Optimization of games or what ?
On some games there is such a thing as for some unknown reason, dramatically raises the temperature of your card without increasing the load on the GPU and the memory. For example, in Fallout 4 was on one of the locations and after a few minutes the temperature jumped from 60* to 82*, but when I ran from this place about 50 meters, the temperature recovered. The same was observed on the maps of the mod in Far Cry 4, and other games.Vidyuhi GTX 780 Ti. The typical load on the GPU=50-60%, memory=50=80%, temperature 50-70*.
Guess that's due to poor optimization of the games in some locations, but it is strange that only increases the temperature, without increasing the load on the GPU and the memory. Who is met and what can I say ?
This had happened only during the increasing load had a fever. she can not so simple in itself to jump for no reason. optimization here has no influence. Look very video card - thermal pads, thermal grease, can cooler wedge or something.
Slipnot6
No, none all right. I wrote that when run away from this place, the temperature is restored.
vftor
Optimization is NOT NECESSARY, at least Far Cry 4, played 123 hours and never saw such jokes, the temperature at the same level. I hope that the vertical sync is not disabled?
And Yes, the best way is to test the graphics card in the program notes, and follow the indicators.
vftor wrote:
The same was observed on the maps of the mod in Far Cry 4 and other games
And that, in other games the same bad optimization?
You need to look the hardware. To drive, for example in FurMark. I have my loaded to the max, 100% load and a temperature in the range of 70-73.
The typical load on the GPU=50-60% - so of course, loaded under 100%, and got a high temperature.
How long have the thermal paste changed? And temperature increases with increasing load on the graphics card, all natural. 50-60 was somewhere in an open area without objects, herbs or something, and was 99% in areas with lots of objects to process, along with the increased temperature.
VOVAN WOLF
Settings in games and Vert.sync. everything is fine. Complete prevention vidyuhi with the replacement of thermal compound was done half a month ago.
Important:
1. Why when I run away from the place in the location 50 meters, when jumping temperature, it is restored ?
2. Why the jumping temperatures do not increase the load on the GPU and the memory ?
this is certainly not normal, but it's hardly optimusone. Is to go through the iron of his bunch
Dead Note
ahaha not drive)) reveal the secret for you dear.
vftor
in afteburnere msi or nvidia inspector/ there is such a thing as the priority temperature. that is, put 81% of all video card does not rise above this temperature.. there's just a little cut back frequency if that..
vftor
Oh, ie, even without load. Then wait for the test results in FurMark, to drive and to look at frequency and temperature chart.
The first idea of the condition of the temperature sensor. It does not grow instantly, it grows gradually, and the higher, the slower - accumulating capacity of iron. Well, the fans can look.
do not increase the load on the GPU and video memory - load should grow with disabling vertical sync, then squeezed ALL the juice out of the system unit, then the load should be at 99%. 100% appears at maximum settings FurMark-E.
Check the numbers there.
I do not recall a single game where the load on the graphics card was so low 50-60%. What is it? On what resolution you play? Video card is outdated, and gives 50-60%. What FPS shows in the game. 60 frames is the vertical sync pulse, each first frame. Gets it to 60 fps?
Parallel to the questions above can be answered. If the load of 50-60%, where the temperature under 82 degrees. Need something to knit.
P. S. in fact, according to the specifications of the gtx 780 TI is the Maximum temperature - 95 degrees. We can assume that 82 degrees is the norm. And make STUPID, get MSI Afterburner and set up a schedule of fan operation if the temperature is annoying.
In General, the test for 10-15 minutes at maximum load and draw conclusions.
Fei Fong
I have EVGA Precision X, I put an upper limit temp.=81* and sound signal at 82*. Is still a gadget, GPU Monitor, which constantly monitor the temperature and load on the GPU and video memory (on the secondary monitor). So on this gadget, when the tramp of temperature on one place in locations I have not seen increasing the load on the GPU and memory.
vftor
) guy, gadgets not always show accurate info. To refer to them is also that on Wikipedia. You wrote above, put your iron stress test
VOVAN WOLF
Actually GTX 780 Ti is a very productive card. Some of its settings higher than the GTX 980 Ti. For Example, The Peak.fill rate texture=210, and GTX 980 Ti =176. And play monitor 1920x1080 (main) + 1280x1024 (control). Games can be adjusted (on high) to the graphics card were loaded on 50-60% when FPS=60. With optimized games it works fine, but with poorly optimized, do lock the FPS to 48 Hz, or even 30.
vftor
Some of its settings higher than the GTX 980 Ti - that's only 980 TI gives more frames by about 30%.
Compare 780 TI vs 980 TIhttp://www.anandtech.com/bench/product/1717?vs=1715
So, to find fault with the figures, you can and another. But, the 780 TI still drags, don't say anything, accepted, thought worse.
Game set up so (high) that the graphics card was loaded by 50-60% when FPS=60 - what is this? I see that the maximum of the gtx 780 TI Far Cry 4 gives 75 frames and temperature at 76 degrees. It seems it is the norm. But, 82, and even with low settings, that's a bit much.
Spoiler
and with poorly optimized - Far Cry 3, 4 and Primal do not belong to them, they always boasted stable and high FPS, not once did I complain. Yes, and posts do not remember.
