[kevs]

I notice this a lot. Recently with Photoshop, in file-browser, if I have not done reboot it few days, just to select images there or deselect them take while. Do reboot and the app is like a new person.

[msuper69]

Rebooting forces all the loose bits to the bottom of your Mac, thus clearing the way for programs to run smoother. Hence, everything is faster after a reboot

[Big Mac]

That doesn't sound right to me. You probably don't have enough RAM.

[wataru]

Things are only faster after a reboot if you don't have enough RAM. You must be doing something that fills up all of your physical RAM, forcing the data to be "paged out" to the hard disk. Then, when you finish your task, you go back to the Finder or whatever, and the system has to retrieve the Finder's information from the hard disk, which is extremely slow. If you had more RAM, ideally, your task and the Finder, etc.'s data could all fit into physical RAM and not need to be paged out.

[Wiskedjak]

Macs behave strange ... must be user error

[kevs]

Interesting. I have 1.25 Apple Ram. I thought this is more enough. Maybe not? Most Photoshop files are 100MB or less.

[Dex13]

jesus what res are you at?

no need to go above 150 .....

[mAxximo]

Originally posted by kevs:
I notice this a lot. Recently with Photoshop, in file-browser, if I have not done reboot it few days, just to select images there or deselect them take while. Do reboot and the app is like a new person.

I've been noticing the same. I have 2GB RAM on a Dual 2.0Gb G5 so you can't blame it on the memory. By the end of the day I can't even switch to freaking Safari without having to wait for the system to finish the stupid pageins/outs. So I have to reboot at least once a day to get some of the performance back. I have a G4 in System 9 that runs for weeks without ever needing a restart for things to be smoother. �Smooth� used to be built-in to the MacOS.

The problem, of course, lies elsewhere.

[wataru]

Even 2GB of RAM will get eaten up by Photoshop. Perhaps one of the apps you run has a memory leak?

My iBook has 640MB of RAM, and I don't find myself needing to reboot that often.

[kevs]

Maxximo:
You think we have more problems with A dual G5 in os X, than G4 is OS 9?



Memory leaK?? what is that?

I have enough ram then, I'm ok?

[OptimusG4]

Usually just logging out/in again will cure most problems.

[gururafiki]

I noticed major improvements with this problem after I went to 3 gigs from 512mb. I added on increments, so I noticed an improvement each time I upgraded. Although, after about a week or two I do like to restart my machine as I do notice it 'lagging' a little after a while.

[Spheric Harlot]

Originally posted by kevs:
Memory leaK?? what is that?

I have enough ram then, I'm ok?

Sometimes bad code in an application will keep eating more and more RAM, even if it doesn't really need it. Since OS X gives out RAM to applications on demand, this will eventually tax the memory management system and cause heavy swapping, slowing down your machine.

Internet Explorer does something like this under OS 9, except by design. Problem is that under the advanced memory management of OS 9, this kind of behavior invariably results in a system crash eventually unless the Exploder is the last app opened. Under OS X, it will usually suffice to quit the offending application.

[Randman]

It's not faster. It's just a glitch that happens from time to time in The Matrix.

[SMacTech]

Originally posted by mAxximo:

The problem, of course, lies elsewhere.

When you have this problem, open up terminal and do top -ovsize and post the results here.

This will output your running apps virtual memory size which will tell us a lot about what's going on. If indeed there is a memory leak, or PhotoShop is consuming all of your physical RAM, it will show here. If you need 3GB of RAM for the job at hand and you only have 2Gb physical memory, paging is going to occur, and possibly a lot of it too.

[Boondoggle]

Some HP Printer drivers can cause major slowdowns.

Oh yeah, and Virex. Those two are the only causes of slowdowns that I've experienced.

Bd

[kevs]

Are you all using 3rd party or Apple ram?

[hyperb0le]

A memory leak seems to be the likely culprit. Do you use Poisoned? I seem to remember that it had a problem with memory leaking.

[SMacTech]

Originally posted by kevs:
Are you all using 3rd party or Apple ram?

A mix of original RAM plus other 3rd party. RAM doesn't seem to be your issue.

Can you use top commands in terminal and post result information as I pointed out in my prior post? This can give us a much better idea what exactly is the issue.

[kevs]

thanks.
Someone want to define memory leak?

Spoke to Apple rep today, and he said just logging out each night help performance a lot. Might start doing that. He said 1.25 GB ram should be fine, but some Photoshop users say that 2 GB is great.

