Feature request
 

I'm realizing my processor is actually powerful enough so that, while running prime95, it cannot charge in an appreciable amount of time. I use my laptop for projection as well as normal use.

So my request is, create an option that says "Do not run Prime95 when computer battery is below 50%", or something along those lines.

Short answer:
There is already such an option in menu Options/Prefrrences, you have a check box saying "run when on battery power". Just take it off (uncheck it).

"TL;DR" version:
[B]Unless you have a very expensive battery[/B] which holds your laptop on full CPU throttle for many hours (like those used for helicopters, of for nissan leaf or toyota prius cars, think about something about 8 to 50 times more expensive than whole your laptop, which can supply it in full for 20-50 hours or so), [B]it makes no sense to think about running P95 on the first 50% of the battery[/B]. Physically, the battery goes down exponentially. If your battery can power your laptop for few hours, say 3 hours, when you work in winword, powerpoint, presenter, whatever, then reaching 50% of battery happens somewhere at say 2 hours point. Using P95 it will short the time to one hour, or even less, 40 minutes (until the computer dies because of no battery, or battery under the set limit). So, in fact, if you run P95 until the battery is 50%, then use the computer "normally" until it dies because of no battery, you will be able to [B]run P95 for about 20 minutes[/B], and then you still can use the computer "normally" for one more hour, or less. Does it make any sense? What is your benefit from it, beside of... shortening your battery life? (you will need to change it after 2 years or less! because of the high stress caused by fast discharging)

[B]My advice is:[/B] use the computer for p95 crunching [B][COLOR=Red]only when it is powered on by the wall socket (plugged in)[/COLOR][/B]. If possible, try to avoid getting it hot (like using "Throttle=" parameter, see undoc.txt, clean it periodically to remove the dust clogs from the fans, etc.), and if it however gets somehow hot, then [B]take the battery out completely during this time![/B]

This is no joke, look on wikipedia for lithium-ion batteries, their life time shorten a lot if they get heated up). I do this for a ten-years-old laptop which is used as router in my house and is mostly (99% of the time) plugged in - i took off the battery (if there is no electricity anyhow I have no internet signal because the other things are powered from the same source) and is doing some P-1 stuff too. The battery is used occasionally, if I need the laptop portable, or when I plug it periodically for charging, like every 6-8 weeks or so. As a result, that batter is still "like new" after about 7-8 years. The original battery died after the first 2-3 years (during that time the battery was inside always, the laptop was plugged in most of the time, at the office at home, and unplugged only during transportation, and of course it was continuously running P95, getting quite hot). For the other laptop which I still use for about 7 years, I have changed 2 batteries already, now I am at the third. The work I do can't stand power failures, so I can't take the battery out. My daughter and my wife laptops are 3 respective 2 years old, and they started having problems with the batteries, diminished capacity, decreasing the "work on battery" time to about half. Of course, both are getting somehow hot when plugged in, from obvious reasons of distributed computing and me being lazy to clean them more often :razz: and the battery is never taken out. Try convince my lass that she has to take the battery out when she is on facebook, and see how many scratches you can count on your face after that.

I have posted things in the past about damaged batteries and bended fans in laptops because of heat. Some people say is not possible, well, your mileage may vary, your money, your fun.

[QUOTE=LaurV;311537]
[B]My advice is:[/B] use the computer for p95 crunching [B][COLOR=Red]only when it is powered on by the wall socket (plugged in)[/COLOR][/B].[/QUOTE]

Umm... I think you misunderstand him. He said that even when his laptop is plugged in, Prime95 causes the power draw to go so far up that he can't charge the batteries.

[QUOTE=LaurV;311537]Short answer:
There is already such an option in menu Options/Prefrrences, you have a check box saying "run when on battery power". Just take it off (uncheck it).

"TL;DR" version:
[B]Unless you have a very expensive battery[/B] which holds your laptop on full CPU throttle for many hours (like those used for helicopters, of for nissan leaf or toyota prius cars, think about something about 8 to 50 times more expensive than whole your laptop, which can supply it in full for 20-50 hours or so), [B]it makes no sense to think about running P95 on the first 50% of the battery[/B]. Physically, the battery goes down exponentially. If your battery can power your laptop for few hours, say 3 hours, when you work in winword, powerpoint, presenter, whatever, then reaching 50% of battery happens somewhere at say 2 hours point. Using P95 it will short the time to one hour, or even less, 40 minutes (until the computer dies because of no battery, or battery under the set limit). So, in fact, if you run P95 until the battery is 50%, then use the computer "normally" until it dies because of no battery, you will be able to [B]run P95 for about 20 minutes[/B], and then you still can use the computer "normally" for one more hour, or less. Does it make any sense? What is your benefit from it, beside of... shortening your battery life? (you will need to change it after 2 years or less! because of the high stress caused by fast discharging)

