021005-3  Cyrix and Lacuna #3       Episode 3   Return of  The Jedi .... M.A.D ?.

In order to find out  Cyrix5x86 x4 100 model and  ACE486 interposer in stock condition, I opened every PS/55s and PS/2s.   Cyrix5x86 100GP x4 capable model was in 9585  together with a PGA DX4.  A stock ACE486 was in 8590/Type-1 25MHz complex with a Cyrix 120GP.   Now, it's time for testing Cyrix5x86 100GP/4.

 9577-BTG 33MHz planar
 Hard Drive ; EIDE Quantum Fireball ST, 2160AT
 CD-ROM  ; Toshiba XM3401BMA  2.1x speed off from IBM SCSI II ( OEM'ed FD MCS-700 )
 RAM         ;  8MB x4 
 Resolution; 800x600, 65K color depth 
 OS/2 Warp 3 Connect  NO FIX Pack applied yet.
 File system; FAT

 Test result of SysBench 0.9.4g 
 Upper five  lines are took from previous page for reference. 

	CPU      Integer	CPU     Float	Dive	PM	Disk I/O	CD I/O
DX4-100	44.159	4.809	10.225	21.411	15.999	17.304
Cyrix5x86 3x         *1	45.262	6.088	10.136	14.230	9.056	17.284
Cyrix5x86 4x ET   *3	51.821	7.756	10.140	14.270	9.355	17.291
Cyrix5x86 4x ET   *5	62.986	7.471	10.111	13.154	6.456	17.270
am5x86 4x  PL586 *7	65.945	5.492	10.249	21.834	17.263	17.312
5x86  120GP x3 model  on ACE486 ( x3 fixed interposer ) Designed for intel DX4-100
Cyrix5x86 3x ET   *8	51.241	5.853	10.140	14.337	9.410	17.290
Cyrix5x86 3x ET   *9	45.221	6.088	10.124	14.228	9.307	17.290
Cyrix5x86 3x ET   *10	50.724	5.930	10.133	14.323	9.344	17.290
Cyrix5x86 3x ET   *11
My favorite 5x86 100GP x4 model  on EverGreen586 interposer
Cyrix5x86 4x ET   *10	63.117	7.461	10.121	13.164	6.473	17.270
Cyrix5x86 4x ET   *11	51.924	7.789	10.171	14.270	9.164	17.286
Cyrix5x86 4x ET   *12	50.724	8.887	10.209	19.850	7.686	17.282
Cyrix5x86 4x ET   *13	63.214	8.602	10.222	19.828	7.690	17.282

  ET means cache driver et586OS2.sys is used 
  Interposer;   EverGreen 586 is used for  *1,  *3,  *5.  CB jumper on
                   PowerLeap PL586  is used for *7 .         WB jumper on.

  *1   Cache driver not used 
  *3   Cache driver; default setting ( /CCR4=1D  )
  *5   Cache driver;  /CCR2=56  /CCR4=1D  /WBE  ( worked fine on 5551-R and Win98 )
  *7   AMD 5x86/PGA on  PowerLeap PL586 on an 486 socket interposer ( like the one for PODP )

  *8   Cache driver setting;  /CCR4=1D
  *9   Cache driver setting;  /CCR2=16   /CCR4=1D  /WBE 
*10   Cache driver setting;  /CCR2=56   /CCR4=1D  /WBE
*11   Cache driver setting;  /CCR2=56  /CCR4=1D  ( with or without  "/CBE" makes no differencege) 
*12   Cache driver setting;  /CCR2=16   /CCR4=38 
*13   Cache driver setting;  /CCR2=16   /CCR4=38  /WBE

 *3, *5  are tested with 5x86 133MHz/4x model.

  Additional information.
  1. Under DOS environment, default statues of Lacuna planar with Cyrix5x86/EverGreen586
      Interposer  set to 4x speed is

      PCR0=0  CCR1=0 CCR2=0 CCR3=10 CCR4=05 SMR1 thru 2=0 PMR=2 DIR0=2C DIR1=5 

  2. Comment for CCR3 and PMR are probably not allowed for OS/2 driver.

  3. WBE comment improves CPU Integer and Floating point marks dramatically  but at the same 
      time  it decreases video and  disk access performance.

  4. Not suitable switches are;
      CCR2=22, CCR4=56, 

  5. Don't know how "/CBE" parameter works.    /WBE means  Write-back cache Enabled.

Impression
  Tired and exhausted !!

  I may say
     1)  Cyrix 5x86  series is not good at DISK I/O on SysBench.exe,  I don't know if Cyrix is really so 
          bad at  Disk Access on every bench program and on usual usage for daily job. 
     2)  Video performance is one of weak points of Cyrix CPU. 
     3)  It's difficult to determine *good* sw parameter for register setting. 

 By the way  5x86 100GP x4 model marked 85.300 for Dhrystone in CPU Integer test. Don't you think
 it's astonishing ?

 As far as OS/2 and Sysbench concerns, it seems that 4x models rated 133MHZ and 100MHz are 
 almost same.

 And in the end, I don't want to say but if I trust the result then  AMD 5x86-133/ADZ in  PGA package
 may be a one to take for OS/2 on Lacuna.  No need to hustle to get appropriate cache driver setting.

 But how do you consider following scores ?
 Here is Today's Final Result for usual Cyrix 5x86 100GP 3x speed model.
 I think I probably succeeded. 

 CPU; Cyrix 5x86 100GP 3x speed model
 Interposer ; ACE486 designed for intel DX4-100, no jumpers no switches. Fixed to 3x multiplying. 
 Cache Enabler statement;  Device=C:\ETI\ET586OS2.SYS  /CCR2=16   /CCR4=38  /WBE
 The statement is exactly same with *13 above used for 5x86 133GP in 4x multiplication, but the result
 for 3x model is very different from that of 4x model as shown below.

   ET586OS2 .sys should be added at the bottom of config.sys ( for OS/2 only.  Not for DOS/Win9x ) 

	Integer	Float P	Dive	PM	Disk I/O	CD I/O
Intel DX4-100 3x	44.159	4.809	10.225	21.411	15.999	17.304
Cyrix 5x86 100GP 3x	51.014	6.794	10.195	21.462	16.756	17.314
2nd execution   *1	50.766	6.779	10.205	21.448	15.707	15.707
Cyrix 5x86 w/o driver	45.262	6.088	10.136	14.230	9.056	17.284
IBM 5x86 100  3x   *2	50.703	6.482	10.194	21.447	16.679	17.302

    *1  Executed the program twice to confirm above result. 
    *2  Added  just for confirmation. Same statement was used.  Tested on 2002.10.26 
 

  2002.10.14
  George Albus finally got following result with his Lacuna 25MHz planar using same driver statement 
  as mine ( /CCR2=16   /CCR4=38  /WBE  ) .
  Integer and Float P are  improved a lot . Dive is a bit bihind of DX4-100.    PM and Disk I/O are both
  almost same with DX4-100.  et586OS2.sys apparantely has brought out Hidden Ability of Cyrix CPU. . Don't you think so ? 

	Integer	Float P	Dive	PM	Disk I/O	CD I/O
DX4-100	21221	3348	16174	20464	15119	31076
Cyrix 5x86  No Driver	20517	3967	12825	12367	6530	26904
Cyrix 5x86 et5x86OS2	35400	4687	13597	19534	15269	31083

 But let's see his coment about disadvantage of the driver ;
    " Only thing is I lost the CTR-ALT-DEL function  :-(  "

 I do not have clue for this symptom.  Don't know if another HEX vallue could solve it or not.