[hyperb0le]

Originally posted by kevs:
Someone want to define memory leak?

Originally posted by Spheric Harlot:
Sometimes bad code in an application will keep eating more and more RAM, even if it doesn't really need it. Since OS X gives out RAM to applications on demand, this will eventually tax the memory management system and cause heavy swapping, slowing down your machine.

[Tyler McAdams]

Could be a lot of things all together... like a bunch of open tabs in safari and other application open at the same time... your cpus getting hot because they've been doing something all day... when you reboot you lose the apps and cool down your system... The systems cache clears...etc

I think this article might be of interest to you:
http://arstechnica.com/reviews/02q3/...sx-10.2-8.html

Look at the "Performance Problems" area... It's talking about Jaguar but I don't think much of this particular subsystem routine has changed much.

[Big Mac]

Originally posted by mAxximo:
I've been noticing the same. I have 2GB RAM on a Dual 2.0Gb G5 so you can't blame it on the memory. By the end of the day I can't even switch to freaking Safari without having to wait for the system to finish the stupid pageins/outs. So I have to reboot at least once a day to get some of the performance back. I have a G4 in System 9 that runs for weeks without ever needing a restart for things to be smoother. �Smooth� used to be built-in to the MacOS.

The problem, of course, lies elsewhere.

I have a DP 2.0 Rev B, currently with 1GB of RAM. There is something seriously, seriously wrong with your system. I never reboot my G5; I never log out. The ONLY time I experience slow downs is for short periods after fast user switching back to a user that has not been used for quite some time. And that is perfectly understandable, because most of the data for that user has been paged out to disk - it hasn't been used in hours or more. The user is fully back up to speed within 20 seconds. If you have 2GBs of RAM on the same system, there is no excuse for the poor performance you're citing.

And, for the record, OS 9 was a great desktop OS for its time, but OS 9's time passed more than 15 years ago. That's the direct truth. I loved the classic Mac OS quite a bit. But if you think OS 9's overall performance is in any respect superior to OS X's, you're either misinformed or kidding yourself. You think OS 9 could run for weeks smoothly? Only if you're using very simple applications is that true. You're probably conjuring up romantic notions of your old Macs. You may not realize it, but you would never be able to push OS 9 anywhere near the levels you demand of OS X. Don't you remember the classic OS's cooperative multitasking - it was not very cooperative, was it? Don't you remember the total system freezes of the classic Mac OS (their frequency depending on usage and configuration)? I used to just *love* it when OS 9 force quit would fail and crash the whole system, which happened about 45% of the time. Don't you remember having to set static memory allocations for classic applications? Oh the horrors we once endured, yet it was better than the alternative. Realize that OS X is a professional operating system built on the industrial-grade BSD Unix core. In contrast, OS 9, at its core, never received serious upgrades and remained throughout very similar to the first Mac OS that was coded for the low-end, 128K original Mac. We used to forgive the Mac's warts, but now the platform has grown up and we can be honest about the past.

Sorry for the rant, but I felt like belaboring the point. Anyway, there are some things that can slow your Mac down, such as a greedy or frozen application taking up inordinate processor time, or an app that is hogging too much real RAM. You can determine whether you're having a problem with a RAM or processor hog through Activity Monitor. Open Activity Monitor and turn on the regular or floating CPU monitor. Is a significant amount of processor power being used while you're not using the machine? If so, you have a problem with one of your applications. Go to the main Activity Monitor window and sort the chart by % CPU. Figure out which application is hogging the processor(s) and quit or kill it if need be. Also take a look at the Real Memory column to see if an application is taking up too much real RAM. You can determine this by closing a given application and relaunching it. If it was using substantially more RAM than it is using after a fresh relaunch, you may have been doing a lot in the app, or it may have had a memory leak. Sometimes Safari instances that have been running for a long time start to take up too much RAM, especially if you've had a lot of tabs open. Closing the tabs doesn't necessarily free up the memory, but relaunching the application will.

Some people think that their Macs slow down if they've been running too long, or they're too hot. Heat does not reduce performance. The thing that heat can do is cause a computer to crash if the heat is far above acceptable levels. And computers don't get tired. Finally, remember, just because you're having problems, it does not mean that OS X is defective or that other Mac users are necessarily having similar problems. Most of us like assisting one another with given issues. It is, however, annoying when one or a couple of people make general statements about the Mac that are incorrect.

[Tyler McAdams]

Big Mac - Heat does not reduce performance.

