Operating System - OpenVMS
1828490 Members
2225 Online
109978 Solutions
New Discussion

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

 
Duverney Barros Arevalo
Occasional Contributor

OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Hi,

Where can i find information about differences between OpenVS 7.3 for Alpha and
OpenVMS V8.2-1 ?
8 REPLIES 8
Karl Rohwedder
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Pls. check the Releasenotes and the 'new Features Guide' for OVMS V8.2.
You can find them at:
http://h71000.www7.hp.com/doc/os82_index.html

regards Kalle
Arch_Muthiah
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Hi,

This link show quick specs and feature of HP OpenVMS for Integrity Servers Versions 8.2-1 and 8.2, and HP OpenVMS on Alpha.

http://h18000.www1.hp.com/products/quickspecs/12156_div/12156_div.HTML

And this link shows the special features and benefits of HP OpenVMS Version 8.2-1 for HP Integrity server systems

http://h71000.www7.hp.com/openvms/features_benefits_chart.pdf

Archunan
Regards
Archie
Antoniov.
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Hi,
let me add some user consideration:
- OpenVMS on IA64 is slightly faster than OpenVMS on AXP
- Integrity server uses standard components so you can save money
- OpenVMS V8 use a new licensing model; just for example user are unlimited
- Itanium uses a unique internal calling model indipendent from OS; may be easy to port linux code on openVMS

Antonio
http://it.openvms.org

Antonio Maria Vigliotti
Robert Brooks_1
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Antonio wrote . . .

Hi,
let me add some user consideration:
- OpenVMS on IA64 is slightly faster than OpenVMS on AXP

--

I'm not really sure what you are trying to say here. Certainly, a GS1280 is a wee bit faster than, say, an RX1600.

Are you trying to say that on comparable hardware, the I64 version of OpenVMS is faster than the Alpha version?

If that is your point, could you please explain in more detail?

-- Rob
Ian Miller.
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

RE: performance - on low to mid range systems then most applications are faster running on current model Itegrity servers running VMS V8.2-1 than current model Alphaservers running VMS V8.2 according to hp.

The GS1280 still has the edge at the top end over the Itanium Superdome. It is expected that Monecito based top end systems will be faster than GS1280 but that remains to be proved.

As already mentioned the new features and release notes for VMS V8.2 and V8.2-1 are the best places to look. Another good place is http://h71000.www7.hp.com/openvms/integrity/resources.html
____________________
Purely Personal Opinion
Antoniov.
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Ian,
you posted quicker than me and said exactly what I wanted to say.

Antonio
http://it.openvms.org
Antonio Maria Vigliotti
Martin Vorlaender
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

>>>
let me add some user consideration:
- Itanium uses a unique internal calling model indipendent from OS; may be easy to port linux code on openVMS
<<<

I have no idea what a processor feature could have to do with porting from *ix.

The main showstoppers in porting from *ix are use of C RTL functions that don't exist on VMS (yet - e.g. fork, select on anything but sockets) and the building process (though in that area, with GNV is has become considerably easier).

cu,
Martin
Robert Gezelter
Honored Contributor

Re: OpenVMS 7.3 Alpha and OpenVMS 8.2-1 IA64 comparison

Duverney,

I must comment on two points:

- I agree with Martin on the question of "whether the calling standard makes a difference in porting".

The calling standard is not a primary cause of porting issues, calls and presumptions about the semantics and models of the underlying OS are the problem. The calling standard differences between, say Alpha and IA-64 are a far lesser issue (actually, almost IMO, irrelevant).

- Performance

As I have been noting, performance is an issue which only matters in the particulars of a given situation. That equivalent priced systems give approximately similar performance is neither a surprise, nor an absolute statement.

Since my initial presentation on porting to IA-64 (see http://www.rlgsc.com/cets/2001/1620.html) and before, I have been advising people to do the science of performance, not guess. I would not advise a client to perhaps plan the budget based upon an estimate of system performance, but I would highly recommend doing baseline performance work on one of the rx2600/1600-series systems before actually signing the Purchase Order for a significantly larger system. As a matter of principal, all performance numbers are estimates, the best performance estimates are based on your actual workload, with the system properly tuned.

- Bob Gezelter, http://www.rlgsc.com