CPU Usage question

Yes it is. Inconceivable, that is. I thought you understood this technology, you certainly talk like you do, but I can't see how you could understand it and dream of trying to make that kind of argument, or start talking about "stop and go" running, or talk about them "sitting idle"... running SL with the frame rate throttled, or even killing every program on the computer and letting it sit there handling timer interrupts, is not "off", or "idling" or "stop and go" any more than running it full speed is: the processor is still flipping all the same switches millions of times a second even at its most "idle".

I don't know if you're trolling me, but if so, bravo. Well played. I'm not going to bother reading the rest of the message, my eyes already hurt from rolling in their sockets getting this far.

I don't know much about such things, but I do know that old valves ran best when running at near max and I have always been led to believe that solid-state stuff was the other way around. My knowledge of basic physics regarding electrickery reminds me about the way they work too and draws the same conclusion.

Nothing lasts forever...'cept maybe taxes.

Look, I said those first few paragraphs were simply an exercise in "silly-logic". I said right at the beginning that I didn't agree with the argument I was about to present. I had thought the reason I was presenting such a factually absurd, but logically sound, argument would be rather obvious, once you'd read the whole post. But I seem to have miscalculated a bit. I didn't anticipate that you'd stop reading. But since you did, please allow me now to try to explain in a better way. Hopefully you'll read this whole post.

The truth of our situation is that one of us is factually incorrect here, just as that "silly-logic" argument I presented was also factually incorrect. The truth is that either a CPU's lifespan will be shortened by SL's heavy usage or it won't. One or the other of us is factually wrong about the subject, but both of our arguments are logically sound. We're both pretty smart, and both pretty sure we're right, but only one of us actually is. Neither one of us can know for sure, since we've got no practical way to test, which is why the best we can hope for is to agree to disagree, as I'd suggested. In the absence of proper testing, and in the presence of equally compelling logic on both sides, there's not a whole lot we can do to persuade each other one way or the other.

Therefore, I figured the best way to shock us both into realizing that this debate wasn't going to go anywhere was simply to present an example of an argument that we both could identify as totally wrong if you already know the facts, but that also could appear to be persuasively right if you don't.

In other words, that you were so easily able to identify this "silly-logic" presentation as wrong was exactly my intention. I'm sorry if I didn't make that clear enough.

What I was hoping to do was to establish that because solidly constructed logic can be based on incorrect facts, neither one of us will able to "outlogic" the other on this. Both of our arguments are equally well constructed. It just happens that one or the other of us has built our logic on an assumption that isn't factually true. And since we can't test, we can't know which one of us that is. We both believe we're right, but we don't really know.

After demonstrating the ridiculous, I then presented my real, totally sensible, argument, which you, by your own admission did not bother to read. That's a little disappointing.




Thank you. Yes, of course all that is true. Now that you've said it, you might be in a better position to understand where I'm coming from with my real argument. As you've now pointed out, and we all already knew, a huge amount of activity is happening inside the CPU whenever it's on, no matter how hard any one program is "working" it. The difference between working 100% and working 1% is not, in my opinion, statistically significant, when compared with the difference between simply being on or off. Therefore, I cannot believe that allowing a program to use a CPU at 100% will wear it out significantly faster than having that same program use it at 50% or 5% or 1%. As long as it's kept adequately cool, it will run for years and years and years. That's all I've been trying to say this whole time.



No, I'm not trolling you. I'd like to think you know me well enough to know that's not the sort of thing I do. I was simply trying to emphasize a point about the argument itself. Perhaps I should have done a better job of explaining that that's what I was doing before I did it, or while I was doing it, so that you wouldn't have stopped reading. I'm sure you would have understood that point had you actually read the entire post instead of stopping after the silly part.

Do you understand it now, at least, even if you still don't agree with my side of the real debate?

No, I'm not going to read the post. This is not a "matter of opinion". The major cause of failure in large scale integrated circuits is electromigration caused by thermal stress. That's not just my opinion, you can look it up anywhere.

No amount of playing word games will change that. You are simply wrong.

Holy Hell, what a thread!

