this page has not yet been verified !! 

 Please check and read 021005-1, 021005-2, 021005-3  for OS/2 Warp 3
 

 980810     Cyrix 5x86 

  Continued from #980809 Windows 98 and PS/55.

  There are few variations in Cyrix 5x86 CPU.
     1)  5x86 100GP  3x
     2)  5x86 100GP  4x  ( 25MHz x4 )
     3)  5x86 120GP  3x  ( basicaly 40MHz x3. Pre-production sample of 30Mhz x4 version rated 120GP
                                        were once sold in the market )
     4)  5x86 133       4x  ( 33MHz x4 )
     5)  IBM 5x86-100 3x
     6)  and others made by TI ( Texas Instruments ) and ST ( SGS Thomson ) . 

It's relatively easy to get  1),  3) and 5) but it's hard to get  2) and 4) .   I  was very lucky that I got these 4x   models.  Among above 5 models,  4x models rated 100MHz  seemed be a different beast.  It even can  tolerate 133MHz  operation if it get cooled  well ( at least for a few hours.  I haven't ever run it in 133MHz for a day or more ).   When  I got 4x 133MHz model I really expected that it would be the best CPU among all of  486/586 class CPUs.   But it seemed  that 133MHz 4x model is a bit different from 100MHz 4x model.  In the same system and with same cache setting,  4x 100MHz version and 4x 133MHz version scored different result.  100MHz rated model gives higher performance result in general as if there are major difference in core design.
      Sorry I'm not an expert about this issue so I can't tell definitely yes or no )
      Added on 2002/10/06 : Sysbench  under OS/2 Warp connect,  performance result were almost
                                           same as if they are same products.   Really strange....

Benchmark programs usually reports  that Cyrix 5x86s are  slower than AMD am5x86-133 but if you run business software  you'll  get  better  performance with Cyrix 5x86 rather than with am5x86.   If you use an appropriate cache program to activate  L1 WB function of Cyrix CPU,  system performance will be boosted.
One of such program is et586.exe which contained in et9060.exe provided by  Evergreen.  et586.exe should be placed in your config.sys with appropriate option switches in HEX value to activate registers.  In order to determine HEX values you can use a nice EXL file which was written by Mr. F355, a guy who reported Clock up Method of Reply M.70 PowerBoard.  To use this program, you have to install analyzing tool using  TOOL>ADDIN>Analyzing Tool  in M$ Excel. 
 

   5551-R :  EIDE without SPOCK SCSI 
   5551-N :  SCSI, IML

   Parameter setting for ET586.EXE ; Install c:\ ETI\ET586.EXE /CCR2=56 /CCR4=1D /WBE
 

		HDBENCH			ETDIAG	RMKS
	Speed	Float  P.	Integer		Dhrystone
5551-R + Win98
Cyrix5x86-100   4x	133MHz	3877	6585		92592	Enabled
Cyrix5x86-100   4x	133MHz	3762	6102		45872	NOT loaded
Cyrix5x86-120   3x	99MHz	3178	4970		65789	Enabled
Cyrix5x85-120   3x	99MHz	3093	4635		45871	NOT  loaded
am5x86-133      4x	133MHz	3583	5759		65789	*3
DX4ODP100 35  x3	105MHz	2691	5373
PODP5V83 33  x2.5	88MHz	2565	4962
5551-N + W95
Cyrix5x86-120    3x	99MHz	3130	4690			not recorded
Cyrix5x86-120    3x	126MHz	3910	5929			not recorded
am5x86-133    42 x4	168MHz	3883	7028
5551-N + Win98
Cyrix5x86-120  *1	120MHz	3716	5575		53763	NOT loaded
Cyrix5x86-120  *1	120MHz	3704	5561		56818	Enabled
Cyrix5x86-100    4x	133MHz	3791	6162			not recorded
Am5x86-133      4x	166MHz	4200	6776		51020	*2

  *1   No remarkable improvement observed on 5551-N  with  or  without ET586.exe. 

  *2   As already reported in 980809, am5x86 and IBM SPOCK will run in MS-DOS compatibility mode
         under Windows98. 

  *3   Write-Back caching  enabled with PL586 interposer. 

et9603.exe       ;  contains  DOS, Win NT and  OS/2 drivers and associated programs
                              run the program with  /d switch ( or  -d whatever you like ) for directory sorting. 

  5x86rege.zip    ;  EXL file for calculating HEX values of et586 cache driver.  Written by Mr. F355.

  SGS4191.pdf   ;  ST5x86 Data sheet by SGS-Thomson. 

  5x86.exe          ;   PeterMoss's 5x86 Utility

Quoted comments regarding Cyrix CPU in the News Group

 Tim Clark wrote to the News Group on 2002/09/01.
   " IMHO, the Cyrix/IBM 5x86 is the better bet, as it has notionally better performance than even a
    PODP83. Especially if you can find a 4x chip for 33MHz core. This would approximate to a P100.

    > :>It's been my experience that the POD 'feels' faster with graphical
    > :>operating systems.  YMMV.
    > :>
    > :>> Also, could this thing run OS/2 Warp 4? Would you get good performance?
    > :>
    > :>It should, although the gurus will probably tell you that Warp 3 is a
    > :>better match for this machine.

     Quite possibly, the "page-tuning" of the Warp 3 "core", although probably lost to some extent by the 
     significant no. of FixPaks issued since, is very likely to give better memory performance on systems
     limited to 64MB max."

 Jim Sherney replied,
   " I will agree that the performance of the IBM-branded Cyrix 5x86 chips is somewhat astonishing. 
    Dunno about the Cyrix-branded ones, I've heard rumors that IBM tweaked the microcode in thier
    version.  Meaningless benchmarks on my P70 page show that the IBM 5x86-100 compares very
    favorably to the AMD 133, despite the lower core speed.  It resides in my P75 now.  I'd love to get my
    hands on the 4x version."

 Tim replied, 
   " It's a difficult question to determine at what point the Cyrix and IBM-labelled chips became the same.
    Initially, the M9/M1SC chip design was outsourced to various fabrication companies, IBM being only
    one (TI and ST also being in there). In a later deal, IBM took on the majority of the fabrication, with 
    one of those "we'd like to brand x% for our own use" deals. At this point there was some further 
    "production engineering" and development work done to increase yield and, I believe, shrink the die. 
    TI and ST continued to lease the design with the older spec. well after the IBM fabrication ended, as
    their core voltage requirements tended to be ~3.6V rather than the Cyrix/IBM 3.45V. I may have some
    of the details wrong, but  that's mostly correct, I think. Certainly, any Cyrix 5x86-nnn/4X will be a
    late IBM fab."