Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Silicon Graphics, Inc. (SGI) -- Ignore unavailable to you. Want to Upgrade?

To: Esvida who wrote (3639)	11/21/1997 2:51:00 PM
From: John M. Zulauf	Read Replies (3) \| Respond to of 14451

Al, educate me. Q1. Are you saying that modern parallel program tools make the sort of hand work I've described unnecessary? On the mas-par I know I am out of date, the last technology I spent any time with was Linda (some persistant object system who's name escapes me). Yes I know that parallizing compilers have made great strides, but tuning and optimizing the last 2x perform is still an issue in #pragma tic terms? Q2. Are you saying that extant technology can easily migrate C++ style objects across non-continuous memory. I've looked and the distribute object technologies, but they all seem to add a huge layer of computation and design overhead. You have to design in some distributed object class specification layer, using some on tool to create the networkable, rellocatable, persistant objects. Compared to a flat 64bit memory space shared over a large number of processors, this cannot avoid being more work and more bloat. Q3. Are you saying you think the HP messaging, and SCI-ring Sun clusters are even nearly as easy to program well as a single image flat memory system? (ahem!) Help me out here. Thanks john.