Let me throw my own two cents down on the table, and then I'll gracefully exit:

I have personally observed the SLVoice gobbling up an inordinate number of CPU cycles - at least on the Linux client - and I have no reason to believe the Windows binary would be any different. If you are not using voice, particularly if you are standing alone on a sim, try disabling voice. Your CPU will thank you.

ETA: Chosen Few: I have been an electronics technician for over twenty years, so I can vouch for Argent's statement. Thermal stress is the leading cause of failure of solid-state and semiconductor components. Frequent hot-and-cold cycles will cause electronic components to degrade and burn out in much the same way as frequently turning an incandescent light bulb on and off will shorten its life. It is simple physics at work here.

Argent, come on, you're just being pigheaded and stubborn now. I never once said that thermal stress wouldn't cause a a chip to fail. I simply said I don't believe running a CPU at 100% is a problem AS LONG AS IT"S KEPT ADEQUATELY COOL. Why you keep ignoring that last part, I can't imagine.

How many times did I say "Get a laptop with a good cooling system" or "Get a good CPU fan" througout this thread? I've been beating that drum constantly. You just keep refusing to hear it. I guess that's what happens when you just read the first paragraph of every post and then decide you know what the whole thing says, rather than actually reading it to find out if you've guessed right. Wow, what a ridiculous way to try to have a discussion. You really have no idea what I've been trying to say, do you?

Let me make it easy for you. It's a simple concept. Good cooling system = chip stays cool = less electromigration = chip lasts longer. I know you're intelligent enough to get that.

He'll have the last word..........LOL

Stalking again, are we? If you're going to comment, please comment on the discussion topic itself. Kindly keep your opinion of the participants under your hat. Thanks.

Because I have sufficiently refuted it already. Modern laptops are simply not built to run at 100% for hours a day every day for more than a few years. It's not cost-effective.

I'm pretty sure whoever designed the cooling system in my G70 would beg to differ. It's got two massive fans underneath, and about 75% of the rear of the chassis is a giant exhaust port. You can literally feel the air blowing out of it from across the room. That thing runs cool as a cucumber, even when overclocked. (And somehow it remains whisper quiet as well, which the most amazing part of all.)

In any case, you haven't sufficiently refuted anything. Even if the lifespan is shortened by overuse, there's no way it will be shortened enough to die before the laptop itself becomes obsolete. What does it matter if the CPU dies in 9 years instead of 10, if I'm only going to use the machine for 4?

They don't have to build laptops to break. People will buy new ones every 2-4 years whether they break or not.

I'm not stalking again........I wasn't even talking to you. I simply stated a fact.........your ego will never let someone else get the last word. Call it stalking........I call it stating a fact. Carry on........as you always do.

As I said, if you have something to say on the topic at hand, say it. Otherwise, you're just wasting your time, and everyone else's. If you really can't bring yourself to stop this behavior, I'd highly recommend you consult your local Yellow Pages, under P for psychiatrists. Obsessive compulsion is treatable.

Now, do you have anything on topic to say, or don't you?

Wasn't this the original topic?




no one really answered the original question...just went off on a tangent....

Yes.


The original question was answered in the very first reply. Did you miss it?

After the first reply, the discussion topic mutated into whether or not the answer SHOULD be the answer, in an ideal world. But no one ever disputed the answer IS the answer.

That's not supercooled because it's an expensive laptop, that's supercooled because it's specifically designed for gaming. Most laptops aren't built to blow everyone's papers off the table in the conference room and piss off your boss.
Because you're not moving some kind of magic slider... it's a statistical function. It's not "it dies in 9 years instead of 10", it's "it's 20% likely to die within 4 years instead of 5% likely".

That's not what I said.

And your point is?

Look, first of all, gaming hardware is what every SL user should be using anyway. So I'm not sure what you're trying to say is so unusual about that.

If SL weren't part of my life, my purchasing decisions would be very different. As a professional artist, I'd much rather buy machines with workstation hardware in them than gaming hardware. But since SL and other virtual worlds are such an important part of what I do, gaming hardware is a must. Anyone who uses SL with any degree of regularity, for any significant portion of time, should be making those same decisions. Always, always, always get what's suitable for the intended task; never ever get what's not. Again, it's a simple concept.

