@OlafS25

Quote:

compared to modern hardware both do not look very good.

I don´t think any Amiga hardware is comparable to mainstream computers for last 10 years... What is real competition for Tabor in our small market?

@pavlor

What I would say is that there is a threshold of comfort with the AmigaNG powerpc systems.
Once they preform the task at a speed that is comfortable to the user then there is no problem.
Or If the don't then you can adjust the task. For example like re encoding videos from 480p mp4 to 480p xvid. Takes only a few minutes on my Windows 7 machine and plays back beautifully on my Sam 460ex.

Quote:

pavlor wrote: @OlafS25 Quote: compared to modern hardware both do not look very good. I don´t think any Amiga hardware is comparable to mainstream computers for last 10 years... What is real competition for Tabor in our small market?

Last edited by Spectre660 on 08-Nov-2015 at 07:26 PM.

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@Spectre660

iPad Air (1.4GHz A7 / "Cyclone" CPU) iOS 9.2

Safari:

Score: 8701
Richards: 12462
DeltaBlue: 15399
Crypto: 11813
RayTrace: 22644
EarleyBoyer: 18185
RegExp: 1145
Splay: 4281
NavierStokes: 7178

Opera:

Score: 473
Richards: 276
DeltaBlue: 280
Crypto: 399
RayTrace: 742
EarleyBoyer: 1397
RegExp: 164
Splay: 1089
NavierStokes: 437

Chrome:

Score: 461
Richards: 263
DeltaBlue: 259
Crypto: 402
RayTrace: 706
EarleyBoyer: 1347
RegExp: 166
Splay: 1087
NavierStokes: 440

I suppose Safari uses JIT and the others not...

But even the non-JIT browsers are 3x - 10x faster than the Tabor.

And 16x faster than the Sam.

The iPad is 2 years old (2013). After that followed the A8 (said to be 20% faster than its predecessor) and after that the A9 (said to be 70-80% faster than its predecessor). iPad/iPhone of course runs completely battery powered and passively cooled.

The Tabor will be sold as a desktop on the 2016-2018 markets.

Think about it...

Last edited by TRIPOS on 08-Nov-2015 at 08:26 PM.

Quote:

ribdevil wrote: I have an AMIGA, the other it's a PC

But what if your Amiga would have been based on a Core i5/i7?



Quote:

pavlor wrote: Quote: compared to modern hardware both do not look very good. I don´t think any Amiga hardware is comparable to mainstream computers for last 10 years...

But wouldn't it be nice if that wouldn't be the case? If Amiga H/W wouldn't be so embarrasing backwards and ridiculously overprised?



Go for x86 already!

@TRIPOS

I think a fair comparison would be tabor vs raspberry pi 2, ie low end hobby grade hardware that's not cranked out in millions of units. Lots of people use pi 2s for "desktop" type tasks like media center boxes.

Anyone here with a pi care to run the same tests and post the results?

Last edited by CodeSmith on 08-Nov-2015 at 08:44 PM.

@TRIPOS

Quote:

But what if your Amiga would have been based on a Core i5/i7?

There was Core i7 based Amiga... Amiga Mini from Commodore USA.

Quote:

Go for x86 already!

Exactly my words!

@CodeSmith

Go for ARM then!

@TRIPOS

Still waiting from my Ipod touch G1 to finish the benchmarks.

Richards at 25.90 .
I think that it cost in the region of US$400.00 in its day .

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@CodeSmith

Quote:

I think a fair comparison would be tabor vs raspberry pi 2

FireBee, ArcMini or other boards targeting similar markets as ours would be better for comparison.

Tabor has decent features for our needs, only real use will show, if its known weak points are prohibitive. As we wrote numerous times in this very thread, final price will be most important factor.

@TRIPOS

Quote:

The Tabor will be sold as a desktop on the 2016-2018 markets.

Quote:

Think about it...

Do you get people buy boards for OS4 with even worser price/performance ratio than Tabor? Yes, it is insane, but even with such burden these computers stay competitive (in number of users) to dirty cheap second hand hardware (for MorphOS) or powerful and cheap PCs (for AROS). Even with facts and numbers, you can´t beat power of Amiga name.

