Windows 10: Show Us Your Temps
Here's mine. Maybe a little too much information but may be helpful to someone in one way or another.
Temps shown at idle, then 100% load, then CPU-Z showing the 3.20GHz CPU is really running at a 4.20GHz overclock.
Some have achieved considerably higher with this Pentium G3258 CPU but mine starts acting up after about 4.20GHz.
It will run at 4.40GHz but not reliably.
Anyway, here are the temps:
CPU 100% load
CPU specs. Notice the Core Voltage & Core Speed
Temps were much higher than this with stock cooler and couldn't handle any overclock at all. Still ran around 55C at idle with no overclock.
I now have installed a Cooler Master D92 cooler and use Tuniq TX-2 thermal compound. Temps are still not what you could call chilly but they are perfectly acceptable and will withstand a pretty decent overclock.
Funny note about thermal compound: You can Google for advice about thermal compound and you'll see more than you ever care to see information posted from major magazine and website (Tom's Hardware, CNET, etc.) 'experts' and can read about it for days and days and still not know what to use.
I've had 4 system over about the past 3 years with various CPUs, thermal compound, and coolers, both air and water.
People get the idea that their temps will drop at least 20 degrees with water or some other outrageous expectation. It won't. Save yourself the hassle. There are air coolers out there that can cool every ounce as cool as a moderate level water cooler.
Anyway, about thermal compound. Throughout all my tests I found a pretty shocking bit of information. The el cheapo basic white $2 a tube thermal compound from Radio Shack worked as good or better than any $40 a tube complex metal particulate compound out there.
It's true. Before you call me crazy, just try it. You won't believe it.
ill have to disagree with you about temps
because when I'm running say gta v or unigine heaven 4.0 triple monitors
my gpu would be as high as mid 80c on the air cooler
now ive got it watercooled via 1x 240mm rad and 1x 120mm rad
it now sits at around low 50c so that's about 30c difference
and when I put the water chiller on it hovers around the mid to high 30c
and that's with the cpu on the loop too
nothing is overclocked though its all stock
I think that's an exceptional result you have there. I never experienced anything near what you described. Look up just about any water cooler tests done by any major magazine or website and you'll see that they consider going to water from air and saving 8 or 10 degrees C is quite an accomplishment.
Here are the results of a review by Tom's Hardware of 7 closed loop water coolers trying to beat the cooling ability of the famed (air) Noctua NH-D14.
(Noctua's results at the bottom)
You'll see that the air cooler beat every water cooler at idle temps. Then was only 8 degrees C warmer than the very best water cooler at 100% load.
Yes I can see what u mean
I had a closed water loop the 1st time I went the watercooled route
Because it was cheaper and less hassle.
I also was abit worried of leaks etc
But I found that just like u said the temps wasn't that great compared to after marked air coolers
So you are right in that respect.
If you go the whole hog and have a full custom loop
The temp benifits are great, in my experience
Plus you just get a constant low hum of the fans and not the sounds of a hair dryer from the gpu fans when u start playing games or benchmarking.
If u go over to the valley benchmark thread I posted my benchmark results but also included my gpu temps in the screen shot to illustrate how low my temps were
1 gpu was on 37c the other was at 39c
I'd be surprised if anyone on air was below mid 60c to low 70c
I re-read my post and I realize that it sounds like the tone of it is "I don't believe you'. I certainly didn't mean to sound like that.
Another thing, too, is that when I read your details I see that I may have been comparing apples to oranges. The Tom's Hardware water cooler review was for closed loop systems (just strap her on and you're done) and yours is a totally different full custom loop.
The two or three water systems that I personally used before were also closed loop. First, the older Corsair H50, then later some crazy off-brand system that had a 45cm tall aluminum cylinder with fins that held almost 5 liters of water with no fans. The pump just circulated the water across the CPU block and through the tall cylinder. Lastly, I tried the Corsair H105.
All of them gave a temp result of very slightly above, below, or equal to an air cooler that I don't think has been matched yet.
It was copper, and about as big as a softball (unheard-of huge fans at that time). I believe it was a Zalman CNPS9000 or something like that. Copper base, fins, and heat pipes and a huge fan inside.
Sadly, when I got my newest motherboard, an EVGA Z87 with Socket 1150, the mount wouldn't work anymore so I had to replace it. Searched everywhere for an aftermarket mount and couldn't find one. I researched until I got tired of researching and finally got the Cooler Master D92, which is also a huge heatsink with 2 big fans mounted in push/pull. This cooler is so big the side of the case will barely close (maybe 5mm clearance and you can forget using a side fan, if you have one.) and it's a big case, Antec 900.
That's another thing, you have to read reviews until your eyeballs fall out trying to decide what thermal paste/compound/grease/TIM (all the same thing) you want. Temps with different air coolers can vary wildly. Some may cool to 65C under full load and some may cool to 40C under full load. As I've mentioned, it's my experience that thermal compound may make make a degree or two difference but nothing to write home about. I've used the very cheap and the very expensive and different application methods (my favorite now, after much experimentation is putting kind of a pea sized ball right in the center of the CPU heat spreader and press the cooler on top and bolt it down).
Anyway, I'm glad you have a handle on the most damaging of all elements in a computer, heat. An engineer said in a recent article about overclocking that it's not excessive voltage that kills CPUs and GPUs, it's the heat generated by said overvoltage.
Have a nice week, my friend,
wow you have tried a few different things
I agree its is all about what we both have personally experienced
plus how much effort we are willing to put in in research and installation
it is a lot of effort
I'm the same as u I havnt seen much difference with the difference thermal paste brands
no I didn't think u came across as you didn't belive me
its just that its easy for people on the net to claim things with nothing to back it up with
I just wanted to let you know that I wasn't just making it up
nice chatting to you
About trying different things, crispy, some of them were against my will. lol
I've been a computer technician for 25+ years and an average of about 3 or 4 systems a day come through my small shop and I've had to resort to trying different things out of desperation once in a while.
I'm a firm believer in the saying that 'nobody knows everything' and I get surprised more often than I care to regarding these machines that I believe were set loose on us by an evil government, designed to drive us crazy. It works, too.
I'm much crazier now than when I worked on my first computer. lol
I just upgraded my Dell Optiplex 790 from an i5-2400 to an i7-2600. Adding new thermal grease and cleaning all of the dust bunnies out of the heat-sink has really lowered temps. This is just the stock cpu fan/heat-sink that comes with the Optiplex 790. Just using generic white thermal grease, none of the fancy silver stuff.
Here are my temps.
CPU is OC'd to 4.0 GHz.
I'm building a PC and I know that many of you here have custom built PC's.
So, I though that maybe it would be nice to have a thread where you could show off your work.
Just a thought,
Here is my community centre PC's Desktop.