And Fallout 4 duck even easier and simpler headlights edge 4, it and FPS is higher, well at least in the wasteland (although the game is difficult to answer, it is a terrible place, but there are more issues with upload and download speeds overall, such issues exist).
In this game on gtx 780 TI max gives 74 degrees.
Fallout 4 PC / 1920x1080
In this case, all that we look. this is only excerpts, there may be sometimes the same ship is not childish.
Gtx 780 TI is reference and (and suddenly so worth it) - he temperature is ALWAYS higher.
And Yes, all the same, what are the tricks, and even with a load of 50-60%, the question arises. Actually, it is floating, and thus to adjust so precisely is generally not possible, at least, it's hard to imagine, still there are places where load will grow, it all depends on location, environment and facilities. The location is ALWAYS different, and different load.
Fps does not vary too? The more fps the more temperature. Generally a combination cooler-graphics card need to not warmed to above 60. Well, or 65 if the card is 250 watts as you, but there is only a hybrid air / water or pure water will save. If you don't want the blade chip.
VOVAN WOLF
Right, 980Ti above 30%, but for the latest poorly-optimized games, starting with Fallout 4 for what she has done with their drivers + increased memory.
A game I set up so as not to ruin the none excess load. The quality of the graphics on high I was quite happy.
kc5
Dear, it is incorrect mark, if I may, I would like to note and correct myself:
From the articles:
- A real dump of the chip from the Board happens very rarely, although happens. Mainly as a result of mechanical impact: impact fees, curvatures, deformations, such as improper installation of massive cooling system or misalignment when installing the motherboard, and can also be the result of defective lead-free soldering, in revidirana temperature conditions, etc.
- First the blade chip is due to the mechanical stress at the soldering and the most noticeable is the influence of alternating stresses. Where they come from — vibration (including the cooling system) or cyclic heating / cooling (thermal Cycling).
Always said, better to let the equipment operates in the nominal mode, last longer.
Ie in real life, not so much the temperature itself, how many DROPS (on / off), mechanical damage, improper installation, vibration.
The video cards have operating temperature range that defines normal operation of the equipment, in this case, there is a limit to 95 degrees. To the limit 95-82 = 13 degrees. The temperature at 82 - this is not a permanent phenomenon, but only temporarily. Usually, as noted by the author, 50-70 degrees.
So, for the new architecture Maxwell (graphics card 900 series) and pascal (1000 series) achieved lower temperatures by reducing energy consumption with more efficiency. There are operating temperatures of 60-70 degrees, are a great indicator. It is important to note the difference between the reference (turbine) and preferences (fans). The reference is ALWAYS the temperature was higher and the frequency of work - below. She could warm up to 85 degrees, where the temperature is considered quite the norm.
Old architecture, with higher numbers of energy consumption, were given higher temperature. The temperature in the range of 70-80 was quite the norm.
From all this we can conclude that the temperature readings are quite normal, if you omit nitpicking. And the maximum temperature is not the only reason the blade chip, metal expands when heated, and contract when cooled, not terrible the temperature itself, as the process itself.
Well, I understand. If wrong, then corrected.
vftor
A game I set up so as not to ruin the none excess load.
I asked the same question, but from a different piece of equipment - HDD. Constant disk access, torrents, downloads/distribution, all this leads to wear and reduced service. It's a dead end, and no you can't use, because WHY BUY and ship so, how long would he live?
This is a sick question, but with the same success can on the shelf and forget to put there then just nothing will happen. Things have to work out the money, and exploitation is not the only cause that can lead to destruction of the device.
Here are a few options: 1) due to power Fluctuations, BP missed and will burn the WEAK link, why not the GPU? Besides, BP and myself fly to the light, you can take something. 2) Lightning discharges. recently was the subject of one friend hooked Mat. fee, and she began to die, he went for a new system unit, the second, he said, lost PC. 3) it Seems to be a good thing, as he can take the gtx 1080, but the EVGA was a bad model, which was burning scheme. 4) Any other accident. Including swap/sale.
You can think of many, so I think to protect something as silly from what? She can live a long time and can cover at one point and that? It's a lottery as lucky.
But it's not our business how things are there. I thought so, you didn't have to ask. To each his own.
Personally, I prefer to enjoy the beauty to the maximum with high FPS 50+.
And Yes, we've deviated from the topic, managed to get rid of the program?
VOVAN WOLF
I believe such arguments are erroneous. It is necessary to protect your health and your technique. The load should be back to normal from 20 to 80%. It is a pledge of reliability and durability. But both man and equipment to withstand momentary loads of up to 100%.
In relation to modern PC hardware, considering the price, it is better to limit the load to 60%.
VOVAN WOLF
In General, I decided to watch the behavior of the graphics card and the temperature in games. It is possible that this school gadget GPU Monitor.
VOVAN WOLF
Now let's buy expensive (by your income) graphics card and I would venture to check the normality of 80 degrees with constant 60-80 drops in the game is because of alt-tab. The new money only after 5 years at least. I have here a card for$ 200 was dead for 3 years. The second one for 240$ from a small bend in 5 years. The viscous solder all abandoned in favor of the fragile and over time it becomes stronger but quite the contrary. rtfm 1)Occam's Razor 2)Planned obsolescence
Think 7 times 1 times to write and not Vice versa.
The typical load on the GPU=50-60%
And the processor what? and what his normal load?