There's a good chance that the iBUYPOWER Element's production at God Speed Casing was seen by Jonsbo earlier this year, when the SI paired with NZXT to build an S340 with tempered glass paneling in the front and side. Since then, Jonsbo has begun producing its QT03 chassis, which bears striking resemblance to Rosewill's Cullinan and Anidees' AI Crystal.

It's “striking” because it's the same. Jonsbo does, convincingly, market its QT03 with the catch-phrase, “chasing on the road, have you already forgotten the beginner's mind?”

With copy like that, it's hard to believe that more manufacturers haven't ordered from the OEM.

Jonsbo is the OEM, here, but that doesn't diminish the fact that the Cullinan is generally a well-designed case – at least, for aesthetics. We'll look at the thermal benchmarks momentarily. The enclosure remains the most modern of Rosewill's current case portfolio, and is poised to attract more eyes to Rosewill's case lineup. This OEM/manufacturer relationship isn't new to the industry, either; we've exposed most of the liquid cooler suppliers and their partners (e.g. Asetek, CoolIT, and relations to Corsair or others), but the existence of a single supply doesn't invalidate a manufacturer's contributions. These companies are often involved in making change requests during development, instituting custom design or firmware changes (like Logitech's changes to the ADNS-9800 sensor from Avago, or Corsair's to Asetek's supply), and still carry recognition and support/warranty systems that matter.

The point is: It's not bad that Rosewill and Anidees are both using Jonsbo as a (likely) supplier, but it's still good to know as a consumer.

Regardless, we're reviewing the Rosewill Cullinan case today. If you're here to learn about our thoughts on the Anidees AI Crystal or Jonsbo QT03, you're in the right place – our thoughts will apply (nearly) equally to all three of these enclosures.

Rosewill Cullinan Specs

The Cullinan has an expected street price of $150. For hard specs, the unit is using 5mm thick tempered glass side panels for the left and right, each secured by flathead thumbscrews in four locations. 4mm tempered glass is mounted within the front plastic housing (no screws). A silicone damper sits between the left/right screws and the glass to prevent unpleasant metal-on-glass contact, with the glass itself sitting loosely on a standoff.

Obstructing the front with tempered glass obviously prohibits any air intake, but this isn't any different from what we're seeing on modern cases making use of sleek plastic or steel fronts, either. The Corsair 400C, 600C, and NZXT H440 and S340 cases all uses flat fronts with no mesh, and then leverage the depth between the front of the front panel and the case frame as side intakes for front fans. It's an approach to design that allows for intake, a cleaner front, and theoretically reduced noise exposure to the user by directing fan noise out and away from the front of the case, toward the back, rather than leaking noise out a front mesh.

To make this work, however, the enclosure needs to have some depth between the front panel and the frame, and a grill within that cavity. We've found that a grill of about 1” is ideal for 'breathability' – the same is also true for bottom intakes on elevated cases – and the Cullinan meets this recommendation with a 1.2” gap.

For fans, the Cullinan includes 3x 120mm front fans with blue LEDs and 1x 120mm rear fan with blue LEDs. Optionally, 3x 120mm top fans can be fitted, though we've generally found that cooling performance worsens when using top exhaust fans with a tower cooler, so it'd be best to dedicate that slot to a radiator.

The Cullinan supports radiators up to 360mm long in the top and front. A 120mm radiator can be fitted in the rear; although 140mm fans are supported in the rear slot, not all radiators will fit depending on tank size.

Optionally, the front fans could be replaced with 140mm units, though you're really not gaining much from that endeavor other than potentially some noise reduction. The front fans are already fairly maxed out on their usage of the entire front surface area, with 140mm ultimately bleeding some of the air channel over the steel paneling.

The case's construction is entirely steel. Without the glass, the enclosure is fairly lightweight and uses 0.7mm thick SGCC steel. The glass helps with structural rigidity when mounted to the case.

Test Methodology

We tested using our Ivy Bridge test bench, detailed in the table below. This particular configuration has been retired and brought back into service, following some changes to our hardware availability. We’ve moved from a one-bench-fits-all setup (which uses a Haswell config in an mATX board) to multiple benches, one for mATX and small towers and one for ATX / mid-towers. This bench is for the latter.

Conducting thermal tests requires careful measurement of temperatures in the surrounding environment. We control for ambient by constantly measuring temperatures with thermocouples and laser readers. We then produce charts using a Delta T(emperature) over Ambient value. This value subtracts the thermo-logged ambient value from the measured diode temperatures, producing a delta report of thermals. AIDA64 is used for logging thermals of silicon components, including the GPU diode.

All case fans are manually configured to their maximum throughput using BIOS. If a fan controller is present, we opt-in and test on multiple settings. This forces testing of case fan performance in addition to the case's air channeling and airstream design. This also ensures minimal variance when testing, as automatically controlled fan speeds can reduce reliability of benchmarking. The CPU fan is set to use a custom fan curve that was devised in-house after a series of testing; setting the CPU fan to its maximum speed can limit the disparity shown from case-to-case as the CPU cooler is extremely efficient, and will create a ceiling for thermal performance if bottlenecked.

The CPU is overclocked to 3.6GHz with all power saving features disabled and EIST disabled.

The video card is configured to run at 55% fan speed at all times. Stock clocks are used. We employ an NVIDIA GTX 980 reference for the case test bench.

Prior to load testing, we collect idle temperature results for ten minutes to determine the unloaded cooling performance of a case's fans and air channels. Thermal benchmarking is conducted for twenty minutes, a period we've determined sufficient for achieving equilibrium. The over-time data is aggregated and will occasionally be compiled into charts, if interesting or relevant. The equilibrium performance is averaged to create the below charts.

Load testing is conducted using Prime95 LFFTs and Kombustor “Titan Lake” stress testing simultaneously. Testing is completely automated using in-house scripting, and executes with perfect accuracy on every run.

We recently validated our test methodology using a thermal chamber, finding our approach to be nearly perfectly accurate. Learn more here.

Continue to Page 2 for the thermal results and conclusion.