I remembered something from the VIA C3 Samuel 2 Datasheet, Here are some quotes from that document:

VIA C3 Samuel 2 (1,6 Volt; 700..800MHz) wrote:

The bus-clock multiplier is hardwired into each VIA C3 Samuel 2 processor chip. ... Actually, it is semi-frozen. A VIA-unique ma […]
Show full quote

The bus-clock multiplier is hardwired into each VIA C3 Samuel 2 processor chip.
...
Actually, it is semi-frozen. A VIA-unique machine specific register allows programming to temporarily change the hard-wired multiplier. This capability, documented in Appendix A, is intended for special BIOS situations and test and debug usage.
...
BUS CONTROL REGISTER 2: Certain bits in BCR2 register are used to program the processor’s clock multiplier.
...
BIT 19: ESOFTBF - Enables Software Clock Multiplier
BIT 26:23: CLOCKMUL - Software Clock Multiplier

It is a bit like the AMD K6-III+ and K6-2+ software multiplier. Potentially very practical, but only when there is a DOS tool to configure this.

Gona wrote:

On Socket 7 in DOS ALI Aladdin V is faster in low resolutions (like 3DBENCH) but VIA MVP3 is faster than Aladdin V in higher resolutions with all CPUs.

K6-III+ and K6-2+ (6x86MX too?) support write combining tof the LFB (MTRRLFBE), which makes all the difference in Vesa high res speed, together with the FSB speed that is.

Would be nice a DOS util to change the locked multier of the Samuel 2.

Just found two Windows programs that allow C3 Multiplier Control:
1) WCPUID by H.ODA. wcpu330.exe (Main site down, search for mirrors)
2) http://crystalmark.info - Crystal CPUID

Great finding!
Maybe Ezra and Nehemiah cores use the same control registers for multiplier.
A Nehemiah 1.2GHz would be good for testing but I have none of them. I have archived these two programs (maybe later I get one later VIA CPU).

Gona wrote:

I was amazed too. The C5A 650 MHz by DOS performance is somewhere between the MMX 200 and the MMX 233 MHz. Unfortunately I have no WinChip 2 to compare them but I would be curious. 😀

According to the 686 benchmark comparison, the Cyrix C3 Samuel, C5A at 550 MHz was about as good as a Pentium P55C at 262 MHz for FPU performance. The C5A-550 scored about the same as a P55C-250 in Quake II. 550 Mhz was the slowest I tested the C5A. 400 Mhz was the slowest I tested a Nehemiah. The 400 MHz Nehemiah with a 66 Mhz bus easily beat the Samual C5A at 550 MHz with a 100 MHz bus in all FPU benchmarks.

Gona wrote:

I have tested three BX board: Abit AB-BX6: not work at all; something Lucky Star: Boot as 550MHz but writes Celeron; ASUS P3B-F: […]
Show full quote

I have tested three BX board:
Abit AB-BX6: not work at all;
something Lucky Star: Boot as 550MHz but writes Celeron;
ASUS P3B-F: Not boot with jumperless, jumpered to FSB 100 will boot as 550MHz whatever is the multiplier.

I have found an other slotket (Socket 370 to Slot 1 adaptor): "AA370TS" like this (Click on the pictures to see larger version):

Image made by 5u3 Thanks! 😀
So this slotket have jumper settings for VIA CPU and I have try it with jumpered to VIA and FSB 100 with my VIA Samuel (C5A) 650MHz on the ASUS P3B-F.
Perfect! It boot as 650MHz. I have tested with system info prorgam and it is true 6.5x100MHz. 😀