Cakes Bottles Smokescreens And Overclocking
Overclocking is one of those so-called "good things in life." In other words, it's something to be enjoyed, but not obsessed over or given disproportionate attention to. I suppose it's a bit like the icing on a cake.
Strangely, though, here I am, working an hour of make-up time, and my mind wanders to some OCing. Specifically, I find myself thinking about how sometimes some people might have an unfocused approach to the dark art.
Consider for a moment that we all have different reasons for overclocking. Some do it to increase usable performance for their machines; they push the hardware hard, but not to the absolute limit. Stability and functionality are still very high on the list of priorities, and so they leave a margin for error. Then there are people who look at overclocking as some kind of sport, a competition, even if it's just one against one's own established limit. They try to go further than they ever have gone before, pursuing raw performance without any considerations for practicality or functionality (at least, from my perspective). In a way, it's like a video game: You try to beat your own best benchmark score.
I'm not at all suggesting one approach is "better" than the other; we all do things for our own reasons and rationale, and it's certainly not my place to pass judgment on why people do what they do with their own computers.
However, I must confess some uneasiness when I encounter (all-too-frequently) stories of how some OCers want to have their cake and eat it too. Sorry to break it to you, but my own (admittedly limited) expertise in the dark art suggests that ultimate performance and ultra-reliable functionality are mutually-exclusive targets. In other words, you can have one, but it will always be at the expense of the other.
My limited time in engineering school taught me enough to know that any design has limits; to change those limits, you have to change the design. This is the fundamental principle at work when we constantly upgrade our equipment piece by piece. But everything has a finite limit of capability that cannot be exceeded if you have certain expectations of performance and reliability. Think of it as a glass, or a bottle: You can only fill the container with so much before the contents spill out. You can even mix the contents, if you wish, but the total volume is a finite limit. If you want more volume, you get a bigger bottle.
When it comes to overclocking a system, often you have to choose which approach you'll be using: Do you shoot for absolute max performance (and deal with a temperamental beast that can crash on you anytime), or do you hedge a little and allow for some breathing room and therefore dial in some stability for your equipment? Some enthusiasts, though, think that they can OC a system with impunity, then wonder why it's crashing all the time. It's like trying to put a black liquid into the bottle and mixing it with a white liquid, filling the bottle to the brim, then finding that one substance spills out when you try add more of one (or the other) to the mix.
I am fully aware that these sentiments are probably viewed negatively by hardcore enthusiasts who think that performance is the end-all be-all, and that's fine. But unless you have a very deep set of pockets full of green, then it's all but impossible to buy the cake, slice it up, and eat it too. For the vast majority, there needs to be a shift in perceptions, at the very least; there ought to be an honest and considered appraisal of your system's capabilities. I think we all need to understand how big our own bottles are, how much they can contain, and then decide what to put in them. I believe we all just need to focus on what our approach ought to be before applying our skills and knowledge and practice the dark art.
Otherwise, it'll be like trying to hit a moving target shrouded in a smokescreen.
Strangely, though, here I am, working an hour of make-up time, and my mind wanders to some OCing. Specifically, I find myself thinking about how sometimes some people might have an unfocused approach to the dark art.
Consider for a moment that we all have different reasons for overclocking. Some do it to increase usable performance for their machines; they push the hardware hard, but not to the absolute limit. Stability and functionality are still very high on the list of priorities, and so they leave a margin for error. Then there are people who look at overclocking as some kind of sport, a competition, even if it's just one against one's own established limit. They try to go further than they ever have gone before, pursuing raw performance without any considerations for practicality or functionality (at least, from my perspective). In a way, it's like a video game: You try to beat your own best benchmark score.
I'm not at all suggesting one approach is "better" than the other; we all do things for our own reasons and rationale, and it's certainly not my place to pass judgment on why people do what they do with their own computers.
However, I must confess some uneasiness when I encounter (all-too-frequently) stories of how some OCers want to have their cake and eat it too. Sorry to break it to you, but my own (admittedly limited) expertise in the dark art suggests that ultimate performance and ultra-reliable functionality are mutually-exclusive targets. In other words, you can have one, but it will always be at the expense of the other.
My limited time in engineering school taught me enough to know that any design has limits; to change those limits, you have to change the design. This is the fundamental principle at work when we constantly upgrade our equipment piece by piece. But everything has a finite limit of capability that cannot be exceeded if you have certain expectations of performance and reliability. Think of it as a glass, or a bottle: You can only fill the container with so much before the contents spill out. You can even mix the contents, if you wish, but the total volume is a finite limit. If you want more volume, you get a bigger bottle.
When it comes to overclocking a system, often you have to choose which approach you'll be using: Do you shoot for absolute max performance (and deal with a temperamental beast that can crash on you anytime), or do you hedge a little and allow for some breathing room and therefore dial in some stability for your equipment? Some enthusiasts, though, think that they can OC a system with impunity, then wonder why it's crashing all the time. It's like trying to put a black liquid into the bottle and mixing it with a white liquid, filling the bottle to the brim, then finding that one substance spills out when you try add more of one (or the other) to the mix.
I am fully aware that these sentiments are probably viewed negatively by hardcore enthusiasts who think that performance is the end-all be-all, and that's fine. But unless you have a very deep set of pockets full of green, then it's all but impossible to buy the cake, slice it up, and eat it too. For the vast majority, there needs to be a shift in perceptions, at the very least; there ought to be an honest and considered appraisal of your system's capabilities. I think we all need to understand how big our own bottles are, how much they can contain, and then decide what to put in them. I believe we all just need to focus on what our approach ought to be before applying our skills and knowledge and practice the dark art.
Otherwise, it'll be like trying to hit a moving target shrouded in a smokescreen.
Comments (0)