Although I doubt it's noticable on a *properly working* G5 you should start Quake 3 on an iMac run the framerate command and let it sit for an hour or 2... you'll see a big difference in the frame rate and it's noticable in the game play as well. Physics... heat causes resistance... electrons flow slower... your system slows down.

[moodymonster]

I've never had memory trouble with PS(7), but Illustrator CS uses 10/15% of the processor even when its doing nothing (as does Flash). Sometimes I've had PS up in the background and not used it for weeks at a time and forgot it was there.

The one program I have found that will eat memory and slow things down over time is Safari. I would suggest quitting and re-opening Safari once a day or less.

As suggested earlier, looking at the activity monitor would help you track down what's going on.

I would have a look at what Safari's doing - it usually takes 60/200MB (real/virtual) or so when first run. After time (I can't remember the real) the virtual can go up to 800MB, usually I quit it there, 'cause it slows stuff down.

On the PB I can tell that Safari's taking up processing 'cause the fan comes on - usually the machine is silent.

10.3.7/PB 17 1.33 1GB

[Big Mac]

Originally posted by Tyler McAdams:
Although I doubt it's noticable on a *properly working* G5 you should start Quake 3 on an iMac run the framerate command and let it sit for an hour or 2... you'll see a big difference in the frame rate and it's noticable in the game play as well. Physics... heat causes resistance... electrons flow slower... your system slows down.

Well, I suppose that properly working is the operative word, yet I am skeptical of the claim that heat alone slows performance. The G5 will run hotter when it is set to Higher Performance, although that's not exactly relevant to the point you're making. The G5 is still running in acceptable temperature levels on the Higher Performance setting, whereas you're claiming in your iMac example that it's running hotter than it should.

The point I'm making is that there isn't a direct correlation between heat and performance in the way you're suggesting. The G5 is not running faster because it's running hotter, but it is running hotter because it is running faster (in High Performance mode). Understand what I am? I'll have to try your iMac Quake 3 example some time. I assume you're talking about a G3 iMac. Higher levels of heat on components could cause unexpected behavior like freezes. But the computational power of a computer does not vary according to its internal temperature. Computers don't run faster when they first start up from a cold boot, and they don't run slower after regular user. I'm not saying heat is never a performance factor. Of course, cooler components do make a difference to enthusiasts who wish to over-clock their systems, for example. Cooler components may mean longer life spans, depending on the degree of differential involved. And, as I said before, hotter machines may end up crashing if the levels are extreme. But in my fifteen years of computer experience, I have not seen an appreciable decline in performance due to extended demanding use. And if the laws of physics affected the function of computers in the way you imply, the computing world would be a world of hurt. Respectfully, it just doesn't work that way.

[Big Mac]

Originally posted by moodymonster:
I've never had memory trouble with PS(7), but Illustrator CS uses 10/15% of the processor even when its doing nothing (as does Flash). Sometimes I've had PS up in the background and not used it for weeks at a time and forgot it was there.

The one program I have found that will eat memory and slow things down over time is Safari. I would suggest quitting and re-opening Safari once a day or less.

As suggested earlier, looking at the activity monitor would help you track down what's going on.

I would have a look at what Safari's doing - it usually takes 60/200MB (real/virtual) or so when first run. After time (I can't remember the real) the virtual can go up to 800MB, usually I quit it there, 'cause it slows stuff down.

On the PB I can tell that Safari's taking up processing 'cause the fan comes on - usually the machine is silent.

10.3.7/PB 17 1.33 1GB

Thank you for the post, moody. Most properly coded programs on OS X should go to sleep when not in use and take up zero processor time. If an application is using processor time while not being used, that application is almost certainly coded poorly. Now sometimes an application can be processing something that you aren't aware of. More complex web pages can cause Safari to take up time in the background, and sometimes Safari has to be quit after that.

In reference to memory usage, the most recent public version of Safari seems to be much better about memory leaks. I haven't had to quit Safari as often as I did months ago. Sometimes certain web pages can cause Safari memory problems even after they've been closed, but other than that Safari has been pretty responsible in terms of memory usage. Check up on its memory usage periodically and quit if it's getting out of hand. Right now my user's Safari is taking 80MBs real memory, a background user's Safari has 40MBs, and another background user (which has been used least recently) has Safari occupying only 20MBs.

