Gtx 480 VS. 8800 Gtx , My upgrade and benchmarks

[dreamphantom_1977]

More Pics (Link below)

http://pbckt.com/s3.wnJ

Today I just got my new Galaxy Gtx 480 and replaced it with my old 8800gtx. I thought I would post pics and I did some benchmarks just in case any of you with an 8800gtx wants to see what upgrading to a Gtx 480 will do.

I am upgrading the Gtx 480 to a water cooling kit, but unfortunately I had some trouble with parts getting damaged and am waiting for my new parts to arrive, also, I just got lucky enough to order a Danger Den GTX 480 Block with nickel top and copper base. So when I get that, and All my parts, Ill repost everything with the new pics. Hope you enjoy.

Here are the benchmarks:

Thermaltake ArmorPlus(Armor+) VH6000BWS
Motherboard Name Gigabyte GA-EX58-UD3R-SLI
CPU Type QuadCore Intel Core i7 920, 3200 MHz (20 x 160)
System Memory 3072 MB 3 x 1 GB DDR3-1333 (Corsair XMS3 CM3X1G1333C9)
Audio Adapter Creative SB X-Fi XtremeMusic
Hitatchi 1.0 terabyte deskstar 7200rpm 32mb buffer
Windows 7 ultimate 64


Graphics cards- 8800 Gtx vs. Gtx 480

All test ran on 8800 Gtx, then card pulled out and switched to Gtx 480 and ran again.

Crysis benchmark

Settings:

1920 x 1200
Anti Aliasing= 8X
DirectX 10
64bit
quality settings- all set to VeryHigh
loops= 1
time of day= 9


8800 Gtx:
Min FPS= 2.65
Max FPS= 8.20
Ave FPS= 6.89

Gtx 480:
Min FPS= 14.67
Max FPS= 30.17
Ave FPS= 25.32



Far Cry 2

Settings:

Ranch Long
loops= 1
1920 x 1200
Anti Aliasing= 8X
DirectX 10
Quality Settings- everything set to very high
accept Ambient and Post FX are set to high
Bloom & HDR are checked

8800 Gtx:
Min FPS= 16.26
Max FPS= 57.07
Ave FPS= 28.38

Gtx 480:
Min FPS= 48.30
Max FPS= 157.05
Ave FPS= 86.94




Heaven Benchmark 2.0

Settings:

DirectX 10
Shaders= high
Anistropy= 16
Anti-aliasing= 4X
Fullscreen
Resolution= 1920 x 1200

8800 Gtx:
Min FPS= 8.6
Max FPS= 36.1
Ave FPS= 18.7

Gtx 480:
Min FPS= 26.1
Max FPS= 91.9
Ave FPS= 47.5




Grand Theft Auto IV

Thought you might like this test as gpu in this situation makes a huge difference.

Settings:

Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: Medium
Shadow Quality: High
Reflection Resolution: Very High
Water Quality: Very High
Texture Filter Quality: Anisotropic x16
Night Shadows: High
View Distance: 100
Detail Distance: 100

other settings:
Vehicle Density= 100
definition= on

8800 Gtx:
AVE FPS= 30.55

Gtx 480
Ave FPS= 64.79






OTHER:

GTAIV GTX 480 Max Settings

Statistics
Average FPS: 63.55
Duration: 37.34 sec
CPU Usage: 37%
System memory usage: 87%
Video memory usage: 85%

Graphics Settings
Video Mode: 1920 x 1200 (60 Hz)
Texture Quality: High
Shadow Quality: Very High
Reflection Resolution: Very High
Water Quality: Very High
Texture Filter Quality: Anisotropic x16
Night Shadows: Very High
View Distance: 100
Detail Distance: 100

other settings:
Vehicle Density= 100
definition= on



Temp Comparison
(i did not do an exact benchmark, these are just the min and max I observed)

8800 GTX: (with aftermarket cooler)
Min Temp= 42C
Max Temp= 56C

GTX 480:
Min Temp=49C
Max Temp=90C
UPDATE: Max temp has hit 91C so far, I will update if I see higher temps.
UPDATE: Min temp so far is 46C
On a side note: My case is a huge case, and is a little loud, but,
going from the 8800 Gtx to the Gtx 480, I have not noticed it being any louder at all. I was expecting a huge difference, but I don't think anyone would notice a sound difference, even at full load. Maybe it's the placement of my case, I dunno.
I actually think the card is very quiet in the thermalright armo+ case.

