EVGA X58 SLI LGA 1366 Motherboard Review

Voltage Regulation

The Voltage Regulation section is where we have a look at the motherboard from a voltage stand point. We don't always get what we set in the BIOS when it comes to voltage controls. This is a common misconception but one that needs to be investigated with every motherboard. It really isn't a fault if a motherboard doesn't supply what is selected, it is just a little bit of a hassle. We'll find out right now whether or not the EVGA X58 SLI tends to provide exactly what we set, or whether it is off by a small margin. To measure the actual readings with a digital multi-meter, we will be using the test points EVGA has provided in the photo below. The multi-meter used will be a calibrated UEI DM393. The numbers under load were obtained using Prime 95 on Blend with 8 threads utilizing all four cores and Hyper-Threading of the i7 processor. Let's get take a look at the results.

As mentioned and shown above, EVGA has provided onboard test points that can be easily accessed for measuring actual voltage of specific components. We have the ability to measure the vCORE, vTT, vNB, and vDIMM. We found that getting at these test points was a bit of a challenge and there was no way to continuously measure the voltages without using a set of hands. We are working on a method for mounting test probe holders but an easier option would be to simply solder leads to the test points for easier access. This is a nice first step for easy access to voltage without having to probe the motherboard but could be developed a little better on future models. Here is how the voltages break down from the BIOS through to Windows at idle and under load.

First off, we are fairly limited in what we can monitor because the BIOS only reports the four voltages we see above. This means we only have access to these four voltage readings in Windows using E-LEET. These happen to be the same four voltages that are available for direct reading from the monitoring pads shown above. The first voltage measured is vCORE and from the chart we can see that under load, vCORE actually increases. This is due to the fact that we had vDROOP disabled from within the BIOS. We will discuss this more in a little while. vDIMM in the chart above, unlike vCORE, shows a substantial drop in voltage going from idle to load. We also see that the actual voltage reading from the motherboard is substantially lower that what is being reported in Windows. The actual voltage reading for vDIMM is actually pretty much identical to what is set in the BIOS, except under load.

IOH vCORE or the NB voltage is similar to vDIMM in that it droops under load but reports exactly what is set in the BIOS. The actual reading is slightly lower that what is selected in the BIOS like the other voltages discussed thus far. The final voltage we are going to talk about is vTT and like the rest, the actual voltage supplied is less than chosen in the BIOS and the voltage drops from idle to load. vTT shows the largest discrepancy between what is reported in the BIOS/Windows and what is actually being supplied. This is rather important to know because vTT is responsible for a lot of the overclocking we are able to do with this motherboard in regards to uncore frequency and memory clocking. We will now take a quick look at the difference between vCORE with vDROOP enabled and disabled followed by a few charts of the other voltages to show their regulation.

Chart from OCCT - vDROOP Enabled

Chart from OCCT - vDROOP Disabled

In the chart above, we saw that with vDROOP disabled in the BIOS, the vCORE reading under load actually increased from idle. The charts above this clearly support those readings. Disabling vDROOP definitely works on the EVGA X58 SLI and will help you keep a lower vCORE at idle then provide the necessary boost in vCORE when the system enters a load state. The small spikes in the voltage charts are likely software anomalies and nothing to be concerned about. Now let's take a look at the voltage charts under the same conditions for vTT, vDIMM, and the IOH vCORE or vNB.

The charts above again support the table of results we looked at first and show a pattern that isn't exactly what we wanted to see. In all of the OCCT charts above we can see a very rippled voltage reading throughout the entire one hour test period. We can't be certain this is actually what is occurring because software voltage readings are limited by accuracy of their readings. None-the-less, a flat line would have been more comforting but at the same time, we can't hold a lot of stock in what we see above. The one thing we can hang our hat on is that VTT and vDIMM seem to droop a good margin under load. We would really have liked to see a more steady voltage from idle to load on these two voltages.