The test protocol:
To test the performance of our Alphacool Core 1, we’re going to pit its results against two other waterblocks we use regularly: the EK-Quantum Velocity² and the Optimus Foundation. These are the two waterblocks I use most often in my tests, so it seemed logical to pit the Alphacool Core 1 against them. For lack of time, we haven’t yet had a chance to add the other waterblock we received from Alphacool, the Eisblock XPX Pro Aurora. We’ll tell you more about it when we test the Eiszeit 2000 chiller.
Test configuration :
Our three waterblocks will be tested on a ROG Z790 APEX motherboard, combined with an Intel i9-13900KS processor. This processor has 8 P-Cores and 16 E-Cores for a total of 32 threads. For the tests, and to prevent the boost frequency from distorting the results, we’ll set the P-Core frequency at 5.6 GHz, the E-Core at 4.0 GHz and the voltage at 1.20 volts, with the LLC at level 7.
Temperatures will of course be high, as we’re going to use a stress test that will load our processor at 100% continuously for 10 minutes with Cinebench R23. Of course, in everyday use of your configuration, it’s never the case that all the cores are called upon for such a long time at the same time! But the idea here is to have an identical basis of comparison for all 3 blocks.
Cooling loop components :
We’ve already presented them in our three tutorials on installing a liquid cooling loop. The radiator is an EK-CoolStream XE 480, measuring 520 x 130 mm and 60 mm thick. It is equipped with four 38 mm thick EK-Meltemi 120ER fans. These fans feature extremely high static pressure and are ideally suited to XE series radiators .
Still with a view to comparative tests, these will run at maximum speed during our tests, in order to maintain an identical protocol with the three blocks. The same applies to pump speed.
Unless otherwise specified by the manufacturer, the waterblock will be positioned so as to be visually “straight” when your motherboard is inserted into your case. We have of course respected the manufacturer’s input and output specifications for each block.
Testing each waterblock :
While we were thinking about setting up this protocol, the question of thermal paste came up. Firstly, we won’t be using the pre-installed paste or the supplied syringe, but will opt for an identical paste. Our choice is Noctua’s NT-H1 thermal paste. We will carry out three series of tests. We clean the IHS of our processor with Noctua NA-SCW1 wipes between each test.
Stress tests with R23 :
The software we’re going to use to warm up our i9-13900KS is the Cinebench R23 benchmark. The special feature of this benchmark is that it can be repeated in a loop, and the user can choose the length of the loop, which in our case will be 10 minutes.
Once the waterblock is in place, we’ll let the configuration run for 30 minutes to allow the water temperature to stabilize. We will then carry out a 10-minute stress test to take the first reading.
The configuration will then rest for 15 minutes, before running the second 10-minute stress test. We’ll do the same for the third reading.
Several readings will be taken during this stress test:
- Liquid temperature, using a probe placed in the tank and our Fluke 51 II thermometer.
- The temperature of our processor via Core Temp software version 1.18, which will give me the temperature of each core and enable me to calculate an average. As the tests are carried out three times, we’ll have three averages in all.
Current temperatures :
It’s never easy to take temperature readings in the heatwave we’re currently experiencing in Belgium. Fortunately, we carried out our tests last Sunday, a rainy day, but we still had a temperature of 24.5°C in the lab. Having also received the Alphacool Eiszeit 2000chiller , we’re certainly thinking of using it for our next comparisons, by locking the water temperature at a certain level.