Stiati ca XP merge pe 386??

[eddy]

XP minimum requirement test This text in English!

A goal of the thing is it to let XP on as weak system run as possible.
For memory, Microsoft demands a processor with 233 MHz and 64 MT RAM.
That might have to be struck nevertheless.

First attempt: Base 5, Pentium 100 MHz & 64 MT RAM
Experimental setup 1:

* Gateway of 2000 bases 5 board
* Pentium 75 MHz over-clocks on 100 MHz
* 64 MT RAM, later > 32 MT
* SCSI non removable disk and CD Rome
* Winner 1000 diagram map (ISA)

The Setup goes problem-free, it lasts only somewhat longer, to Windows finds a Pentium, it makes thereby no difference
whether it is a Pentium I, II, III or IV. The 64 MT RAM are with the Setup MINIMUM. Otherwise the XP breaks simply
off. Also installing and starting on a 486er are not possible. It was tried naturally.

Naja after the Setup is XP no longer completely so fastidiously, one can the RAM disarm. With 16 MT
RAM does not start Windows however any longer. It must be already somewhat more, this shows the second test.

Achja to the CCU, after the installation was not only disarmed the RAM, the CCU became also again in the original clock
operated. The 16 colors come therefore the XP no more drivers for ISA diagram maps make available.

[eddy]

First result: Base 5, Pentium 75 MHz & 24 MT RAM

Second attempt: Base 3, Pentium overdrive 83 MHz & 64 MT RAM
Experimental setup 2:

* Arstoria AS496 Base 3, SiS 496/97 chip set
* Pentium overdrive @ 83Mhz and 63M H z
* 64Mb RAM, end the conclusion 18M B
* IDE non removable disk 6.4G B and CD Rome
* S3 Trio32 P CI diagram map

In the course of a renewed attempt to install XP on a 486er system it succeeded to Antz Windows XP at a base 3
To install Main board. However with the Pentium overdrive processor. This miracle of the technology is adequate
Pentium core, with 32 bits a memory CONTROLLER the extra for the 486er base was designed.

With this board it was also not possible a 486er to then use (does not load simply), but it became in
new processor and memory record set up.

[eddy]

The CCU has a clock composition of 33 MHz system bus with a multiplicator of 2,5 (2.5 x 33 = 83 MHz),
Antz has simply the system bus on niedrigst possible placed: 25 MHz. as a result of 2,5 x arises 25 MHz
then the new depth record of 63 MHz (XP rounds off, therefore 62 MHz).

[eddy]

The memory record thereby reached which one the 32-Bit-Speicher-Controllers of the overdrives use made itself:
A simple Edo memory latch plate has 32 bits an interface, the normal Pentium processor has however 64 bits an interface,
so that one runs always 2 latch plates necessarily thereby it. The Pentium overdrive for 486 bases can do against it directly one
individual memory latch plate respond.

Thus came to 24 MT (2x8 MT + 2x4 MT Edo memory), came 20 MT (2x8 MT + 2x2 MT Edo memory) and
finally 18 MT (2x8 MT + 1x2 MT Edo memory). By the way XP shows the 18 MT as 20 MT on...

[eddy]

in the task manager one experiences however the true Speichergroe ss e

[eddy]

Second result: Base 3, Pentium overdrive 63 MHz & 18 MT RAM

Supplement:

Antz is to be lowered it succeeded the clock further, one must in addition simply the exhaust of the CCU
down-take. The exhaust is supplied via the CCU with river, and so this notices even if
it is missing. As consequence as a result of it a simple multiplicator arises. As result one has then
a clock of 1x25 MHz, instead of 2,5x25 MHz. Starting time of the system amounts to minutes.

[eddy]

Third attempt: Base 4, Pentium 60 MHz & 64 MT RAM
Experimental setup:

* Elitegroup SI5PI AIO (REV. 1,1) base 4
(installation: Base 5 board from test 1)
* Pentium 60 MHz
* 64 MT RAM, later 32 MT
* SCSI non removable disk and IDE CD Rome
* Voodoo Rush PCI diagram map

Around a deeper clock than Antz to reach I use the first Pentiumprozessor, the Pentium 60 MHz. It runs along directly
the Frontsidebus. It has thus a multiplicator of unity.

[eddy]

The CCU, a normal Pentium 60 MHz, without division errors, for the base 4

To the beginning there is a bad surprise, XP falls with the Setup. Each mark! It could be because of the SIS chip set, I has it
does not herausgekriegt. XP worth itself against all attempts.

