http://www.evga.com/images/hcftw/inside1.jpg
GTX 470
Load Heatsink: 83c
Load Waterblock: 38c
GTX 480
Load Heatsink: 95c
Load Waterblock: 49c
http://www.evga.com/forums/tm.aspx?m=271190
Printable View
http://www.evga.com/images/hcftw/inside1.jpg
GTX 470
Load Heatsink: 83c
Load Waterblock: 38c
GTX 480
Load Heatsink: 95c
Load Waterblock: 49c
http://www.evga.com/forums/tm.aspx?m=271190
Pretty shoddy machining, but at least it gets good temps. The heatpipe/VRM cooler is interesting....
Looks like Swiftech used the pin-matrix design they used in their XT.
Looks a bit dodgey. As long as it works I guess. I wonder what EK can come up with.
The delrin cover says Swiftech on it. :DQuote:
Originally Posted by dM@n
Yep. Similar to the XT.
Nice, didnt catch that. That must have been a killer deal for Gabe. I think Swiftech GPU blocks are the most innovative and its good to see EVGA feels the same way.
The XT has MUCH smaller pins. Glad to see pins in this hot GPU, better heat reduction. Channels just don't cut it no more I guess.
I don't want to start a water block argument, but the EK HF tested in Vapour's review performed better than the Swiftech XT. The EK HF has channels, the XT has pins, so that shows channels do still cut it, and in fact outperform pins. Further, the HF, while performing better, also has less flow restriction.Quote:
Originally Posted by Conumdrum
Well, maybe so. I'll say it looks like in the long run for overclocking controlling the VRM temps is what matters. In my case, I'll pretty much sit back and wait till the fall. See whats on the market. Be a good time to make a move before the next crazy Christmas announcements. Til then I'm done.
I think you are misunderstanding him, there is a HUGE difference between the very small channels on the HF, and the much larger channels found on most full cover vid card blocks, including the EK ones. I don't think he meant to mean that channels were worse period, just the way they are typically used in vid card blocks.Quote:
Originally Posted by CryptiK
I'm disappointed to notice how the heatpipe seems to have no direct flow...
http://www.evga.com/images/hcftw/channel.jpg
Anyway it's wrong to say that VRM temps are the only ones that count, this might be partially true for recent ATI cards but Nvidia's designs didn't ever rely much on VRM temps. GPU cooling is fundamental, especially on Fermi chips.
It surprises me however that most of you failed to notice how Swiftech used this very kind of pin matrix in all of their komodo blocks for a long time.
kool block shtty cooling for other chips. those unisink or w/e its call thingy dont cool as good as it looks either. i think vrm should b lil worry if it gets too hot no.. 280 vrms get hot with swiftechs unisink on them its hottest part on card 60c + . n u get gfx coruption flashesQuote:
Originally Posted by JaD
dd blocks full cover no lil unisink
I think the heat pipe to the block is enough. While the water doesnt directly cool the VRM's it does cool the heat pipe which is cooling the VRM's. This has been shown to be pretty effective with the old 4870 block and MCW60.
It's not a bad design, my only concern would be if the heatpipe had enough capacity for OC'ing.
Well, how did these work out:
http://www.techpowerup.com/img/08-10-27/mcw60-4870.jpg
Anyone know? Maybe Gabe could provide some VRM temps...
Hmm... so how will this compare to the DD
DD all Cu, gona cost more, but I'd go for that.
It's nice to see that they use Swiftech blocks for their watercooled models.
The question is if there's going to be a European model and if it's going to use the same Swiftech blocks?
Perhaps they did learn from their mistake with the EVGA GTX 285 Hydro Copper.
If I know swiftech, this is all Cu as well with a nickle or chrome plating.Quote:
Originally Posted by the finisher