!!!!!!!!!!!!!!!!!!!!!!!!!!! >From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] >On Behalf Of Stephen Reed >Sent: Saturday, November 30, 2002 10:19 PM
>At Cycorp, we switched from Symbolics Lisp Machines to >Linux three years ago, and although we have a port of >the Cyc knowledge base for Win32 we are now blocked by the >memory model of Win32 which reserves 2GB of virtual >memory for the OS. Linux reserves only 1 GB and because >the Cyc KB is currently memory resident we cannot contain the >full Cyc KB in Win32 (max user address space is 2 GB) but >we can contain the full Cyc KB in Linux (max user address >space is 3GB). >..applications suffering in that memory paging occurs >even if sufficient RAM exists. The Win32 server OS is much more >expensive and still does not completely solve this >performance problem. !!!!!!!!!!!!!!!!!!!!!!!!!! Yes, there is that paging problem. I haven't checked out Win64 on Itanium, hopefully this has been addressed(pun?). For the AI that I've been doing Windows has the essential ingredients. Developing these "Hunter Gatherer" nodes, they intricately exploit the Internet Explorer object model that is on 10 zillion Windows desktops. The nodes(one per machine) avoid paging penalties by keeping themselves way below major paging thresholds. In fact they try to keep themselves running healthy on machines with less than 64 megs. But to harmonize processing power they look for other nodes in IP address-space subsets, link themselves loosely, and generate/relay summary information through the lazy mesh. The nodes function as "constellations" but they don't penalize resources too much and all the glory of the IE objects are exploited. But if the AI app needs to "swim" in large amounts of RAM there is that paging situation... ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]
