[SOLVED] Fractal Design Define 7 Performance?

[Gregow]

Hey y'all,

I'm considering the Define 7 as a new home for my PC. I have an overclocked 8086K running around 1.34 volts that I cool with a 360mm radiator with three Noctua A12x25 fans. The GPU is an aircooled 1080Ti. Currently I have the Phanteks Enthoo Luxe case, and what I don't like about it is that I have to top mount the radiator. Meaning, I can either pull hot air from the GPU through the radiator and see my CPU temps rise, or I can get fresh air through the rad and not be able to sufficiently cool the GPU.

That's why I'm looking for a new case, where I can front mount the rad.

Now, I like Fractal Designs cases. I have an old one for my server and it's solidly built case. I also like the minimalist design. Being able to open the door for better airflow isn't such a bad idea either. Airflow however...

Well, according to the reviews I've seen it seems the front is a bit restrictive. Same thing goes for the dust filter. That's using the stock fans though, and as always reviews don't give the full picture.

So I'm asking you guys if you have any experience with these cases and especially if you have front mounted rads. What's the cooling performance like? Can you get really good thermals for both CPU and GPU?

 

[Phaaze88]

Was Gamers' Nexus Define 7 review one of them?

View: https://www.youtube.com/watch?v=XeTxUjUrw4A



IMHO, with minimalist cases like this and similar(NZXT's H500 comes to mind), what should be given priority to liquid cooling is the hottest part in the chassis, so that the airflow restrictions are less of an issue.
People tend to do it wrong by liquid cooling the cpu, leaving the gpu to run warmer in there, when the cpu comes nowhere close to the gpu's output.
In your situation, the 8086K VS the 1080Ti. Even overclocked, I'd bet the 8086K doesn't even exceed 200w - it's a golden sample 8700K after all.
It's likely closer to 175w in the worst case scenario; Prime95 small FFTs, AVX off.
The 1080Ti, on the other hand, 300w easy - even more with some of the fancier models.

 

[Gregow]

Yes, GN's review is one of them. It's a good review but it doesn't tell me what to expect when using a front mounted radiator with good fans. I mean, my Noctuas might do a better job of overcoming the restrictions of the dust filter.
Would also be interesting to know if there are cheap mods to replace the filter with something less restrictive.

Now, I don't intend to water cool the GPU. I will mod it with two 120mm fans, which should give good thermals provided it gets enough fresh air.

 

[Karadjgne]

For front restricted cases, which is pretty much any without a full mesh front, for a rad, use Pull. It's far more effective at lower rpms than push, is not reliant on a skinny amount of room between fan and door, you automatically add an extra 25mm of airspace. Most will use push anyway since that's how rgb/argb fans are setup. Fashion over function.

Also benefits the gpu since the fan now has unrestricted airflow directly to the gpu instead of loosing pressure through the rad.

 

[Gregow]

Sure, that makes sense. Question remains though: How good airflow can you get in practice with this case?

 

[Karadjgne]

I have a R5 Window. Used a kraken X61 and the 2x stock 140mm (swapped them around a bunch for testing purposes), so 4 fans total, never bothered with a rear exhaust, no point with a top mounted exhaust. Door open or closed really didn't make much difference, still got better temps at 4.9GHz on i7-3770K than with 5 fans and Cryorig R1 Ultimate.

So I'd hazard a guess the airflow is ok, but not spectacular, more beneficial for liquid cooling over air. But then again, my fans never go over 700rpm unless I'm stress testing, so higher rpm fans with better cfm might make a difference.

All hdd bays removed, so R5/R6/R7 share similar airflow properties, the differences are more on the tooling/cosmetic side of things.

Even with the door closed, there are still many cases with worse airflow, at least with the top dampers unused ports can be sealed off preventing most bleed suction.

 

[Gregow]

What voltage are you running on the CPU and what temps do you get on CPU and GPU when stress testing (like running Prime95 and 3DMark, for example)?