Comparing my iBook 466 to my DP 2.0, I have found that a faster Mac will compensate for less RAM. For example, for awhile I ran my G5 with only the stock 512MBs and did not experience too much sluggishness despite using multiple logins. Of course, in addition to the gargantuan difference in CPU/GPU performance, my G5 also has a faster, roomier drive for virtual memory. Not exactly a fair comparison, but it's the best I can offer. Btw, with only a GB of RAM I keep four users open all of the time. And I only restart to install updates. Right now I have 28 days uptime; before my last software update I was up to 40 days. I don't mean to boast. The point is, the G5 is a powerhouse, and OS X is rock solid.

[Tyler McAdams]

Originally posted by Big Mac:
Well, I suppose that properly working is the operative word, yet I am skeptical of the claim that heat alone slows performance. The G5 will run hotter when it is set to Higher Performance, although that's not exactly relevant to the point you're making. The G5 is still running in acceptable temperature levels on the Higher Performance setting, whereas you're claiming in your iMac example that it's running hotter than it should.

The point I'm making is that there isn't a direct correlation between heat and performance in the way you're suggesting. The G5 is not running faster because it's running hotter, but it is running hotter because it is running faster (in High Performance mode). Understand what I am? I'll have to try your iMac Quake 3 example some time. I assume you're talking about a G3 iMac. Higher levels of heat on components could cause unexpected behavior like freezes. But the computational power of a computer does not vary according to its internal temperature. Computers don't run faster when they first start up from a cold boot, and they don't run slower after regular user. I'm not saying heat is never a performance factor. Of course, cooler components do make a difference to enthusiasts who wish to over-clock their systems, for example. Cooler components may mean longer life spans, depending on the degree of differential involved. And, as I said before, hotter machines may end up crashing if the levels are extreme. But in my fifteen years of computer experience, I have not seen an appreciable decline in performance due to extended demanding use. And if the laws of physics affected the function of computers in the way you imply, the computing world would be a world of hurt. Respectfully, it just doesn't work that way.

No... that's wrong... the cooler the system is the faster it will run. that's why overclockers use liquid nitrogen to cool the systems... Overclocking the cpu ups the heat output and will burn the chip up if it even starts up otherwise. The cooler the system *stays* over a long period of time the faster whatever the job is your running will complete... which is exaclty why your fans come on in your G5. Perfomance mode produces a lot of heat and needs to be cooled off. PowerPC chips are known to not put off a lot of heat but any chip that is doing work produces heat and thus needs cooling. Just like a motor in your car when it gets hot it needs better cooling... your fan starts to spin at a higher rate and/or your intercooler sprays water on the fins to disipate heat. Any "engine" has a side effect of heat when producing work... it's basic physics. The heat is what slows down everything and that's why the *efficiency* of your heat dissipation is a key issue to the durability of your cpu/motor/whatever. And since an iMac has extremely poor cooling it's easy to see the results by just putting on quake and letting it go until your system stops to a dead crawl or freezes up completely. If you reboot the machine it will play quake again but for a lesser time than before... sooner or later it will be at a crawl for everything and you'll just have to shut it down completely. Where this can be seen the best in the compuer world is in render nodes that are overworked... the heat becomes too much since the systems are at 100% cpu utilization for an extremely extended amount of time... the cluster locks up in the middle of the night and the sad SA gets paged out and has to go in to work and reboot the entire cluster.

[Boondoggle]

Originally posted by Tyler McAdams:
No... that's wrong... the cooler the system is the faster it will run. that's why overclockers use liquid nitrogen to cool the systems... Overclocking the cpu ups the heat output and will burn the chip up if it even starts up otherwise. The cooler the system *stays* over a long period of time the faster whatever the job is your running will complete... which is exaclty why your fans come on in your G5. Perfomance mode produces a lot of heat and needs to be cooled off. PowerPC chips are known to not put off a lot of heat but any chip that is doing work produces heat and thus needs cooling. Just like a motor in your car when it gets hot it needs better cooling... your fan starts to spin at a higher rate and/or your intercooler sprays water on the fins to disipate heat. Any "engine" has a side effect of heat when producing work... it's basic physics. The heat is what slows down everything and that's why the *efficiency* of your heat dissipation is a key issue to the durability of your cpu/motor/whatever. And since an iMac has extremely poor cooling it's easy to see the results by just putting on quake and letting it go until your system stops to a dead crawl or freezes up completely. If you reboot the machine it will play quake again but for a lesser time than before... sooner or later it will be at a crawl for everything and you'll just have to shut it down completely. Where this can be seen the best in the compuer world is in render nodes that are overworked... the heat becomes too much since the systems are at 100% cpu utilization for an extremely extended amount of time... the cluster locks up in the middle of the night and the sad SA gets paged out and has to go in to work and reboot the entire cluster.

