If you need a replacement cooling?
HIS 7950:
What say the experts?
There probably is more likely to irritate noise 1 cooler than temperature, 74-75 working, the VRM standards, too, especially for stoves of the type 7950-7970
Depcool V6000 help, or Dracula, there is still a Gelid icy is hard to find.
bloomm
Well,Yes,at 4,000 rpm the plane
But this theme I like:
http://www.deepcool-ru.com/Product/V6000/
[Denis]
But this theme I like:
www.deepcool-ru.com/Product/V6000/
and what? another Chinese junk disguised as neborachko. Such disguises heap - just search to correctly use necessary. Although , when you consider that the market is anything from solutions to try to improve the heatsink VRM's (to alter it, I mean make a new one ) - to the second valve of the fan blowing on it efficiently.. rebooted WITH should not be.
WanRoi
The most simple: fasten a hull cooler to the radiator of the graphics card, 80mm, or 120mm - which is handy. The price for them in the area of 300 rubles. for 1300-2000 rpm. Really need to take into account the condition of the radiator!
LOL )))))
bloomm
.... VRM standards, too, especially for stoves of the type 7950-7970
85 degrees on the transistors? even know what 85 degrees? It's not Fahrenheit.
WanRoi
P. S If not a secret: what jaccaci makes under the video card?
Really good question... at Least someone noticed.
p.s. Generally cool on the PG always, do not care that he has slipped relative to the zero. But you can always see a bunch of lulz from the same bunch of users - and sometimes neighing under the table ))))
p.p.s
Damn, people something generally become arrogant. You buy their BU cards for the sake of experiment, and they with them like a pig treated. Not too noble is for some thing to buy maps of BU, even at pretty hot prices.
[Denis]
to kill a card I think will not work if it is not under significant acceleration especially
dangerous because the temperature is always 100-110 (and faces for the most part the loosening of the PCB and otvalivayas from soldering and high temperature fluctuations during operation) but then the map itself most likely would drop the frequency if there is turboboost
well, there were listed all sorts of tips to remove the cover and configure the ISI afterburner the dependence of the cooler 40C=50% 60C=100%
although my pitch to the graphics card turbine or around the turbine of a type not prone to large sensitivity to ambient temperature
but if the gasket between the cover and the memory has decayed or not at all then again the temperature is little affected by
WanRoi
MunchkiN 616
bloomm
In General,removed ohladu with 7950,Tempesta dried,but I was not that confused:
Between the radiator and the complete lack of memory of the conductor.4890 on some filipucci was:
Maybe someone put something similar?Or knows someone that can put?
[Denis]
the lack of gaskets commonplace for bugten-budget segment as I Peck suspected
or what to put not sure because the map is designed for this kind of work and intervention may lead to Pitch with unpredictable consequences
MunchkiN 616
If the heat is easier to be allocated,what consequences can there be?Maybe the gaskets are removed,so memory is not persecuted?That worked on a temperature maximum
HD4890 funny version. She set Spitfire G - way must be suitable (a large and a small radiator). But it is not for sale . But for the HD7950 should be designed for VRM'Noi region. Gemorrah something is a bit....
[Denis]
If the heat is easier to be allocated,what consequences can there be?Maybe the gaskets are removed,so memory is not persecuted?That worked on a temperature maximum
Forget about laying them now shaft in any region. You on the other need to worry )). The memory drive is meaningless, and at the drain it is not strongly heated. It is DDR5 not DDR3. The memory of nothing.
Resoldering polypropylene condensatori to the other side. so as not to interfere designed the heatsink for the VRM to deliver? But they are so hard to use.... + it is necessary to observe the polarity.
Easier to sell before the expiration date. Anyway )). 5K or 6K Mona to help out.
[Denis]
As to do with it? It is the memory at 10-13° heated more
Bollocks! ))) Where did you see it? Show on the screen in red?
Sergey067
Bollocks! ))) Where did you see it? Show on the screen in red
You're either a Troll or blind.In the first fat and the second go down to the optometrist
WanRoi
Yes he is no.But cover:
WanRoi
The idea is (second photo), metal base should tightly cover the memory chips. Perhaps in some test environment, they have calculated that ahlad good, the need of gaskets is eliminated.
Damn, I thought the radiator is rectangular, and it is small for such a length of the graphics card.
I repeat not in this case. Well she GDDR5 with heavy load prone to slight heat. But it is small. To bask in the memory area can only be because a number of the graphics chip. Heat is transferred from it around the area - and the memory also receives additional heating. But this is solved very simply - buy the thermal pads, cut to the desired size, blur, och thin thermal compound , that has not flown and placed in memory.
But this does not solve the problem. If you need a quality change WITH the heatsink on the power transistors.
+ Optionally, you can associate the conductivity of the memory and heatsink WITH stock memory will get extra nagrev, as the temperature of the chip at the GPU-Z over 70 deg. Good or not Good - good question. Now everything is clear - why the thermal pads on the memory put?
[Denis]
You're either a Troll or blind.In the first fat and the second go down to the optometrist
well what are you gonna do... Dammit, they'll never learn, Bloede.Dh'oineHocuspokusabrakadabra, bollocksandarses....+ yourbunnywrote
Where GPU-Z to monitor the temperature of the memory? In the first screenshot it is not visible.
PS And this person with such knowledge wanted to switch WITH, and even dismantled the map.... well, well
P. P. S. [Denis] , sell. Then the elbows will is, that to listen to )).
[Denis]
it is sealed not with the aim that is persecuted and to optimize the substrate owing cheap ahlada since this preference
https://overclockers.ru/lab/show/49792/Testiruem_raznye_varianty_Radeon_HD_7950_Sapphire_HD_7950_Dual-X_OC_XFX_HD_7950_Core_Edition
if you look at the 7950 in the review the xfx strip and sapphire has a simple heatsink on all the memory and picolinic
the khis fat be likely to mold them nowhere because the sole is small. from the turbine of the reference, in turn, everything seems to have been
MunchkiN 616
The sole is not much small.The turbine is a little harder due to the presence of tubes.Today I disassembled both and compared
Sergey067
And che to tell? 3-5 minutes.And VRM I differently try to solve it.I admit, thought it was the memory overheating,and transistors.To sell still will not
[Denis]
And VRM I differently try to solve it.I admit, thought it was the memory overheating,and transistors.To sell still will not
Oh, so enlighten us how would you solve this problem with VRM????
The most interesting part about the VRM , what is it I him 100500 explained what it is. But he listened to it MunchkiN 616 at a tip. ))) what interesting logic.
Don't want to sell you is worse. WITH better than this monitored by PWM chopper you can't do.
Sergey067 wrote:
85 degrees on the transistors? even know what 85 degrees? It's not Fahrenheit.
you th bombing, the working rate it =) gtx 970 under control of the condom bukletom under 97 rode =)
[Denis] wrote:
.I admit, thought it was the memory overheating
It's not overheating, forget it =) If you take Off 10 degrees, hail, and so it's operating temperature would not be out there sensors and did not even know would be there at 90 =) Put some kind of deep cool Dracula is cool and chip and VRM, downvolt also an interesting pastime, as an option.
People protocolsa to the new cold maps, and now they think everyone should work not above 60 degrees and that would Pytalovo guys are not warming up and the memory is cold as ice it was, yeah of course =)
bloomm
In General,I've found a way of reducing the temperature of the VRM,it's after 10 runs of Metro LL bench a) :
Removed the cover from the card,took the side panel of the system unit and down blowing on the valve 800об 120mm/min aerosol.The air flow directed on VRM1.Re-built the bike,of course,but it works.Thanks for the help:
