Linked by Anton Klotz on Thu 18th Jan 2007 18:16 UTC
This article tries to explain why workstations are no longer an appropriate tool for the present working environment, what the alternatives are, and what consequences it has for the development of OSes.
Permalink for comment 203566
To read all comments associated with this story, please click here.
I think this is one of the main points here, the gap between the top and bottom of the computing platforms in use has narrowed a lot, such that it isn't at all clear what people mean when they say workstation.
E.g. I have a windows machine at my desk, that is I beleive called by the manufacture a workstation, because (maybe it has 2 CPUs in or more than a couple of monitors attached ) - but its a few years old and probably no more powerful than a modern bottom end "dual CPU". All would be more powerful that the ageing sun "workstation" I also use from time to time to work with Solaris code.
In addition, the other major change, is the way we scale computing power, its no longer normally just ever faster bigger/faster boxes, as you move from workstation to server.
Now our most powerful computers are no more than clusters of boxes similar to those we have on our desktops, which has lead to design changes in software, where once our final code may have run on seriously heavy iron, we may have needed a "workstation" just to test / develop it.
Instead we now have software that it more componentised and can be run/tested /developed in smaller isolated blocks.
Sometime ago, there was a PC for day to day tasks / documents e.t.c. But it couldn't really run the more complex systems we would build even in development, those simply needed more memory / more CPU / more throughput to even test them. Then our production servers were generally large, expensive multi CPU boxes.
Now the deployment system tend to be "grids" of smaller systems, rather than big hulking machines.. so the software can also now run in smaller units, which means we no longer
So yes, I would predict a continued demise of the workstation..
Member since:
2006-01-18
I think this is one of the main points here, the gap between the top and bottom of the computing platforms in use has narrowed a lot, such that it isn't at all clear what people mean when they say workstation.
E.g. I have a windows machine at my desk, that is I beleive called by the manufacture a workstation, because (maybe it has 2 CPUs in or more than a couple of monitors attached ) - but its a few years old and probably no more powerful than a modern bottom end "dual CPU". All would be more powerful that the ageing sun "workstation" I also use from time to time to work with Solaris code.
In addition, the other major change, is the way we scale computing power, its no longer normally just ever faster bigger/faster boxes, as you move from workstation to server.
Now our most powerful computers are no more than clusters of boxes similar to those we have on our desktops, which has lead to design changes in software, where once our final code may have run on seriously heavy iron, we may have needed a "workstation" just to test / develop it.
Instead we now have software that it more componentised and can be run/tested /developed in smaller isolated blocks.
Sometime ago, there was a PC for day to day tasks / documents e.t.c. But it couldn't really run the more complex systems we would build even in development, those simply needed more memory / more CPU / more throughput to even test them. Then our production servers were generally large, expensive multi CPU boxes.
Now the deployment system tend to be "grids" of smaller systems, rather than big hulking machines.. so the software can also now run in smaller units, which means we no longer
So yes, I would predict a continued demise of the workstation..