• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

a serious question

Ice Czar

Ice Czar

Inscrutable
Joined
Jul 8, 2001
Messages
27,174
OK well a few of you may know Im not a gamer
(barring my flash version of Asteroids :p )

so I need some help here, as Im calculating alot of "realworld" power consumption senerios in the PSU forum

when your gaming, what is the maximum CPU utilization you see?
for a given CPU
(figuring the GPU is at 100% or near it)

and if you could note approximately what other processes are running
or if your multitasking
 
When I am playing games, such as CS:1.6 the CPU usage is around 50-75 percent. My compeuter specs are in my sig. I can almost never hit 100?% CPU usage and keep it at that for even a minute. When multitasking, I once tried running Norton AntiVirus, Some benchmarks, Ad-Aware, and Spybot at the same time. The CPU usage hit 100% and then dropped to 90 % and never stayed at 100% for a long time. I think this is due in part to HT..but I dont know.

I dont know if that helps but maybe I can later gather some more info on the CPU usage.... :)
 
Mine runs a little higher than that. JK2 pulls, averaged, over 20 min of play. 68%.. HL2 76%.. xp 3000 and a 9800pro both slightly overclocked.. It pegs to 100% a number of times. but doesn't stay there for anywhere near a minute at a time. Prolly just while loading the next level.. According to rescource meter anyway..31 processes running. zone alarm, avg antivrus, Steam, (only while I was running HL2), otherwise just the reguluar background stuff.. In truth, cpu usage is going to vary alot, even on nearly identical comps.. Background programs, game settings, and general system overhead, is going to vary alot, from one user to another.
 
Only time cpu hits 100% is when windows alt-tabs between graphic-games and regular windows in my experience.

~Adam
 
GORANKAR said:
In truth, cpu usage is going to vary alot, even on nearly identical comps.. Background programs, game settings, and general system overhead, is going to vary alot, from one user to another.
Click to expand...

granted, but Im simply trying to figure out if the worse case "real world" senerio
would be that both the CPU and GPU hit 100% at the same time ever ;)

or what a reasonable 100% GPU and X% CPU value might be for a few assorted configs
 
well 100% CPU by itself is pretty easy to get
just need a borked driver for that :p
 
I'm curious now. I'm going to see if I can get my cpu usage and gpu usage to a 100% for an extended time.. Perhaps HL2 with the eycandy to max @1600x1200 with 6xaa and 16xaf while running folding @ home, and converting a DVD to divx.. Will let you know.. HalfLife2 @ 4fps here I come. Have no idea how to measure gpu usage. I'll just assume that HL2 @ those settings will be murdering my poor 9800
 
with modern CPUs hitting 7 to 9 amps and some cards at 4 amps on the +12V rail
knowing if they will coincide at 100% would be very useful ;)

even if its just for a few seconds (which would exceed most "peak" amp ratings
 
running axis & allies during a huge tank battle seems to do the trick when it comes to 100% cpu
 
I did it and, my cpu usage was @ 85 to 100%, for the 15 minutes I did it for.. It rarely went lower. Like I said earlier, I can only assume my 9800 was maxed @ those settings. I actually peaked @21fps. My average was close to 4fps. I'm suprised it ran at all, honestly... That's not really, a real world situation though..Maybe with a dual cpu rig..

Some one with a 6800 Ultra and a P4 EE or 3800+ needs to try it.. I don't think my comp can pull the amps your looking for.

Perhaps there is a testing or benchmarking program out there that will utilize the cpu, gpu and disk systems to their full extent simultaniously, regardless of system specs. Anyone know of one..

Ice Czar, What are looking for to happen? If my psu wasn't up to snuff, it prolly would have just rebooted.
 
GORANKAR said:
.

Ice Czar, What are looking for to happen? If my psu wasn't up to snuff, it prolly would have just rebooted.
Click to expand...


Just trying to determine a formula for calculating a worse case senerio
a little cut and paste
----------------------------------------------------------------------------------------------------------------------------
kesv said:
inet said:
Do the math here: http://www.jscustompcs.com/power_supply/
There's a good chance you will need a bigger PSU than the 470..
Click to expand...
And once again. Let's not forget that these calculators are absolute worst case.
Click to expand...

well as far as calculators goes, that one sucks (sorry)
its very important to actually determine the amps per rail
and not all calculators are automatically the worse case senerio
I use > takaman's Power Supply Calaculator v2
which does have the option to set a utilization percentage, and to manually enter amp values for a given component
which is very handy if you have real world test values like
Power Consumption of Contemporary Graphics Accelerators: ATi
Power Consumption of Contemporary Graphics Accelerators: Nvidia
(all the better because they break out the draw between the AGP and the auxillary connector, which on an ATX12V v2.0 is on different +12V rails)
and various databases like Processor Electrical Specifications
(if the board has a +12V auxillary you add to the +12V otherwise it goes on the +5V rail,
and you convert the watts to amps for the appropriate rail, AMPS = Watts / Voltage)
or the actual spec sheets on the components (like say from Seagate)

that way Im able to run multiple calculations
first I do a "full" calculation, of all the additive maximums of the components
then for an ATX12V v2.0 supply or an EPS12V supply I repeat that for the +12V1, +12V2 and additionaly for EPS12V +12V3\+12V4 rails

then I run a "realworld" worse case senerio,
Its just as likely Im helping someone calculate a NAS as a lean gaming rig
so a "spinup" calculation is the first criteria
all the fans and HDDs\RAID arrays at 100% value and a 25% value for the CPU\GPU, often that will tell me if a more powerful supply is needed simply to boot or if the NAS\SAN need to employ a controller with a delayed spinup option
(common on SCSI controllers, and found on quite a few ATA)
once the fans and HDDs are spinning they will drop to 1\4 of their full rating, so I recalculate with that value as a static load, and throw in a 100% CPU\GPU and single optical (full spinup draw) for a worse case senerio realworld


Interpreting the veracity of a manufacturer is the tricky part
since we dont get to actually see the engineering test specs
which will typically say sometghing like
rated for full power at 25C decreasing linearly to no power at 70C
and your operating temperature being around 40C

generally I start with deducting a third off the rated amps for temperature
and then start to match to the baselines, factor in thermal considerations (PSU placement, type of thermal solution for the rig,) and special factors like overclocking or usage patterns (NAS\Server\Workstation\word processor :p ),

when I actually make recommendations however I typical build in a healthier safety value than the "real world" worse case senerio unless someone has specifically stated they are running a UPS
--------------------------------------------------------------------------------------------------------------------------


so with 125 viewrers this morning Im trolling for more information :p
 
and a bump, providing one or more of the 601 lurkers currently viewing this forum to contribute some more info ;)
 
Hope this helps
listening to Streaming music
strolling through the [H]
writing this
and runnin Real time [H]DR IBL at 1183x 645
CPU 100%
GPU temp 61c
VC ambient 35c
AS per Everrest:
MB 31c
CPU 42c
CPU core 1.74v
+3.3 = 3.31v
+5v = 5.05v
+12v = 11.98v
-12v = -12.03v
-5v = -5.05v
Digimeter: 5v = 5.09
................12v=12.13v Taken at molex ,should I tweek these volts at psu?
Hope this helps
 
thanx guys (& gals? never can be too careful :p)
looks pretty reasonable to assume that the CPU and GPU will hit 100% at the same time at least for some portion of a gaming cycle ;)
 
Another good way to pin your cpu @ 100% is to put a stoopidly high number into the calculator and then hit the factorial button, it will run @ 100% until its finished, but it won't max 2 cpu's on a dual system.
 
You must log in or register to reply here.
Back
Top