UPDATE: I have noticed in some games when I'm playing for a bit that the fan does get loud, it's not extremely loud, but it is noticeable. It's like it's warning me, "hey, im getting hot, hurry up and water cool me"..

 

[L1qu1d]

Dont be I'm running a Micro ATX its going to be like to Sumo wrestlers in my case lol

 

[EXT64]

Oooh, very shiny

 

[OvrClkr]

That block looks killer, cant wait to see the final outcome.

 

[notty22]

Do the people that sold you that awesome looking water block give you a game plan on converting the card over to water. What I mean is, cooling for your ram and key VR components. Thermal pads to the block ?

 

[dreamphantom_1977]

Do the people that sold you that awesome looking water block give you a game plan on converting the card over to water. What I mean is, cooling for your ram and key VR components. Thermal pads to the block ?

If you go on Danger Dens site, they have plenty of pictures, and video's explaining everything, from removal, to installation, temps, and overclocking.

http://www.dangerden.com/Page-2.html

Disassembly
http://www.youtube.com/watch?v=SQPD4k_kB50&feature=player_embedded

Installation
http://www.youtube.com/watch?v=L_0w2_dXt5A&feature=player_embedded

 

[dreamphantom_1977]

Well, Yesterday was my day off, and I got my fans in, so I decided to start my build yesterday. I worked on it from 7pm to 3am, and I am just about finished. I am at work right now, so I haven't plugged it in to test it yet. Last night and this morning I was leak testing it though, and so far everything looks good. I took a bunch of pics. I am not finished yet, I ordered 4 fans, but could only put two on, because the screws that came with the radiators were too long as they are meant to be used with radiator grills, and my fans didn't come with any, luckily I had two lying around so at least I have two fans. Gonna try getting two grill tonight and get the other fans on. Either way, I'll have it up and going tonight, and i'll post some temps and some more pics, possibly overclock. Anyways, hope ya like the pics.

On a side note: The stock cooler for the gtx 480 was just about impossible to get off. They used threadlock on the small screws that held the cooler on. And I had to apply a lot of pressure to get the screws to loosen up. Three of the screws almost stripped out, and I was using the proper screwdriver size. In order to get it out, I had to have my wife hold the card, while I clamped the screwdriver with a pair of pliers, pushed down on the screwdriver and turned the pliars. Finally it came loose. I thought I would have to drill them out. Hopefully the card didn't crack or anything . I guess I'll find out tonight.

All pics
http://s177.photobucket.com/albums/w237/Dreamphantom_1977/

U.V. Pics, I put my black light by the case, no pics with fluid in it yet i'll take some tonight.

 

[dreamphantom_1977]

I wanted to post 11 pics but the forums only allow 10 per post so here is the eleventh.

When I bought the case I wanted to get it the inside powder coated metal flake blue or emerald green, but unfortunately I didn't have enough to have it sent out to have done, and there is no place I could find around me that does powder coating. So, hopefully that will be a future project.

 

[cromedome]

dude this sh!!Ts insane

 

[rofl_my_waffle]

Aww your running a dyed liquid. Straight water is the way to go.

Are the anti kink coils really necessary? I mean under what circumstances would they ever kink.

 

[dreamphantom_1977]

Aww your running a dyed liquid. Straight water is the way to go.

Are the anti kink coils really necessary? I mean under what circumstances would they ever kink.

Well, I am running HydrX because one of the guys from toms hardware team suggested using it, and he said he ran it in his system for 2 straight years with no problems with sludge buildup, algae or corrosion. I'm really not worried about the dye in it, Although I would have rather had blue or clear to match my tubing. Either way, If I want to change it later on it's only a drain away, no big deal.


I actually put the anti-kink coils on hoping they would help with the kinks. When I started running the loop I originally wanted, the hose kinked, and the loop didn't do anything to stop it from kinking, so I ended up changing my loop configuration. I'm thinking, I would need metal loops of internal loops to stop it from kinking. Either way, I like how it looks with the anti-kink, I love the blue. To each his own I guess.

Thanks for your posts.

Can't wait to get home and test it out.


EDIT: oh, just so you know, the bottom pics in the above post are empty loops. The hose is uv blue, and when I shine my 3 foot U.V. light on them it turns almost solid blue color. I did check out how it looked last night with the U.V. green in them, they actually turn almost a solid white color, while the coolant in the reservoir turns green. It looks really cool in person actually. Hopefully they will come out with a clear or UV blue HydrX in the near future.