@Spectre660

G1 ipod touch cant do the benchmark. Stuck after Richards.
Apple you Swine!!!!!

Quote:

Spectre660 wrote: @TRIPOS Still waiting from my Ipod touch G1 to finish the benchmarks. Richards at 25.90 . I think that it cost in the region of US$400.00 in its day .

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

What could be perfomance advantage of AOS4 if Tabor was based on T1022 or T1042 instead of P1022?

Is T series much faster than P series?

I mean considering that AOS4 does not have support for 64-bit or multicore. What will be pure e500v2 core vs e5500 core result?

Last edited by dooz on 08-Nov-2015 at 09:30 PM.

@dooz

Quote:

What could be perfomance advantage of AOS4 if Tabor was based on T1022 or T1042 instead of P1022?

Unless you know someone that has a T1040RDB or a T1042RDB-PA, that is a moot question (and those cost $1500).
Besides, that is just self flagellation.

@dooz

Quote:

I mean considering that AOS4 does not have support for 64-bit or multicore. What will be pure e500v2 core vs e5500 core result?

e500v2 vs e5500:
2.3 DMIPS/MHz vs 3.0 DMIPS MHz
slow/uncompatible FPU vs normal FPU
1200 MHz vs 1400 MHz

@iggy

Quote:

iggy wrote: @dooz Quote: What could be perfomance advantage of AOS4 if Tabor was based on T1022 or T1042 instead of P1022? Unless you know someone that has a T1040RDB or a T1042RDB-PA, that is a moot question (and those cost $1500). Besides, that is just self flagellation.

OK, only data from CPU specification is available. I can read that e500v2 is 2.4 MIPS per MHz, and e5500 is 2.9 MIPS per MHz. Also T1022/1042 can be clocked at 1.4 MHz, while P1022 maximum is 1.2 GHz.

That is 2880 MIPS for P1022 vs 4060 MIPS for T1022!

Also T has normal FPU.

I dont know what system bus and memory speed numbers are, but I guess that P1022 has faster bus. I would like to know that since it can have significant impact on overall system perfomance

But from MIPS data and clock in MHz T1022 can be 30% faster than P1022. MIPS can be compared if CPU is from the same familiy.

Personally I would like to have T1022/T1042, but in the same moment I am not really impressed with only 30% faster CPU

@dooz

Freescale lists various DMIPS/MHz values for its products (2.3 and 2.4 for e500, 2.9 and 3.0 for e5500 etc.). You may also compare Coremark scores:

P1022 1.2 GHz: 3720 (single core)
T1023 1.4 GHz: 5180 (single core)

@pavlor

Quote:

pavlor wrote: @dooz Quote: I mean considering that AOS4 does not have support for 64-bit or multicore. What will be pure e500v2 core vs e5500 core result? e500v2 vs e5500: 2.3 DMIPS/MHz vs 3.0 DMIPS MHz slow/uncompatible FPU vs normal FPU 1200 MHz vs 1400 MHz

May I add that if board is designed for T1022 from the begining then also faster versions can be easily produced since several CPU types from that family are pin compatible (just found that information)!

T1022: 2 core e5500 1.4 GHz
T1042: 4 core e5500 1.4 GHz
T2081: 4 core e6500 1.8 Ghz

I guess you can then just make several versions of the same motherboard with different CPU? Even e6500 is possible?

I guess that will be the next 12xx project from A-Eon.


Q

@pavlor

it will everything depend on the price

@CodeSmith

Raspberry is unbeatable cheap, I think 30 EUR for the board. It is a toy but a very cheap toy. The PPC devices are much more expensive and in range of well equipped good ARM or Intel/AMD based systems (not to talk about used systems).

@pavlor

from the comments I read from Fab the chance for getting the JIT on PPC are pretty small so the existing difference in performance will stay (not to talk about the hardware). For me PPC based systems in 1 Ghz range have no chance.