Originally posted by sammorris:

---

Auto voltage usually has a maximum allowable voltage. For the 1.325V E4300 I had, neither of the boards I overclocked it with would auto a voltage any higher than 1.4V.
Manual overclocking and manual transmission is an interesting analogy, I kind of see where you're coming from with that, but to say a 'true overclocker' pushes all the settings, while true, is a bit of a farse these days. Automatic overclocking has got my i5 beyond 4Ghz nice and stable. The benefits you get from manual overclocking, in the terms of a realistic long-term stable setup, not just a benchmark muncher, are so slim nowadays it hardly seems bothering with. This is at least true of LGA775 and LGA1156/1366, whether it's true for AMDs I can't say but if not, that's a disadvantage, not really confirmation of the statement.
Not sure how you worked out the maths for the Q9550, 1.337-0.287=1.05V, and that's not the stock for them, it's usually 1.25V, I think the lowest they ever were was 1.20V.
Saying Vdroop is less in AMD gigabyte boards is purely coincidental, it depends on the load of the CPU, and how good the voltage regulation is. The more you spend the better it's likely to be. It's a bit like saying 'GMs are faster than Fords' or vice versa, it's a totally irrelevant statement without specific cases.

---

Originally posted by Estuansis:

---

Mmm loving my 790X-UD4P even more now. Took the ECO like they were match made :D

Russ, can you further explain where the AMD fan should go on my board? The NB or the VRM heatsink? I have the fan, and several others, from old and new coolers alike.

---

Originally posted by Estuansis:

---

Well I'm set then. Two of the Scythe SFF21Fs on the way as we speak.

---

Originally posted by sammorris:

---

You want a 7-blade fan like the SFF21F for a radiator really. The pressure produced by the slipstreams is too weak for that environment, you may as well just be using sn SFF21E instead of a 1200rpm slipstream, though of course the slipstreams are cheaper as they're lower quality fans.
You can often wire up graphics fans to a fan controller, especially if it's a two-pin with no fan control, but this method works with 4-pin PWMs too:


These adapters that don't have the cover, can be forced into the tiny fan power sockets that graphics cards use, allowing you to control the speed with a dedicated fan controller, or presumably with the motherboard. I see no reason why that wouldn't work though I haven't tried it myself.
I ran the fan on my Asus 3870 like this from my Zalman ZM-MFC1 so it was running at 5V. Much quieter than stock, and because the 3870s put out so little heat, there were no issues.
Of course, being Asus that card didn't last long anyway, but nonetheless it was useful while the card lasted.

---

Originally posted by sammorris:

---

Yeah as case fans the slipstreams are excellent, hence why I use six of them, though I did have reservations about their longevity when I bought them after having read some reviews. So far one has failed and been replaced, but the other five have been fine. They came cheap from an ebay store anyway, as did the replacement, so in terms of outlay they've been reasonably good. £40 for 6 case fans as good as the slipstreams are is pretty reasonable value, I paid that much for three FM121s, and almost that much just for two Noctuas [the original ones]. The Noctuas were terrible as well, the bearing whine was hideous, and it is something SPCR complained about in their review of one of their samples, but not the other.

---

Originally posted by sammorris:

---

Hmm, I tend never to use push-pull setups as they're not really suitable for air cooling, thus I can't say exactly what effect would be. I imagine rear exhaust rather than front intake would be better on a radiator for a 9-blade fan though as there's not the same issue of backpressure. Still, if it were me I'd be using a pair of S-Flexes, pricey though they are.

---