 

[L1qu1d]

Wait water? Doesn't reg water have a high evaporation rate and less heat is needed to change the temp of water compared to coolant? Its like putting water in an engine than water coolant. It will heat up if temps go over 100 Degrees, even with the fans.

I forgot the scientific word for high heat capabilities when talking about materials. (heat resistance and how much it takes to change the temp of a certain object...damn grade 11 Science...yrs back)

I think u should most likely use water to clean out the system and use either 50/50 coolant (Although I think most users use around 10-20% coolant) or depending on the pump power more or less coolant.

I might be mistaken though!

 

[notty22]

Sometimes hose will want to flatten after 'time' especially at the apex of a curve. I've seen that happen. I think it looks cool. I only recenlty realized people were making the water glow, its novel + cool.

 

[dreamphantom_1977]

Wait water? Doesn't reg water have a high evaporation rate and less heat is needed to change the temp of water compared to coolant? Its like putting water in an engine than water coolant. It will heat up if temps go over 100 Degrees, even with the fans.

I forgot the scientific word for high heat capabilities when talking about materials. (heat resistance and how much it takes to change the temp of a certain object...damn grade 11 Science...yrs back)

I think u should most likely use water to clean out the system and use either 50/50 coolant (Although I think most users use around 10-20% coolant) or depending on the pump power more or less coolant.

I might be mistaken though!

Oh, i'm sorry, I meant I am running HydrX plus distilled water. I'm actually running less HydrX than It specifies. On the bottle it says add 1 bottle to .5 litters. I have added 1 1/4 bottle to 2 liters (aprox. 1/2 gallon) of distilled water (should be using 4 bottles according to the bottle). I am planning on adding the rest of the second bottle when I get home (i was waiting for leak testing first), but either way, it's mostly distilled water.

 

[soundefx]

Mad props man.

 

[EXT64]

Yeah, can't wait. Hope it all works well.

 

[dreamphantom_1977]

Well, I after work I stopped at my best friends house, and he gave me two fan grills so I could mount the fans. Took me awhile but I wired up all the fans, and leak tested it for another two hours. Anyways, I finally got the system up and going, and everything is going great.
I didn't take many pics, only a couple, but I did manage to run some temp tests. I ran msi afterburner and kombuster on extreme burn in, anyways here are my temps.

Kombuster extreme burn in @ 1600 x 1200 for 1/2 hour
Max temps = 55C GPU and 39C for PCB

To get the minimum I wrote the temps down when the computer was up and running idle for a few minutes, and then after combuster I let the computer sit and seen how long it took to drop back down to those temps. I didn't time it, but it only took a couple of minutes.

Min temps = 32C GPU and 28C for the PCB

Anyways, i'll probably do some more testing tomorrow. Here are a few pictures. Hope ya like them.


Kombuster screenshot taken @ 1/2 hour mark

 

[microterf]

Very nice. Impressive

 

[L1qu1d]

GOOD

 

[EXT64]

Wow! That is awesome.

 

[rofl_my_waffle]

That is very impressive. What kind of clocks can you get on that?

Im using a single 360 radiator and my 5970 + i7 + north bridge. The 5970 barely breaks 30 and the i7 averages at 45. Suprised that using two 480 radiators you got over 50 with the two cards.

Pure distilled or deionized would be a better coolant because glycols have higher heat capacity but conducts heat a lot slower. Since a liquid cooling loop doesn't get anywhere near the boiling point or freezing point of water, you don't need it. Also glycols have slow down water flow. Anyway it is insignificant.

I worked with lots of coolants and hydraX was one of them. They are all the same product. Some form of glycol + dye. The dye can build up in parts of the loop if you have a highly restrictive block. But those GPU blocks look very smooth and should be fine. You aren't adding too much either. I just don't like the dyes because they stain everything like my hands and anything it gets on while providing pretty much no performance increase.

 

[notty22]

His system is performing right in line with Danger Den's informational video. Made for the GTX 480 and their block.
Installation
http://www.youtube.com/watch?v=L_0w2_dXt5A&feature=player_embedded[/quotemsg]

 

[cromedome]

DreamPhantom... Just by looking at the cooling and stuff. Did it not cost as much as a small car? lol... Should you not have just bought another 480 to get way more performance than a single 480 + liquid cooling???

 

[OvrClkr]

Why dual 480's? A single 480 watercooled should be more than enough, especially on water where he can up the voltage pretty high w/o worrying about temps..

