 is definitely a derivative of the Z800 and it can be even slower than the Z80 at the same frequency. Shocked! Those tests missed 16-bit arithmetic. Sad
It is interesting what was a Z380 based system used for tests in the benchmark? What was a year of publishing?

Page 5/7
1 | 2 | 3 | 4 | | 6 | 7
My MSX profile