Second, expense has little to do with it, once you get past the total cheapos. The main point here is you said no modern laptop has an adequate cooling system, so I simply rebutted by giving an example of one that does. And it's certainly not the ONLY one. There are tons and tons on the market with similar or even better systems, all across the price gamut.

So for the umpteenth time, if your laptop is likely to die from overheating just because you're using the CPU at 100%, you simply shouldn't be using that particular laptop. There are plenty to choose from with good cooling systems. Get one.



Let's not exaggerate. It couldn't even blow a piece of paper anywhere if you put it directly in front of the exhaust port and attached a parachute to it to catch 100% of the air flow. I said you could feel it from across the room, but I didn't say it was with force. If it's on the coffee table, and you're wearing shorts, you can feel that the air hitting your knees is warmer if you're in its line of fire than if you're not. That's all.




And this is where we MUST agree to disagree. You don't have any method of testing to verify that alleged statistic. You're just going with your gut feeling. If you get 10,000 laptops and run them for 4 years straight, half of them running SL every day, and half not, then we'll have something to talk about. Until then, neither one of us can verify our beliefs.

You ready to just admit that we're not going to convince each other on this? I've certainly been for quite some time now.

So if this is normal, why don't i experience the problem...why didnt you ask about memory, graphics card, background programs, all that jazz...like i said in the middle of this post somehwere, my work computer has a celeron with 1 gig of memory, adn a pretty weak graphics card, and doesn't run at 100 percent cpu....

I also stated earlier that I don't believe that's normal behavior.

I see SL sometimes runs the CPU in the high 90% range, but normally it's much less than that.

I do not see running the CPU at nearly 100% as normal behavior for SL on my computers, either. It does occasionally, but normally it's between 50% and 70%.

First, it's not a problem.

Second, I don't know why it's not using 100% on your machine. All I know is what I've observed over the years. If it is indeed behaving differently on different hardware configurations, that's worth exploring.

But they're not. Most people using SL don't organize their life around SL. Many of them aren't gamers. Most of them can't afford to buy a new computer just to play SL, and the idea that you should buy a new computer just to play SL is sheer arrogance.

No normal non-enthusiast laptop, then, is built to run hours a day at 100%. I've ordered hundreds of laptops for salesmen and engineers over the years, and repaired and replaced them. I haven't bought into the gamer market, but over the past several years heat has increasingly become an issue for most laptops.
Don't be silly. There's been an enormous amount of testing done on this. It doesn't matter whether you're running SL or other games or even flash animations or high-CPU screen savers.
I'm not trying to convince *you* of anything.

If SL is limited by the GPU performance rather than the CPU performance you're less likely to be seeing the CPU running at 100%... unfortunately there's no standard meter for measuring GPU utilization.

Arrogance has nothing to do with it. The simple truth of the matter is one should never ever EVER approach any task, no matter what it is, without adequate tools to do it. If you're a baseball player who decides to take up golf, you don't say, "I've already got a bat and glove, so I'll just skip over buying golf clubs." If you're a business class laptop user who wants to take up SL, you SHOULD get a machine that's up to the task. If you don't, then you must surrender to the fact that you won't enjoy the experience nearly as much, and you MIGHT even damage your equipment.

So once again, if you're using a laptop that will overheat just because you're running the CPU at 100%, you shouldn't be using that particular one. You should be using one with a better cooling system.

Is that the answer that everyone's going to want to hear? No, of course not. That's why I said, "As much as I hate to say this," right at the beginning. But whether anyone likes it or not, it is the truth.

Is everyone going to take that advice? Again, no, of course not. A certain percentage of people will continue to blame the software and/or LL for their woes, because that's easier (and even more fun to some) than just to admit they bought something that wasn't practical for what they're now using it for.

You seem to be under the impression I've been expecting everyone who reads this to run out and buy a better machine. That couldn't be further from the truth. I expect no such thing. If I'm hoping for any good to come of this at all, it's simply to encourage people to weigh their options more carefully the next time they are in a position to make a computer purchasing decision.



