Talk:IBM System p
|WikiProject Computing||(Rated Start-class)|
This page is practically worthless. I wish I could add something useful but I came here looking for information on some systems I recovered. Even the pathetic AS/400 page has more data. dreddnott 21:48, 24 July 2006 (UTC)
Deep Blue: 100M or 200M?
This articles says that Deep Blue is capable of making 100M calculations per second. I have been at the Computer History Museum in Mountain View, CA, USA and I have seen Deep Blue in person and while there I read that Deep Blue is capable of making 200M calculations per second. Which one of the two figures is the correct one?
- According to IBM's Deep Blue page there was a second, updated version of Deep Blue that performed 200M chess positions per second. I don't know if it was the early version or the updated version that beat Kasparov. Someone might have to dig deeper into this. -- Henriok 20:48, 18 December 2006 (UTC)
Henriok, thanks for the feedback. I checked the IBM Deep Blue article and I see that. The Deep Blue I saw at the Computer History Museum was an RS/6000 SP2, probably one of the updated versions from 1997 (200M x s).
- If my memory serves me well, there were two matches between DB and GK. The first was lost by DB, and it was upgraded before the challenge was issued again. Thus I would say the updated version is the one which have beaten the human champion.
- Cannot recall what were the nodes of the first version (I think they were 58H) but am sure they were POWER2 nodes, and the upgraded version for sure used P2SC nodes. -- 188.8.131.52 (talk) 21:18, 18 July 2010 (UTC)
The RS/6000 hardware used to be called a "rios" by many IBMers, after the name of the early POWER chip. A namecheck would be handy.
Great work on creating a section of System p hardware through the ages. This list, however, competely takes over the entire article. Wouldn't it be better to create an article of its own, lets say List of System p hardware, and just keep the current models on this page? -- Henriok (talk) 00:02, 7 December 2007 (UTC)
- After my recent edits to make the section in question more "tidy", it has become obvious that the article deals more with the RS/6000 than the System p. I think it would be better if we stopped redirecting "RS/6000" and all the other former names of the System p to this article and turn those redirects into articles for discussing the machines that used those names. If there are no objections, I will carry this out in a week or so. Rilak (talk) 08:47, 31 August 2008 (UTC)
The table is very large and tedious to edit. It would be more manageable to edit if it was splitting into neat little chunks. Since I just put the existing list into table format, I'm asking for comments as to how the table should be splitting just in case someone had a system set up that I didn't know about. I propose that the table be split up into "Types". For example, most of the Model 2xx systems belong under "Type 7011". I'm assuming that "Type" denotes a common system architecture, so splitting the table up like this will also permit future comment on a group of systems that share common characteristics. The "Types" will then be grouped into application, for example, low-end servers/workstations, mid-range, high-end and so to make the existing information more useful. Currently it is hard for people to put the systems into context as most of the entries do not have specifications that will at least give hints as to what the system is used for (example: the 24 CPU Model S80 is not likely to be a workstation). What does everyone think? Rilak (talk) 10:58, 22 August 2008 (UTC)
eServer pSeries is here, should the RS/6000 be too?
IBM eServer pSeries just redirects back here, as "eServer pSeries" -> "System p" was a name change. Was "RS/6000" -> "eServer pSeries" a name change, or did it also reflect a split between PowerStation workstations and eServer pSeries servers? Guy Harris (talk) 18:39, 1 June 2012 (UTC)