Swiftech H220 Review by TTL.

[eshbop]

You're all missing it. TTL tested them at stock configs, minus the TIM. Just look at the H100i fans, 7 large blades are generally better for static pressure, compared with the the 9 thinner blades on the H220. It allows less air to come back through the fans, that's just how it works (just look at the Noctua NF-F12s, Noctua obviously know what their doing). Also, the H220 is meant to run quieter (and therefore lower fan RPM) than the H100i while giving the same (+/- a few) degrees, so it is no surprise he got these results. Of course the Swiftech unit has a better pump, but in reality this is going to effect the temps very little, the all copper rad & CPU block design is much more important. Surface area, CPU block & rad composition, that's what's important. The reason for the better pump is simply because of expandability, no more. And IMHO he is 100% right in having a go at the long tubing, there's no need. Even if you were to upgrade to a full loop in the future, you're going to get ride of the supplied tubing because it's not going to match the other tubing you've bought. Also, next to 100% of people installing this are going to mount it in the top of their case, fact. So why isn't the tubing shorter? Yeah you can cut it down to length, but that just means draining, filling, bleeding, etc. By the same logic the few people that want to mount it elsewhere can buy longer tubing (it's not that expensive folks) if they so desire. It just looks silly with all that extra length. As I said, yeah you can cut them down, but who buys an AIO (that's what this unit essentially is) planning to upgrade in the future only to have to cut tubing and refill it on day one?

And all this makes sense when you look at Linus' results using the same TIM and fans for all the units he's tested. Of course the H220 is going to perform better with this config, that's a no brainer. But TTL wasn't testing the units like this. With lower RPM (& therefore noise) fans that the H220 is supplied with I'm not at all surprised with the results he got.

It's a parallel when testing the H110/X60 with the H100i, which perform almost equally but the H110 and X60 are quieter, that's their selling point. Their 140mm fans don't run as fast, nor generate the same static pressure as 120mm fans (even 140mm Noctua vs 120mm Noctua, the 140's are far behind in static pressure) but the larger rad surface area make these units equal.

[Mysticode]

You're all missing it. TTL tested them at stock configs, minus the TIM. Just look at the H100i fans, 7 large blades are generally better for static pressure, compared with the the 9 thinner blades on the H220. It allows less air to come back through the fans, that's just how it works (just look at the Noctua NF-F12s, Noctua obviously know what their doing). Also, the H220 is meant to run quieter (and therefore lower fan RPM) than the H100i while giving the same (+/- a few) degrees, so it is no surprise he got these results. Of course the Swiftech unit has a better pump, but in reality this is going to effect the temps very little, the all copper rad & CPU block design is much more important. Surface area, CPU block & rad composition, that's what's important. The reason for the better pump is simply because of expandability, no more. And IMHO he is 100% right in having a go at the long tubing, there's no need. Even if you were to upgrade to a full loop in the future, you're going to get ride of the supplied tubing because it's not going to match the other tubing you've bought. Also, next to 100% of people installing this are going to mount it in the top of their case, fact. So why isn't the tubing shorter? Yeah you can cut it down to length, but that just means draining, filling, bleeding, etc. By the same logic the few people that want to mount it elsewhere can buy longer tubing (it's not that expensive folks) if they so desire. It just looks silly with all that extra length. As I said, yeah you can cut them down, but who buys an AIO (that's what this unit essentially is) planning to upgrade in the future only to have to cut tubing and refill it on day one?

And all this makes sense when you look at Linus' results using the same TIM and fans for all the units he's tested. Of course the H220 is going to perform better with this config, that's a no brainer. But TTL wasn't testing the units like this. With lower RPM (& therefore noise) fans that the H220 is supplied with I'm not at all surprised with the results he got.

It's a parallel when testing the H110/X60 with the H100i, which perform almost equally but the H110 and X60 are quieter, that's their selling point. Their 140mm fans don't run as fast, nor generate the same static pressure as 120mm fans (even 140mm Noctua vs 120mm Noctua, the 140's are far behind in static pressure) but the larger rad surface area make these units equal.

Linus' testing has pretty much become "let's test a rad and pump", by using the same fans in all tests.

It's a no brainer to me that with some quiet but yet powerful fans like the Noctuas, the H220 will do better. It's using a watercooling enthusiast grade radiator and pump.

[Helltech]

H100i pull nf-f12 x2 71C

H220 pull stock fans 82.1C (pump 100%)

H220 pull nf-f12 x2 61.3C (pump 100%)

H220 pull nf-f12 x2 61.9C (controlled pump)

Noctua fans realy make difference...

Swiftech H220 Liquid Cooler Review Linus Tech Tips

So Tom's results weren't wrong or false by any means, just crap fans by Swiftech, and this is why I trust Tom, he has no reason to lie. His results were confirmed by Gabe also.

[peanutsrevenge]

Unsubscribed.

The award was upped to a gold due to the price (which I believe is still more than H100i, although it's yet to be stocked in most e-tailers).

It wasn't so much the hose length as the fact it came out the sides that TTL didn't like (though length clearly annoyed him).

The H100i Can control the fans by itself in reaction to coolant temp, where the swiftech requires additional set up.

It was the pump that made the most noise I believe, which it seems is too powerful at max chat for the loop/enclosure.

There will be more investigation done by Linus and co and hopefully TTL also as this product's been rather weird.