I think what you mean is the cooler a system the faster it "can" run, if over-clocked, not "will" run all by itself, without over-clocking.

If a given chip is running in spec i.e. within it's nominal clock speed and normal temperature range I do not believe the CPU temp has a significant impact on speed. You're implying that a hotter chip there is a measurable and significant slowing of the electrons, even within it's operational temperature range. I don't believe this is the case. However as chips heat up past the acceptable range they are subject to failure, which is another matter, and that is why over-clockers use extra cooling methods. The cooling does not make the chip faster, the cooling allows the extra heat of higher speed to be dissipated fast enough to avoid chip failure.

Liquid Nitrogen is used for serious over-clocking because liquid nitrogen is highly efficient at reducing the temperature. It is hardly practical however as far as I can tell...
http://www6.tomshardware.com/cpu/20031230/

In your iMac example you suggest that the cooling system is inadequate to dissipate heat indefinitely at full CPU usage to maintain a temperature range that does not result in failure. That may be the case.

A G5's fans come on to prevent the chips from getting too hot and failing or sustaining damage at high temperatures. The fans do not make the chips significantly faster at a given GHz

I don't believe extra cooling makes a practical system significantly faster at a given Ghz as long as you're not at super-conducting temperatures. And you're not.

[moodymonster]

Originally posted by Big Mac:
snip�OS X is rock solid. [/B]

damn right!

Its 'cause the rest of it does just 'work', that anything less stands out.

Safari has got better over the last few months, and I'd agree that the web pages viewed will affect the amount of processing time required. Safari does seem to not let go of VM once its got it, whereas other programs do.

I run multiple accounts as well, usually just one or two up at a time. Don't notice any slowdown while doing this.

The last time I rebooted was for the 10.3.7 update, also running as a MySQL/PHP/website testbed thingy+bit of video/PS/Flash etc

I do know that processors can't use all their capability at once 'cause of heat though. I don't think that would affect the performance. Freescale say their chips will run @ 105�C for 10yrs

http://www.freescale.com/webapp/sps/...=DRPPCDUALCORE

[Big Mac]

Originally posted by Tyler McAdams:
No... that's wrong... the cooler the system is the faster it will run. that's why overclockers use liquid nitrogen to cool the systems... Overclocking the cpu ups the heat output and will burn the chip up if it even starts up otherwise. The cooler the system *stays* over a long period of time the faster whatever the job is your running will complete... which is exaclty why your fans come on in your G5. Perfomance mode produces a lot of heat and needs to be cooled off. PowerPC chips are known to not put off a lot of heat but any chip that is doing work produces heat and thus needs cooling. Just like a motor in your car when it gets hot it needs better cooling... your fan starts to spin at a higher rate and/or your intercooler sprays water on the fins to disipate heat. Any "engine" has a side effect of heat when producing work... it's basic physics. The heat is what slows down everything and that's why the *efficiency* of your heat dissipation is a key issue to the durability of your cpu/motor/whatever. And since an iMac has extremely poor cooling it's easy to see the results by just putting on quake and letting it go until your system stops to a dead crawl or freezes up completely. If you reboot the machine it will play quake again but for a lesser time than before... sooner or later it will be at a crawl for everything and you'll just have to shut it down completely. Where this can be seen the best in the compuer world is in render nodes that are overworked... the heat becomes too much since the systems are at 100% cpu utilization for an extremely extended amount of time... the cluster locks up in the middle of the night and the sad SA gets paged out and has to go in to work and reboot the entire cluster.

No, Tyler. You misunderstood my point, and you still don't understand the relationship between heat and performance. I specifically said in my post that overclockers want cooler components because lower temperatures allow them to increase overclock the system. Overclockers put in sophisticated cooling systems to keep the components cool when they are overclocked, or else they could overheat. It all depends on the components and temperatures involved. Every so often a semiconductor giant like Intel will demo a super clocked version of a current generation chip, but in order to get the clock speed that high they need to put it in an refrigeration unit.