 

[dreamphantom_1977]

DreamPhantom... Just by looking at the cooling and stuff. Did it not cost as much as a small car? lol... Should you not have just bought another 480 to get way more performance than a single 480 + liquid cooling???

Lol, no I didn't buy a second radiator, I actually got the second one for free. What happened was one of the radiators got severely dented on both ends during shipping. I ordered it from performance-Pcs.com, and they didn't wrap it well. I sent them pictures, and was gonna send it back with an rma, I wasn't gonna use it cuz I was worried about leaks, but they told me I could keep it, and sent me another one. Which, by the way, also came with a couple of small dents. Neither were packaged well.Speaking of performance-Pcs.com, I have used them before, and never had trouble, and they have great stuff, so I highly recommend them, just tell them to wrap your radiator good when you order, they have a note section in the order form.

Anyways, I decided that if they didn't leak then I would use them, so I leak tested them both, and neither leaked, so I ended up using both. I didn't do it just for cooling a single card, I knew I wouldn't get better performance out using an extra one with a single card. Anyways, I decided to add two radiators because I want to upgrade down the road and figured I'd get the work out of the way so i'd be set when I upgrade. I decided I want to upgrade to a cpu water cooler, and add another gtx 480 when I can afford it.

Besides, I get the extra benefit of being able to run the fans on a lower speed if I want so it will be quieter, and I can crank down the pump so it will last longer and still get decent temps. (I can also add more fans to the top radiator if I decide to)

Plus I think it looks cool.

That is very impressive. What kind of clocks can you get on that?

Im using a single 360 radiator and my 5970 + i7 + north bridge. The 5970 barely breaks 30 and the i7 averages at 45. Suprised that using two 480 radiators you got over 50 with the two cards.

Pure distilled or deionized would be a better coolant because glycols have higher heat capacity but conducts heat a lot slower. Since a liquid cooling loop doesn't get anywhere near the boiling point or freezing point of water, you don't need it. Also glycols have slow down water flow. Anyway it is insignificant.

I worked with lots of coolants and hydraX was one of them. They are all the same product. Some form of glycol + dye. The dye can build up in parts of the loop if you have a highly restrictive block. But those GPU blocks look very smooth and should be fine. You aren't adding too much either. I just don't like the dyes because they stain everything like my hands and anything it gets on while providing pretty much no performance increase.

On the danger den site they used furmark to do a burn in test, and actually they got a max temp of 60C after 1/2 hour. I used MSI Kombuster which is actually more stressful. So I think my temps are decent for this block. Also, my thermalpaste still needs time to settle in, this morning I noticed my min temps lowered by 1C. But not sure it might be ambient that changed it.

Using distilled water will only net you a few degrees C difference. Here is a comparison showing a difference the coolant makes, only like 2-3C difference. I decided not to add the other 3/4 of a bottle yesterday, so my mix is mostly distilled water, so I'm guessing switching would only give me 1C max.

http://www.bit-tech.net/modding/2008/02/16/watercooling_fluid_shootout/4

You have to take a lot into account when you are doing comparisons, I did some research, and not sure what block you are getting, but your temps seem average compared to what other people are getting with watercooling on the 5970. A few things to take into account are-

1. Way you tested
2. ambient air
3. cooling block efficiency
4. flow rates (using two radiators actually adds restriction- so I might not benefit from them until I upgrade two SLI and add on my CPU)
5. fan CFM I originally wanted silverstone 110 Cfm fans, mine only push 75 Cfm, which makes a big difference - gonna upgrade when I can afford it
http://www.newegg.com/Product/Product.aspx?Item=N82E16811999344
6. Radiator efficiency

How did you test your maximum temps? You should try running MSI Kombuster @ 1600 x 1200 on Extreme burnin mode and post your max temps after 1/2 hour and post some screenshots. I'm curious about your system though, can you post screens and specs?
Either way, I think both our systems are average, and you really can't compare, as you are comparing apples to oranges. Just too many variables to do a direct comparison.

But I would still love to see your system, I love pc modding. do you have a link?

Everyone should post there systems pics in this thread.

http://www.tomshardware.com/forum/258655-29-pics-custom-rads-watercooling

Either way, thanks everyone for your comments. Love the feed back.

I might do a little more benchmarking & overclocking tonight and i'll post what I get, if I got time.

 

[cromedome]

dam straight.. it looks so jahs dude. and it would be so champ with the CPU water cooled as well..