[B]My advice is:[/B] use the computer for p95 crunching [B][COLOR=Red]only when it is powered on by the wall socket (plugged in)[/COLOR][/B]. If possible, try to avoid getting it hot (like using "Throttle=" parameter, see undoc.txt, clean it periodically to remove the dust clogs from the fans, etc.), and if it however gets somehow hot, then [B]take the battery out completely during this time![/B]

This is no joke, look on wikipedia for lithium-ion batteries, their life time shorten a lot if they get heated up). I do this for a laptop which is used as router in my house and is mostly (99% of the time) plugged in (if there is no electricity anyhow I have no internet signal because the other things are powered from the same source) and is doing some P-1 stuff too.

I have posted things in the past about damaged batteries and bended fans in laptops because of heat. Some people say is not possible, well, your mileage may vary, your money, your fun.[/QUOTE]
I think you've misunderstood. I don't run Prime95 on battery power. I run it only plugged in. I have that setting set.

I mean, if I leave my laptop plugged in starting from a low battery point (say 14%) then to get to 100%, the charging process does not produce charge fast enough to charge it in an entire school day.

My suggestion is: Add an option so that when plugged in, prime95 doesn't run unless the battery charge is high enough. That would allow people like me who run their laptops hard to be able to charge fast enough.

Oh! Sorry for misunderstanding. To solve your problem use Throttle. You can find a very good compromise between the charging time and working time, and this will also protect your laptop (less heat) and battery.

For example, try doing some experiments and see how long you battery need to charge to 50%, how long your computer is plugged in, how long it works on battery. Then set the throttle in such a way than you do the same amount of work per 24 hours as you would do if "stop p95, charge battery, start p95 at 100% throttle".

[QUOTE=LaurV;311541]Oh! Sorry for misunderstanding. To solve your problem use Throttle. You can find a very good compromise between the charging time and working time, and this will also protect your laptop (less heat) and battery.

For example, try doing some experiments and see how long you battery need to charge to 50%, how long your computer is plugged in, how long it works on battery. Then set the throttle in such a way than you do the same amount of work per 24 hours as you would do if "stop p95, charge battery, start p95 at 100% throttle".[/QUOTE]

Perhaps, it is just simpler to just run fewer workers? I mean, with all the throttling and everything, the aggregate thruput might not be all that different.

Yes, run one less thread to save power!
 

You can monitor CPU power draw on some cpus using CoreTemp (and others). On an I5 2500k at stock speed, running all four threads, P95 forces the cpu to draw 57 watts. Reduction of just one worker thread reduces draw to 50 watts.
Reduction of two threads reduced draw to 38 watts.

[url]http://www.alcpu.com/CoreTemp/[/url]

[QUOTE=swl551;311875]You can monitor CPU power draw on some cpus using CoreTemp (and others). On an I5 2500k at stock speed, running all four threads, P95 forces the cpu to draw 57 watts. Reduction of just one worker thread reduces draw to 50 watts.
Reduction of two threads reduced draw to 38 watts.

[URL]http://www.alcpu.com/CoreTemp/[/URL][/QUOTE]

Really?
I thought you couldn't measure power consumption internally?

You can measure CPU wattage on Sandy Bridge.
 

No knowledge of any other platforms but for sure on Sandy Bridge

Those options produce some of the desired effects but with significant drawbacks. Cutting a worker thread reduces the throughput by 25% (presumably). Throttling, no matter what calculations I do to take into account 24 hours or even just an entire school day, will still keep my battery low through, say, a room change or an early morning meeting, and also reduces throughput even when battery power is above 50%. Letting prime95 start only when the battery power is above a certain amount is stil by far the better option.

A work around if p95 does not get the feature.
 

Here is a script using WMI to get battery status. [url]http://www.planet-source-code.com/vb/scripts/ShowCode.asp?txtCodeId=8092&lngWId=4[/url]

There is probably a converted script to use PowerShell out there and seems more correct if you using Win7.

You could code up a mini battery monitor with the script and stop/start P95 accordingly

or if you want to access the information at a lower layer
[url]http://msdn.microsoft.com/en-us/library/windows/desktop/aa394074(v=vs.85).aspx[/url]

Code up a mini battery monitor? Funny :smile: This still defeats the point of all the automaticity that comes with prime95.