But all of that has nothing at all to do with normal systems that aren't overclocked. And whether or not a system is overclocked, it will either run at a specific speed or it will crash. Computers don't naturally gain or lose speed from temperature levels. Modern computers do use software controls to adjust clock rate and fan speed in accordance with temperature. But if you think computers spontaneously, unintentionally gain or lose performance depending on temperature variations, you're just plain wrong. That's what you're implying with your iMac example. I apologize for having to be blunt, but we're apparently experiencing a failure to communicate.

[P]

Big Mac is mostly right... or rather he is right for CPUs a few years old. These days CPUs can clock down if they notice that they're getting to hot. For notes on this from the x86 side of things, see the following:

http://www6.tomshardware.com/cpu/20010917/index.html

Other noteworthy articles here:
http://www6.tomshardware.com/cpu/20041114/index.html
http://www6.tomshardware.com/column/20011029/index.html

I can't tell you how it work on our side of the fence, but it's very likely he same deal. I do know that every G3 and G4 has a thermal sensor to read out the die temperature. It is wildly inaccurate, and the latest G4s don't allow reading it out, but it is acurate in determining changes in temperature. If the the temperature keeps rising, the CPU could presumably clock itself down. I think that may be what you are seeing. Of course, Apple and IBM are both into integrating stuff like this into the OS, so maybe it happens there instead of in the CPU logic for the G5s, but the result should be the same.

[Big Mac]

P, please read what I wrote again, because I covered your point:

I wrote:
Modern computers do use software controls to adjust clock rate and fan speed in accordance with temperature.

Tyler is saying that computers unintentionally, spontaneously get slower as they got hotter. That's the point I'm refuting. Take a look at Boondoggle's post - we're saying the same thing.

[Spheric Harlot]

Big Mac mentioned that modern designs employ temperature sensors that will clock down the CPU if it runs too hot, but his point is that this is controlled by complex circuitry, not some inherent physical property.

[Tyler McAdams]

Hold on, this does not have anything to do with the clock speed of your computer. I'm not saying the clock speed changes at all. What I'm saying is that the work you can get out of a processor over time becomes less due to the heat building up. It is a fact that copper looses conductivity the hotter it gets... you can google that. The electrons flow slower... Because they flow slower the work you can get out of it gets less.

Big Mac all you have to do is put Q3 on your G5 and let it go for a week... put the frame rate command in and take a look at the frame rates... throughout the entire week you'll notice that the framerates drop. Your system will still be clocked at 2ghz or whatever but the frames per second will drop due to the heat.

[Chuckit]

If it is going slower, how can it still be doing 2 billion cycles a second? And conversely, if it's still doing the same number of cycles a second, how can it be going slower?

[Tyler McAdams]

Originally posted by Chuckit:
If it is going slower, how can it still be doing 2 billion cycles a second? And conversely, if it's still doing the same number of cycles a second, how can it be going slower?

I think that in most cpus today the clock speed is artifically governed. But just because the clock is set to 500mhz does not mean that you will see the same performance since there are more variables to performance than clock speed. Have you ever heard of cas ratings for ram? The ram can be at 500mhz but have a cas rating of 3 and still not be as fast as 500mhz ram that has a rating of 2. I'm not exaclty sure what slows down in this case but you'll probably not see a decrease in clock speed... but the actual hardware is slowing down because of the resistance in electron flow increasing.

I'm sure Big Mac has somewhat of a point here but I can't just dissmiss away the results of my own experience and 5 years of Physics classes. Perhaps the older G3 models are different but I'll put down a year's paycheck that my iMac test is easily repeatable for anybody who tries it.

http://www.newton.dep.anl.gov/askasc...0/phy00414.htm

[Spheric Harlot]

Originally posted by Tyler McAdams:
I think that in most cpus today the clock speed is artifically governed. But just because the clock is set to 500mhz does not mean that you will see the same performance since there are more variables to performance than clock speed. Have you ever heard of cas ratings for ram? The ram can be at 500mhz but have a cas rating of 3 and still not be as fast as 500mhz ram that has a rating of 2. I'm not exaclty sure what slows down in this case

Well, you can be absolutely sure that in this case, it AIN'T the heat.

And no, a CPU's processing capabilities per clockcycle are exactly defined and do not vary, not with temperature, nor weather, nor time of day.

With all due respect to your physics classes: computer design DOES tend to respect the basic laws of thermodynamics, regardless of your personal experience/perception.

There are an awful lot of aspects governing the perceived speed of your computer, but the temperature isn't one of them, unless things get so hot that the hardware starts malfunctioning or the clock speed is artificially throttled down.