Thus which do? Simply times take a board ( as in the Windows 98 on 386er attempt ) which goes. Thus base 5 taken
on attempt 1. Whole history with two-way board change (of the base 4 to base 5 and back) has then over 4 hours
lasted. XP needed to start then elendig for a long time around first times on the Pentium 60, since the hardware was again recognized,
but it went. No notion which the Setup wanted...

[eddy]

Why I hung now to this board (and not a Intel board took, which I also have) am completely simple. It leaves itself
quite simply jumpern. For this the Jumper S0, S1 and S2 is responsible, which one can do in addition, umjumpern, e.g. needed that
Attempt:
S0 (J7) S1 (J6) S2 (J5)
2-3 1-2 1-2

25 MHz!

[eddy]

The Jumperblock, the Jumper 5.6.7 regulate the Systemtakt=CPU clock. Here 25 MHz attitude.

Besides I had found also still 50 and 33 MHz. Naturally the memory was again halved, on 32 MT. Less
was not possible unfortunately, since the board supports only 4 memory rows. I had unfortunately only double-sided 8 MT of modules, and
would be thereby under the memory border please (16 MT, there further memory not recognized would be).

[eddy]

If one lets the Desktop without programs run simply, so that the CCU carry nothing great out, speak to load must, (IDLE),
one by the way receives the following picture:

[eddy]

A view times into the equipment manager. The system equipment list is not nevertheless quite short, the Rush has
Driver and the CCU are missing somehow.

[eddy]

But I do not give myself yet struck, it am continued to try:
S0 (J7) S1 (J6) S2 (J5)
1-2 1-2 1-2

20 MHz!!!

[eddy]

Astonishment with boats, 20 MHz!

This clock is not times possible on most base 3 board, but on this board. To which also always goes, that
is gives rumors around a Pentium 50, but the officially lowest is stop of the 60ziger. Possibly it stands in the connection,
with the turbo-switch (which also always may cause (that is not fixed))?

The system start (thus end of BIOS start to the Desktop with symbols is developed) takes nearly 7 minutes. But to then Windows
reacted it lasts still times a certain time. I would like to remind also still times it a rather naked system am:
No network, no sound, no USB....

[eddy]

The CCU is working at full capacity even in the no-load operation to 2/3. There one must run with today's CPUs a quantity of applications of backgrounds
let in addition-get around.

[eddy]

Accordingly one needs much patience for all patience. Is them for example of the color changes of the half dozen
Blue stages BEFORE the welcome side been noticeable? No? If each individual color from above downward constructs itself and
in some seconds of goes, then one notices it...

Also each window develops itself second-long elementenweise:

[eddy]

By the way Windows XP with 20 MHz runs substantially more badly than with 25 MHz. With 20 MHz everything lasts around some longer, and
also the cold-started does not function always...

Third result: Base 4, Pentium 20 MHz & 32 MT RAM

Bench mark to the attempt 3 system:
Pentium 20 MHz
(Win XP) Pentium 25 MHz
(Win XP) 486 SL 25 MHz
(Win 95)


The system
Norton SI 4,0, DOS 63,4 79,2 53,8
SiSoft Sandra 2001 Dreystone (Windows) 43 MIPS 54 MIPS not possible
(Win 95)
SiSoft Sandra 2001 Whetstone (Windows) 22 MFLOPS 21 MFLOPS not possible
(Win 95)
DrHardware 5.5
Hard clay/tone (DOS) 16743 20998 11088
DrHardware 5.5
Soft clay/tone (DOS) 7287 9142 4902
PrimJNold
(Windows) 393.60 seconds 305.13 seconds. 311.40 seconds.

[eddy]

Copy Right tradus de mine imi cer scuze moderatorilor dar vroiam sa iasa frumos in plus pacat ca s-a bagat maimuta aceea peste mine mertia fiecare minut pierdut pe acest topic

PS : Scuze AIWA stiu ca sunt o pacoste dar acest topic chear merita ... sper ca sanctiunile sa nu fie prea drastice

[PigBrother]

Hai ca a ajuns unde trebuie. Am indraznit sa-ti corectez titlul threadului... "stiati" FFS
Amuzant si interesant, dar puteai sa le comasezi in 2-3 posturi.
In fine...funny shit

[Johnny]

Right off the bat: ce e cu engleza asta de galeata uitata sub niste chirpici in Vrancea? Mi s-a facut rau incercind sa parcurg complet toata saga. Boo!