[QUOTE=swl551;311914]Here is a script using WMI to get battery status. [URL]http://www.planet-source-code.com/vb/scripts/ShowCode.asp?txtCodeId=8092&lngWId=4[/URL]

There is probably a converted script to use PowerShell out there and seems more correct if you using Win7.

You could code up a mini battery monitor with the script and stop/start P95 accordingly

or if you want to access the information at a lower layer
[URL]http://msdn.microsoft.com/en-us/library/windows/desktop/aa394074(v=vs.85).aspx[/URL][/QUOTE]

Coding up a script is a proposed work-around.
 

Yes, I realize it does cause you to withdraw your "feature request", but I thought I'd give you some ideas for possible solutions until your request is implemented.

I'd like to make a second push for this feature request. My laptop charges pretty well but I turned Prime95 back on while I was teaching for two periods, plugged in, and my battery power went only from 15% to 31%. That's really not enough for me to simply let Prime95 run in the background while plugged in because it doesn't let the laptop charge fast enough. I suspect others have the same issue.

[QUOTE=dominicanpapi82;318143]I'd like to make a second push for this feature request. My laptop charges pretty well but I turned Prime95 back on while I was teaching for two periods, plugged in, and my battery power went only from 15% to 31%. That's really not enough for me to simply let Prime95 run in the background while plugged in because it doesn't let the laptop charge fast enough. I suspect others have the same issue.[/QUOTE]

Out of curiosity, what is your CPU?

[QUOTE=axn;318145]Out of curiosity, what is your CPU?[/QUOTE]

AMD A6-3420M APU with Radeon HD Graphics 1.5GHz
4 gigs ram
Windows 7 Home Premium, 64-bit

[QUOTE=dominicanpapi82;318143]I'd like to make a second push for this feature request. My laptop charges pretty well but I turned Prime95 back on while I was teaching for two periods, plugged in, and my battery power went only from 15% to 31%. [/QUOTE]

Today is your lucky day. Add BatteryPercent=90 to prime.txt. This will tell prime95 to halt until battery level reaches 90%. This feature has always been available in Windows but is not in undoc.txt (I'll fix that).

[QUOTE=Prime95;318164]Today is your lucky day. Add BatteryPercent=90 to prime.txt. This will tell prime95 to halt until battery level reaches 90%. This feature has always been available in Windows but is not in undoc.txt (I'll fix that).[/QUOTE]
You're my favorite person for the next 23 hours and 59 minutes. Congratulations!

You are also lucky you could charge at all. I am pretty some people have lost charge when using prime95.

a comment and a question
 

Comment: I am running Prime95 on several laptops (always only on AC power) and have never had an issue with the charger not keeping up. Could it be you need a new/different battery?

Question: does anyone have experience using Throttle=n?
I just tried it on my i5-2520 Laptop (Dual core Sandy)and found the results "interesting"
- no Throttle: temp stable at 77&84; iteration time .023
- Throttle=80: temps fluctuate between 65 and 88; iteration time: .021
Why are my iteration times better? Is the ocassional 65 degrees really enough to have it run faster overall? Could it be only timing during the 80% of time it is running?
Might constant temp changes of 20 or more be harder on the hardware?
Could my cooler be being fooled that it gets hotter at times with throttle?

I put both of my laptops on Zalman fan bases. The big, slow fans have a soothing sound. I am using the ZM-NC3 model.

[QUOTE=petrw1;328062]Why are my iteration times better? Is the ocassional 65 degrees really enough to have it run faster overall? Could it be only timing during the 80% of time it is running?
Might constant temp changes of 20 or more be harder on the hardware?
Could my cooler be being fooled that it gets hotter at times with throttle?[/QUOTE]
That is normal. Nothing magic, your CPU is not "faster" at 65C, but it does "hardware" throttle (thermal protection jumps up and turns on) upper, when you don't use the "software" throttle of p95. The one employed by p95 is just "work x% of one time slice, stay the rest". Your CPU has enough time to get hot and cool in cycles (I mean visible cycles, one second or half second each period). "Time slice" depends of how fast p95 can fill and empty its caches. It could be nice to be possible to adjust that time slice (period), beside of the percent, but that depends of how fast the data is loaded/tasks are switched, it may not be so nice to stop it and clear the cache too often, therefore is hardware coded inside of P95, you can only specify the percent. P95 will chose that time slice depending on the work it has to do, and the percentage.

top producers: life time and 365 buttons
 

With the temporary(?) Removal of the ability to choose a date range on the top producers report ... Which I am not opposed to ... We have 2 choices:
- accept the default of 365 days
- enter a start date like 2008-01-01 to request life time.

Could I request there be radio buttons to choose either of the above.
My fat fingers have trouble typing the date on my black-berry.

Thanks