-s*

[Tyler McAdams]

Originally posted by Boondoggle:
I think what you mean is the cooler a system the faster it "can" run, if over-clocked, not "will" run all by itself, without over-clocking.

If a given chip is running in spec i.e. within it's nominal clock speed and normal temperature range I do not believe the CPU temp has a significant impact on speed. You're implying that a hotter chip there is a measurable and significant slowing of the electrons, even within it's operational temperature range.

Well it's actually a mix of the 2... the cooler a system the faster it can be overclocked yes... to a point at which super conducting starts and then you can not do any more cooling for overclocking.

Within a normal operational range you are correct, no you will not be able to detect a difference in output... but if you get past that it will be noticable. It will not result in "failure" in that the chip will not burn up as long as you're not in a badly ventilated room. You will burn the chip up, however, if you have an extreme spike in voltage or temp.

But to sum up your first statement it is actually a mix of both and because you probably never push the envelope of the normal temp range of your chip you will not be able to detect the difference... the only real way to decet is it to run your system at 100% cpu utilization for and extended time period... your iMac or laptop is not made to be used like that and becasue of that you can actually see the change.

[Tyler McAdams]

With all due respect to your physics classes: computer design DOES tend to respect the basic laws of thermodynamics, regardless of your personal experience/perception.

Ohm's Law: V=IR is one if the basic laws of thermodynamics and that's exactly what I'm talking about.

There are an awful lot of aspects governing the perceived speed of your computer, but the temperature isn't one of them, unless things get so hot that the hardware starts malfunctioning or the clock speed is artificially throttled down.

It's obvious to me that I can't prove this to you but that is a governing factor.

[Spheric Harlot]

Originally posted by Tyler McAdams:
It's obvious to me that I can't prove this to you but that is a governing factor.

It WOULD be, were not the clock speed the limiting factor.

Clock speed is artificially set below the maximum possible precisely to ELIMINATE heat as a relevant factor from the equation.

Anyway: a processor will fit the same amount of work into a clockcycle regardless of temperature.

What you're claiming is akin to claiming that a person doing long division on paper will be able to fit less digits on the paper as the ambient temperature rises. Which is wrong. He will get increasingly unreliable, but he will still be able to utilize his resources as effectively as when it's cool.

-s*

[Spheric Harlot]

Originally posted by Tyler McAdams:
Hold on, this does not have anything to do with the clock speed of your computer. I'm not saying the clock speed changes at all. What I'm saying is that the work you can get out of a processor over time becomes less due to the heat building up. It is a fact that copper looses conductivity the hotter it gets... you can google that. The electrons flow slower... Because they flow slower the work you can get out of it gets less.

Since the processor is still running at the same clock speed, the slower electrons must mean, by your theory, that they're not fast enough to supply the processor with the information it needs to be able to actually run each cycle, which means the processor must "misfire" constantly, since it's still running a fixed number of operations per clock cycle, and, as you say, the clock speed does not change.

In computers, that doesn't slow down your machine, that crashes it.

Pure and simple.

Don't mix analog theory with digital processing - they are NOT the same thing.

[Tyler McAdams]

Hey fellas all I know i that I'm looking at my iMac right now that's been on since this debate started and it's frame rate has dropped 15fps already. What this means to you I'm not sure... Anyway, I've got to get to work.. I'll post a little later.

[Big Mac]

I'm surprised this conversation has continued. Tyler, I do believe you when you say you're good in Physics courses. But Physics formulas alone don't make you a computer engineer. Computer performance simply is not determined in such a random way. No one else has offered an opinion that agrees with yours, and you won't be able to find support for your claim in any engineering or computing text. The best you can offer is anecdotal experience with an iMac. I am certain that whatever drop in performance you are experiencing on your iMac is due to purposeful design - bus slewing perhaps - rather than unintended physical limitations. Oh, regarding your comment about RAM timings, the CAS ratings aren't assigned or governed by chance and constant variation, as you seem to imply. Computers work at a definite, engineered, measurable rate and continue working at the level they were designed or modified to work at for their lifetimes. If you see significant performance disparities measured on the same machine, the differentials cannot possibly be ascribed to unintended physical consequences. Why do you persist in asserting something that we're telling you is absolutely not the case? It's okay to admit you're wrong. I'm not trying to be an arsehole about it, but come on...

[kevs]

Could someone please explain what virtual memory means in activity monitor? I thought virtual memory was a thing of the past with OS9.

Right now , for example, I've got lot of things listed using 200 mb of VM and they are just in hide mode.

