Subject: Re: Upper limits of CL From: Erik Naggum <erik@naggum.net> Date: Sat, 15 Jun 2002 18:57:45 GMT Newsgroups: comp.lang.lisp Message-ID: <3233156264920068@naggum.net> * Christopher Browne | In comparison, I can walk down the street and get an IA-32 motherboard | that supports 2GB of RAM for maybe a couple hundred bucks. But it won't | support 8GB. You are aware that IA-32 has support for 36-bit addressing, right? (The PSE36 CPU flag reports its presence in a particular processor.) Chipsets with support for the full 64GB address space are certainly available, and Linux supports them in recent kernel versions, giving you 4GB of physical memory per process. I am not so sure the raw need for 64-bit processors will be all that imminent, given the ability to provide so much RAM to the 32-bit processors. Not that 32 bits is enough, of course, but we may find better ways to work with 64-bit units than switching the whole processor. The only serious short-coming in modern implementations of IA-32 is the register starvation. It is quite fascinating what Intel (and AMD) have been able to do with this architecture. -- In a fight against something, the fight has value, victory has none. In a fight for something, the fight is a loss, victory merely relief. 70 percent of American adults do not understand the scientific process.