Skinnee Labs - GTX480 Block Round-Up (working thread)

[Johnny87au]

Well... there will be no more testing of the GTX480...



A GTX480 has been sacrificed in the pursuit of data.

noooooooooooo, mind letting us know which blocks did shine out compare to the others whilst testing?

[wesley]

Boy do I feel ~

- Systemlord

No worries dude. All of us are learning new stuff too.

Wes

[Hondacity]

hmmm weird how that card died. mosfet short?...can you clean it up? and pics

sorry for your loss skinnee its a community loss as well...

[Systemlord]

No worries dude. All of us are learning new stuff too.

Wes

It's cool, if I can learn something new everyday I'm doing pretty well! I'm going to Google what Temperature Delta is for two things learned today. Mosfets is what they are, that answers my question.

- Systemlord

[Boogerlad]

ugh, that means you'll have to retest everything on a new 480, because chances are the sensors won't be calibrated to the same degree.

[Systemlord]

ugh, that means you'll have to retest everything on a new 480, because chances are the sensors won't be calibrated to the same degree.

He will not be retesting anything, below he stated...

Well... there will be no more testing of the GTX480...

A GTX480 has been sacrificed in the pursuit of data.

[skinnee]

Yeah, I'm not retesting... all the thermals are complete, was just plowing threw some stock cooler runs for a comparison. Test finished, idling down... walked back to my chair and continued working on images... poof, and then the smell of fried electronics.

[Vapor]

The last bits of data he didn't get to obtain were the least important of the bunch

He has decent (2 mounts rather than 3) stock clocks + stock cooling data that can be compared to his good stock clocks + waterblock

If there's a secondary component to this testing (MCW80 + unisink, MCW80 + stock sink bracket, etc., etc.), that can be done with a new card and maybe bring over and retest one or two fullcovers for those interested in lots of cross comparison

[Systemlord]

Yeah, I'm not retesting... all the thermals are complete, was just plowing threw some stock cooler runs for a comparison. Test finished, idling down... walked back to my chair and continued working on images... poof, and then the smell of fried electronics.

Do you expect the warrenty to kick-in for the toasted GTX 480?

[skinnee]

yeah, definitely don't need a full on retest for the core only blocks, just enough for a valid cross comparison.

Its shocking going from all TIM testing to stock cooler and seeing 93C for your GPU core.

warranty... no, I'm well past the boundaries of warranty.

[NOFX]

skinnee

When a you posting the results?

This is the second GTX480 card dead on this Month and mosfets again.

[Systemlord]

yeah, definitely don't need a full on retest for the core only blocks, just enough for a valid cross comparison.

Its shocking going from all TIM testing to stock cooler and seeing 93C for your GPU core.

warranty... no, I'm well past the boundaries of warranty.

How many water blocks in your testing could use 100% TIM without the need for thermal pads? I bet it's like going from a cold winter to summer in Death Valley!

[p2501]

Jebus I'm sorry about your loss. Do you plan a proper burial? Or, now that the card is dead do you have anyone that could go and try to change the burnt components to see if he could pull a frankenstein on that card?

[ell]

Can we please have an ETA Skinnee?

[Biffa]

Hey there,

It doesn't add up that way. If it's a 7c thing on 1 card, it's 7c difference on all 3 cards. Btw, the ~2-7c is just an approximation. Please don't take it as gospel.

Wes

Agreed, sorry if I confused anyone, but I meant 7°C on the entire loop would be important for me, especially if the loop included an i7.

Understood that we are talking ballpark here, but theoretically it would be important.

In a GPU only loop, maybe/maybe not. I haven't seen any Fermi blocks not able to keep a thrashed GTX480 sub 60° and a GTX470 sub 50° (depending on ambient of course)

This is just in my experience, I bow, obviously to the more extreme veterans experience, but I can only go on what I see in my own builds.

[JoeChuo]

Uuuh! GTX 480 dead! Skinnee .. you are real extreme! Looks like 100% TIM is no go for the long run. $400+ card to kiss goodbye!
Hope that EK block perform on average ..

[skinnee]

Can we please have an ETA Skinnee?

Wish I could nail it down for you, just working through all the writing each evening... soon.

[gmat]

Agreed, sorry if I confused anyone, but I meant 7°C on the entire loop would be important for me, especially if the loop included an i7.

If the GPU temps is higher, it doesnt mean the water temp is higher ! 7C more on the GPU wont affect other component temperatures, actually could be the water is even cooler due to less efficient heat transfer from the waterblock. Dont mix up the component temperature and coolant temperature...
Actually the flow restriction will affect more seriously the rest of the loop.

In a GPU only loop, maybe/maybe not. I haven't seen any Fermi blocks not able to keep a thrashed GTX480 sub 60° and a GTX470 sub 50° (depending on ambient of course)

And i have yet to see any significant difference in GPU overclocking using room temp watercooling, be it 40 50 or 60C... You have to go chilled or subzero for that.

[skinnee]

And i have yet to see any significant difference in GPU overclocking using room temp watercooling, be it 40 50 or 60C... You have to go chilled or subzero for that.

Air to water on the GTX480 has lots of OC headroom.

[Vetalar]

Air to water on the GTX480 has lots of OC headroom.

yep. fried mosfets already told us about that

[gmat]

Air to water on the GTX480 has lots of OC headroom.

What i mean is, do you reach the thermal limit on this chip ? This has not happened on GPUs yet (for conventional watercooling), usually at room level temperatures the OC limit is reached long before the chip overheats or throttles down (the VRM cooling is a separate issue).

[Systemlord]

What i mean is, do you reach the thermal limit on this chip ? This has not happened on GPUs yet (for conventional watercooling), usually at room level temperatures the OC limit is reached long before the chip overheats or throttles down (the VRM cooling is a separate issue).

That GTX 480 has been to hell and back, I'm not that shocked that it happened. Could have been the quality of the mosfets or maybe the mosfets had poor contact with one of the water blocks for a period of time. All I know is those mosfets were deep fried!

- Systemlord

[mk-ultra]

What i mean is, do you reach the thermal limit on this chip ? This has not happened on GPUs yet (for conventional watercooling), usually at room level temperatures the OC limit is reached long before the chip overheats or throttles down (the VRM cooling is a separate issue).

its not only about thermal limits, cooler core means higher stable oc

[UrbanSmooth]

Ah, I hear the graceful typing of one Skinnee pulling together ever most precious data compiled over the last 90 days or so. Kind sir, many await your prized data with baited breath and tender wallets! But, do take your time, sir, and make it the most perfect of presentations. Again, thank you!

[Blacky]

Air to water on the GTX480 has lots of OC headroom.

How much compared to air cooling ? Interested to know I hope you included OC tests with all blocks too ! Anxious to see results !