Photoshop is in hide mode-- no photos up, nothing being done there -- and it using 117 of real memory/ 433 virtual.

Also, I was burning a DVD with Toast today, and Toast was not listed, but at the top it said "window server" twice, both at 22% CPU. Was that Toast?

And for evaluating if your ram is sufficient what others things should you look at in Activity monitor, and how would you know if your ram is sufficient?

And finally, does adding ram makes things happen faster? thanks!

[Tyler McAdams]

Here we are:
http://www.xbitlabs.com/articles/cpu/print/p4-temp.html

Read down the to the middle:
"There are a lot of erroneous ideas about the Thermal Control Unit mechanism. The most frequenctly occurring delusion implies that Pentium 4 reduces its nominal clock-rate when it gets overheated. That is, if it works at 2.2GHz, it will drop its frequency to 1.8GHz or even lower. It is not quite correct if we talk about the nominal CPU frequency set by its frequency generator. Let's make it clear now. First we have to recall the way the clock-rate is generated in Pentium 4 2.8GHz CPU.

Suppose the mainboard sends 133MHz frequency to the CPU. This frequency is multiplied by a coefficient, which equals 21 in Pentium 4 2.8GHz. This 2.8GHz frequency is the nominal frequency mentioned in the Pentium 4 marking as read by programs like WCPUid. This frequency determines the working frequency of the processor arithmetic units. Thermal Control Unit can affect this frequency. When the temperature is normal, ALUs will receive the same 2800MHz frequency. But when the CPU temperature is above a certain value, Thermal Monitor sends its PROCHOT# signal to enable Thermal Control Circuit. The latter modulates the frequency sent to the CPU and determines how many clock cycles should be omitted to reduce CPU heat dissipation. The modulation of the clock signal sent to the CPU is shown in the following diagram:

As a result, some clock cycles may be excluded from the normal 2.8GHz. In other words, they are set by the CPU multiplier unit, but left out by the null cycles control system which is enabled by the PROCHOT# signal. So we will have a lower resulting frequency sent to the CPU ALUs. Of course, CPU performance drops down as well as its heat dissipation, although the mainboard and internal clock-rate generator will keep producing the same 2.8GHz. Intel claims the resulting frequency can be as low as 30-50% of the nominal, depending on the CPU model.

However, as the temperature goes down, the Thermal Control Circuit unit will start returning the CPU to its nominal working mode, by reducing the number of null clock cycles and this way increasing the end frequency of the CPU (here we mean the internal frequency, which is set by the processor multiplier unit and is modulated with reference to the Thermal Control Circuit). As soon as the core temperature goes down by about 1oC (the so-called temperature hysteresis number) below the critical value, Thermal Monitor stops sending PROCHOT# signal. After that, Thermal Control Circuit will stop generating null cycles and the effective (resulting) frequency will equal the nominal one - 2800MHz in our case."

Maybe this is what is going on in the iMac.

[Big Mac]

You're significantly changing your argument Tyler, and if you don't recognize the difference between your previous claims and what you're asserting now this conversation will get even more difficult. Before you were claiming that variations in speed were due to unintended physical consequences rather than purposeful design. But the article you cite speaks of purposeful engineering choices. Please, this is really getting ridiculous.

[Tyler McAdams]

Originally posted by Big Mac:
Oh, regarding your comment about RAM timings, the CAS ratings aren't assigned or governed by chance and constant variation, as you seem to imply.

That was just an example of performance increase/decrease with no coorelation to clock speed... I was not trying to imply it was random.

[Tyler McAdams]

Originally posted by Big Mac:
You're significantly changing your argument Tyler, and if you don't recognize the difference between your previous claims and what you're asserting now this conversation will get even more difficult. Before you were claiming that variations in speed were due to unintended physical consequences rather than purposeful design. But the article you cite speaks of purposeful engineering choices. Please, this is really getting ridiculous.

I was saying that it was because of heat but this article seems to show what it was. I'm not changing my "argument" I'm simply adding to the discussion. Nobody here besides P has cited any techical references and that's what is needed to clear up what exactly is going on. This however still could be proof of what exaclty is slowing down the orginal posters machine.

[Big Mac]

I never said clock speed is the only performance factor in the mix. I simply said that you cannot ascribe performance changes to spontaneous, unintended physical consequences. You can't start arguing an entirely different matter in order to claim you were right. Or are you admitting you were wrong?

I really didn't want this thread to be derailed.