Pro/Forums - View Single Post

Cathar · 09-05-2006, 07:15 PM

Agree with many of your comments Joe and I think you've implicitly hit the nail on the head with the $100 to $500 CPU comment.

I do think that one very important point to highlight is the lack of a "killer app", in the sense of some application that everyone wants to use that taxes current systems to the absolute limit whereby there is actually something worthwhile to be gained by overclocking everything until it bleeds.

I remember as much as 3 years ago, the beginnings of grumblings of those in the overclocker crowd starting to question, "Why bother?". Sure, some people were still banging away at 3DMark and SuperPI (and still are), but aside from that crowd computers almost quite suddenly got "fast enough" for most people. Heck, even to this day I still have an old crusty Socket A system that forms the basis of my primary PC gaming platform, and I am fully aware that the reason why I still have it is because I haven't suddenly found any pressing need for it go any faster. Sure, I've updated the video card a couple of times to keep up with gaming advances, but the CPU at the heart of it is still more than adequate.

Recently I picked up a camera, and started encoding videos of my riding, and yeah, while some newer CPU might be 2 or even 4x faster at video encoding, it's not a pressing issue. I can move data around in the video editing window just fine, so the user-interface suffers from no noticable slowness. It's only when I click to start the encoding process does the CPU actually do anything, and even then I can just let it run while I go have a coffee, or have a meal, and come back and it's done.

So much for multimedia being the "killer app" to drive sales and need for faster CPU's.

I really believe that this is the root cause of the malaise. The need is gone. "Need is the mother of all invention" as the saying goes, but for a long time now there has been little real pressing need for the enthusiast crowd to spend the big dollars to make the CPU's go faster. As people often say, they can go buy a cheap, adequate, quiet air-based heatsink, run the CPU at 90% of the speed of what they could do on water-cooling, and not even once think that they are lacking for that extra 10% of CPU power.

The malaise came when the engineers, and even the enthusiasts, became fat and happy with the status quo. Sure, water-cooling is starting to push up against some pretty hard walls in terms of performance as well, but if there was a pressing need, people would be trying harder. There is no need though, and almost collectively the water-cooling world has thrown its hands up in the air and said "Why bother? It's too hard and I'm happy enough anyway".

09-05-2006, 07:15 PM	#11
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Re: watercooling lost its magic Agree with many of your comments Joe and I think you've implicitly hit the nail on the head with the $100 to $500 CPU comment. I do think that one very important point to highlight is the lack of a "killer app", in the sense of some application that everyone wants to use that taxes current systems to the absolute limit whereby there is actually something worthwhile to be gained by overclocking everything until it bleeds. I remember as much as 3 years ago, the beginnings of grumblings of those in the overclocker crowd starting to question, "Why bother?". Sure, some people were still banging away at 3DMark and SuperPI (and still are), but aside from that crowd computers almost quite suddenly got "fast enough" for most people. Heck, even to this day I still have an old crusty Socket A system that forms the basis of my primary PC gaming platform, and I am fully aware that the reason why I still have it is because I haven't suddenly found any pressing need for it go any faster. Sure, I've updated the video card a couple of times to keep up with gaming advances, but the CPU at the heart of it is still more than adequate. Recently I picked up a camera, and started encoding videos of my riding, and yeah, while some newer CPU might be 2 or even 4x faster at video encoding, it's not a pressing issue. I can move data around in the video editing window just fine, so the user-interface suffers from no noticable slowness. It's only when I click to start the encoding process does the CPU actually do anything, and even then I can just let it run while I go have a coffee, or have a meal, and come back and it's done. So much for multimedia being the "killer app" to drive sales and need for faster CPU's. I really believe that this is the root cause of the malaise. The need is gone. "Need is the mother of all invention" as the saying goes, but for a long time now there has been little real pressing need for the enthusiast crowd to spend the big dollars to make the CPU's go faster. As people often say, they can go buy a cheap, adequate, quiet air-based heatsink, run the CPU at 90% of the speed of what they could do on water-cooling, and not even once think that they are lacking for that extra 10% of CPU power. The malaise came when the engineers, and even the enthusiasts, became fat and happy with the status quo. Sure, water-cooling is starting to push up against some pretty hard walls in terms of performance as well, but if there was a pressing need, people would be trying harder. There is no need though, and almost collectively the water-cooling world has thrown its hands up in the air and said "Why bother? It's too hard and I'm happy enough anyway".