[Info-vax] VSI OpenVMS Hobbyist Program Announced.

[Jan-Erik Söderholm]

Den 2019-06-16 kl. 18:06, skrev Dave Froble:
> On 6/16/2019 7:15 AM, Jan-Erik Söderholm wrote:
> 
>> $ sh sys
>> OpenVMS V8.4-2L2 on STUDNT 16-JUN-2019 12:35:17.08 Uptime 0 00:03:51
> 
>> $ tcpip sh ver
>>
>>    HP TCP/IP Services for OpenVMS Alpha Version V5.7 - ECO 5
>>    on an AlphaServer 400 4/166 running OpenVMS V8.4-2L2
> 
> This confuses me.  2L2 is for later Alpha processors, EV6 and such. Could 
> be wrong, but I thought the AlphaServer 400 used an EV4 processor.
> 

Yes, you are right. I didn't thought about that. It does run anyway.

I have a slight recollection from the Rdb kits (that also has a kit
optimized for later processers) that the EV56+ extra instructions are
emulated in software on earlier processors.

So it *might* be that VMS V8.4-2L1 would have run faster on this
particular emulation platform than the current 2L2 kit. I have no
idea, just speculating here... :-)

It would be interesting to run 2L2 on a real physical EV4 based
AlphaServer 400.



Part of the SHOW CPU/FULL output:

$ show cpu/fu

System: STUDNT, AlphaServer 400 4/166


CPU 0    State: RUN          CPUDB: 81C16000     Handle: * None *
        Process: SYSTEM         PID: 00000096
   Capabilities:
         System: PRIMARY QUORUM RUN RAD0
   Slot Context: 8300E1C0
      CPU  -  State..........: RC, PA, PP, CV, PV, PMV, PL
              Type...........: EV4 (21064), Pass 2 or 2.1
              Speed..........: 1216 Mhz
              Variation......: VAX FP, IEEE FP, Primary Eligible

The speed is fine anyway... :-)

But:

On a physical DS20E 666 MHz:
$ @ CALCULATE_VUP.COM
  Approximate System VUPs Rating : 1426.3 ( min: 1423.0 max: 1428.0 )

On this kit:
$ @CALCULATE_VUP.COM
  Approximate System VUPs Rating : 172.0 ( min: 149.0 max: 180.0 )