Review: Distroplate for OBT by Niuulh’s Workshop

Today we are testing an OBT bench top distroplate made by Niuulh’s Workshop. This is a product created for a specific request. The goal is to centralize the inputs and outputs in one area using quick disconnects.

This distroplate is available for 120 € in acrylic. The acetal version that we have will not be available for sale for the moment.

[nextpage title=”The distroplate for OBT:”] We start the tour of the product by the internal design of it. Here, we have two inputs / outputs on the ends that cross the distro. The two middle slots are connected to each other.

And here is the distroplate mounted with the QD3 quick disconnect from Koolance. We can notice some imperfections on the outside machining, imperfections that would come from the acetal material (POM) too soft for the current machining. Consider acrylic (PMMA) for a better finish.

[nextpage title=”The test configuration:”] The test configuration used here will be the new basis for future thermal paste and waterblock tests. Not much has changed from the previous configuration. There is the addition of a couple of QDC QD3s and the addition of the distroplate for OBT. This will have an impact on flow and thermal performance which we will quantify.

Here is the configuration that will be used for the test:

• Motherboard : EVGA Z690 DARK KINGPIN (BIOS 1.15)
• Processor : INTEL CORE I9 12900KS
• Memory : 2x16gb SK Hynix A die
• Graphics Card: NVIDIA RTX 3080 Founder Edition
• Cooling system : Watercool Heatkiller IV PRO + MO-RA3 420
• SSD: Western Digital SN850 500gb
• Power supply : Be Quiet ! Dark Power Pro 12 1200W
• Operating System: Windows 11 64 bits (version 21H2)
• Display: DELL S2421NX

The cooling is provided by a Watercool MO-RA3 420 radiator ventilated by nine Arctic P14. The pump used is a Watercool D5 PWM mounted on a Watercool Heakiller D5 200mm tube tank. The whole system is controlled by an Aquacomputer Octo and an Aquacomputer high flow USB meter. Here are the components of this cooling loop and their respective prices:

[nextpage title=”The New Protocols:”] The objective of this test is not in itself to see the performance with this distroplate for OBT but rather to quantify its impact. Here, we will compare the measurements obtained with this new system to those obtained a few weeks ago on the old one. We will therefore directly compare throughputs and temperatures to the old test results of Arctic MX-6 and Eisblock GPX-N RTX 3080 FE.

Protocol for CPU measurements :

The I9 12900KS is set to get 250W of load on cinebench R20. Pcore is at 5200MHz, Ecore at 4000MHz and cache at 4400MHz for a voltage of 1.335V bios LLC +25%.

Thermal paste applications are done crosswise on the CPU, this is the method that works best with all thermal pastes on the market. The waterblock is mounted and tightened to the screw stop to let the springs apply the right pressure.

Then, a cinebench R23 stress test is run to warm up the thermal paste with 10min of heavy load. To finish, three R20 cinebench are launched in a row. While taking well into account to have the same water temperature between each beginning of benchmark.

Each thermal paste is applied three times with three cinebench R20 per application. The minimum delta T° value between the water and the CPU is recorded among these nine measurements.

Protocol for GPU measurements:

The RTX 3080 FE is set to a power limit of 320W, without overclocking.

The thermal pastes are applied by spreading on the GPU, this is the best method for a GPU use. The waterblock is mounted and tightened to the screw stop to let the springs apply the right pressure. The backplate is mounted to ensure the rigidity of the whole.

Next, we run the Superposition benchmark in 1080p Extreme three times in a row. Temperatures are taken and then a dynamic stress test of 3 x 12 hours is run to simulate an accelerated gaming type of aging. At the end of this test, we run the Superposition benchmark in 1080p Extreme three times in a row.

Each thermal paste is applied only once with a pre-test check. The minimum delta T° value between the water and the GPU is taken before and after the 3 x 12h aging.

Note:

The GPU protocol mentioned above is not applied for this test. However, the card has aged more than 36h with this thermal paste but no disassembly has been done because of lack of time.

These protocols will nevertheless be implemented for the next thermal paste tests.

[nextpage title=”Comparative tests:”]So we arrive at the crucial moment, when is the new performance with this distro? The result is better than expected, we have to start with a flow rate that has gone from 220 L/h to 180 L/h. A minimal decrease that will have little impact on the performance of the waterblocks used here.

The second big change comes from the cpu coldplate, which has been upgraded to a nickel-plated PRO version. We don’t notice any significant differences, however, with the same temperatures on Cinebench R20 and 0.5°C less on Cinebench R23. Thus, we could consider not redoing all the measurements made before the change of configuration. But, this would not be very professional.

Regarding the performance on GPU, we have here almost the same cooling performance as we had before. We will take advantage of the next thermal paste tests to integrate the graphics card to the thermal paste tests.

The tests were performed with Arctic MX-6 for the CPU and KFA² TG-002 for the GPU. [nextpage title=”Conclusion:”]We are coming to the end of this test of a very different product than what we usually test. With this distroplate mounted on our bench and the QDCs mounted on it, it will be even easier to change configuration in the future. The pictures will look even better thanks to the neat pipes. Regarding the performance, it is not really impacted by the loss of flow. But, a switch to a double pump D5 solution could be considered to give more flexibility in flow for future waterblock tests.

As a reminder, this distro for OpenBenchTable is available for sale for 120 €.

Niuulh’s Workshop OBT distroplate

We liked:

• The low throughput after installation
• The performances are still there
• The reactivity for the creation of a new product

We would have liked :

• A lower price
• A cleaner surface machining

This first distroplate for OBT opens the way to products that nobody really needs. But, that every MO-RA3 and OBT owner would like to have. In this version we have one system in series but it is also possible to have two systems in parallel on both sides of the OBT for the IN and OUT. It is in any case a favorite for this first distrOBT by Niuulh’s Workshop.