What is important is that
the Output is V=5.0  or V=5.1
with at least 0.5 Amp, better is more e.g. 2 A

Meaning, explicitly, I can safely use that U.S. cell phone charger I have ?

Reason I ask is that lost of old gaming consoles were destroyed by using a wrong power adapter and while V4 is certainly built using modern-era components, a wrong voltage/amperage can always kill a board...

As general coding advices and guidelines we can give:
1) Be not afraid to use long instructions.
MOVE.L #123456,D0 is as fast as MOVEQ #1,D0
2) Ideally mix independant instructions.
ADDQ.L #1,D0
ADDQ.L #1,D1
ADDQ.L #1,D0
ADDQ.L #1,D1
ADDQ.L #1,D0
ADDQ.L #1,D1

is faster than

ADDQ.L #1,D0
ADDQ.L #1,D0
ADDQ.L #1,D0
ADDQ.L #1,D1
ADDQ.L #1,D1
ADDQ.L #1,D1

It's connected to my 32" LCD that I used for Jaguar. I need to buy a bigger table - I can't fit anything bigger than 32 next to the primary 50" TV that I use as a monitor.

I was worried how would I boot it up after seeing so many issues from other people, but I guess you guys configured some default resolution that works at most TVs ? My TV says it's 576p, either way I am very happy I didn't have to troubleshoot it remotely :)

I have this warm and fuzzy feeling now, playing with various demos, that I didn't think I would experience ever again :)

 

Gunnar von Boehn wrote:

Vlad you do code this in ASM right?

Technically, it's my own language - Higgs - but basically - yeah - assembler. For example, all the floating point code I was writing recently had to be strict ASM as I don't yet have support for it in Higgs.

As long as the output is 5V and at least 2A its fine. Voltage is the 'force' and the A is the capacity, 2000mA = 2A.
Your cellphone plug has the exact same output as the supplied V4 plug you have listed.
You can even use a higher current rating greater than 2A and the V4 will only take what it needs, its really just the voltage that must be correct.

I tried connecting both using the CAT5 cable (no switch, just directly from V4 into PC) and playing with Workbench->Network and after clicking OnLine, I get the caffeine output with :
Io0: flags = 9 (UP,LOOPBACK> mtv 1536
      inet 127.0.0.1 netmask 0xff00000000
net0: flags=822 (BROADCAST,NOTRAILERS,SIMPLEX)
      inet 0.0.0.0    netmask 0xffffffff00  broadcast 0.0.0.255

My PC only sees it as an Unidentified network, but can't seem to access the properties when clicking on LocalArea Connection inside Network and Sharing Center.

Any pointers on how to network them together ? Maybe at least doing some kind of local FTP server ?

Sure! But I'm not talking about the intrinsic power of the respectives CPUs but on the speed of the gcc-generated code on the vamp when targeting various CPUs.
 
Experiments indicates that code generated by c-compilers for 030 do perform better on the vamp than when generated for 040. I've been observing this on quake with various gcc versions as well as other compilers and recently Arczi on reported the same to me with latest gcc with diablo's support libraries.
 
See for instance the following excerpt of my old quake benchmark where asm optims was to quite present (so it is almost pure C):

quake.sasc.040 1.01.build#1589 (08/08/2018)   969 frames 41.92 seconds 23.12 fps     quake.sasc.030 1.01.build#1588 (08/08/2018)   969 frames 41.84 seconds 23.16 fps     quake.gcc-6.4.1b.040 1.01.build#1583 (08/08/2018)   969 frames 40.12 seconds 24.15 fps     quake.gcc-3.2.2.040 1.01.build#1595 (08/09/2018)   969 frames 40.00 seconds 24.22 fps     quake.gcc-2.95.3.040 1.01.build#1592 (08/08/2018)   969 frames 39.93 seconds 24.27 fps     quake.gcc-3.2.2.030 1.01.build#1594 (08/09/2018)   969 frames 39.69 seconds 24.41 fps     quake.gcc-6.4.1b.030 1.01.build#1582 (08/08/2018)   969 frames 39.68 seconds 24.42 fps     quake.vbcc.040 1.01.build#1586 (08/08/2018)   969 frames 39.59 seconds 24.48 fps     quake.gcc-2.95.3.030 1.01.build#1591 (08/08/2018)   969 frames 39.45 seconds 24.56 fps     quake.vbcc.030 1.01.build#1585 (08/08/2018)   969 frames 39.42 seconds 24.58 fps

The 040 version are always slower than the 030 for a given compiler.
 
Of course this could be some kind of gcc "bug", but then it seem to affect other independant compilers too. So there must be something that the vamp likes about 030 code generated by C compilers.
 
Or... maybe.. it is not the 030 generated-code which is pretty good, but the ones for 040 which is perticularily bad. I have a (semi) wild guess about this: at the time of gcc 2.x, amiga users (and compiler's authors/maintainers for the amiga) had better knowledge of widespread 68030 than of less frequent 68040. This kind of bias code generation in favor of the 030.
 
Now it is certain than this doesn't affect hand written asm. For instance with my latest quake, ASM opts are more present and the results are much higher (>30fps) but closer from one another because costly routines run the same vamp-optimized asm code.
 
So yeah, asm is better than C compilers yet.

0.3A is what the mains is supplying at whatever voltage. This is converted to the correct voltage and current available by the eu/v4 power supply. In general mains is transmitted at higher voltage and lower current as this results in less loss over the power lines. Once at the mains/ device it is then converted. Oversimplified you convert high voltage low current to low voltage, high current.

Is serial connection an option?