I don't know if it's most in terms of models, but I'll agree with you that it's probably most in terms of units sold. Most people buy the cheapest crap they can get their hands on. That's not likely to change.

But again, what people do do and what they should do are two different things. I've been talking about what people should do. I have no illusions about what they actually do do.



Care to cite some links to some hard statistics, then?

I'd really like to know, from a solid scientific perspective, just how likely it is that a laptop CPU will fail from overuse before the laptop itself becomes obsolete enough that it should be retired anyway. If you can cite ANY peer-reviewed study on that, which can refute anything I've been saying, I'll certainly recant. But in the absence of that, both you and I are operating on assumptions, nothing more. Why that's seemingly so hard for you to accept, I can't understand for the life of me.



Then why have you been keeping this going? I've offered to agree to disagree several times now. What's the reason you won't take that offer?

Another really bad analogy.

A general purpose computer is a general purpose computer.

So everyone who wants to play SL should fork out two or three grand up front for a G70?

But it doesn't need to be the truth. Adding the ability to throttle FPS would not degrade anyone's SL experience, and would increase the number of people who are able to play, so what the hell are you arguing against it for? Why are you arguing that people SHOULD have to pay for a top of the line computer before they get into SL?

I'm not saying they shouldn't do that, if that's the way they roll, I'm just saying it shouldn't be necessary.

Why do I keep saying that? Because you're not the only person reading this.

Okay......let me contribute something useful to question put forth by the OP. If you would kindly go back to the initial post starting this thread...........read the last sentence. The question was (and still is, as far as I can tell) is CPU usage at or near 100% normal? Your response in the very next post is that that behavior is normal. I believe your answer is incorrect.......actually, it's just plain false. As you can see there have been several others post answers directly contrary to your assessment and you have chosen to ignore those observations. Instead you decided to derail the thread to discuss how CPU usage effects the lifespan of that device. You're contention that it's neglible because the CPU will become obsolete before it fails is silly.........actually inaccurate (false). We all know a CPU will generate heat as the load on that CPU increases........that is a fact. All electronic devices have a design limit. A point at which exceding that limit the device is subject to damage or failure..........the design limit on my processor is 104 C. Every processor has one and as long as you do not excede that you should not experience problems. And to be honest, I actually do not think SL running your processor at or near 100% would cause the processor to reach that temperature.......but, I can not tell someone that simply because I don't know that for a fact (it's my opinion only and, unlike you, I don't claim expertise on the subject).

Argent's statements concerning higher temperatures on the core of the processor is accurate. Heat is the killer of all eletronic components.........another fact. Argent's claim that even if that heat is below the design limit will cause a shortening of it's lifespan is true.........and you even seem to agree on that point. Yet, you insist that it is causing no harm...........because the CPU will become "obsolete" before any failure. That may or may not be true...........but it very likely is causing harm. You are wrong.

I'm including a link to a screen shot of my computer screen while visiting Sine Wave Island (one of the busiest places I could think of in SL). Look at it and notice the CPU usage is well below 100%. And while you are looking at it take notice of the core temps too...............notice the lesser usage on each core shows lower core temperature. This behavior on this computer is typical for 4 of the five different computers I've used (or have been associated with) for Second Life over the 4 years since I first joined. The 5th computer that did, in fact, show 100% all the time was a Pentium D..............and guess what. It died prematurely.

To answer the OP............no. 100% CPU usage is NOT normal. People like our resident expert on everything Second Life seems to think it is..........take that to heart at your own risk. And furthermore people who tell you such "facts" are very likely causing many others to ignore a potential failure of their CPU's. Something in thw way you have your computer set up is driving that usage higher than it should be. Background programs, anti virus, email checkers, Skype, etc all may be contributing to the excessive usage. I would suggest you find the problem and fix it..............or risk an unexpected failure way before the lifespan of the CPU is reached.

http://i323.photobucket.com/albums/nn474/PeggyPaperdoll/SineWaveCPUusage.jpg

Now, lets see how my favorite expert disects this post and attempts to convince me how stupid I am. Any bets?