VARYC

Overview | Learn More

Below are example results that show the value of migrating to optimal controllers.

Fan Controller - 24 % Speed Reduction

The following tests are made on a motherboards from a popular motherboard vendor and an Intel processor. The purpose of the test is to compare an optimal predictive fan controller with the controller already used by the vendor.

24 % Fan Speed Reduction

The optimal predictive fan controller, designed by VARYC, outperforms the current fan controller used by the vendor by an average of 24 %, with a peak difference of 41 %.

.. at the cost of only 1.13 ℃

The temperature difference is very low compared to the large reduction of fan speed. Some systems have a strict temperature requirement and the controller can be tuned to accommodate this request while still reducing the average fan speed.

Multi Pump Controller - 25 % Power Reduction

The graph shows a derived power model for a multi pump system consisting of three pumps. The power optimization algorithm calculates, for every requested pressure, what pump speeds are needed if you use only one pump, two pumps, or three pumps, and what the power usage will be for every option. After the calculations the controller naturally chooses the scheme with the lowest power consumption. In this example running two pumps with 1800 RPM is the lowest power consuming solution. With these analyses the pump vendor knew exactly the benefits of an optimal controller, and we delivered an implementation strategy.

CPU Controller - 15 % Power Reduction

The objective was to compare an optimal CPU controller with the most popular CPU controller on the market.

15 % Power Reduction

The tests are performed with different CPU workloads. The above figure shows how the optimal CPU controller chooses the optimal power states without saturating the CPU.

The below figure shows the savings as a result of this optimal power control.

3 ℃ Temperature Reduction

Because the optimal CPU controller on average runs at a lower power state than the current CPU controllers, a significant reduction in heat dissipation from the CPU is seen. This heat reduction means power reduction on cooling and, as an added bonus, noise reduction.

VARYC CPU Controller Temperature Savings

Power Controller - 99 % Optimization

It is very important that power production companies produce the exact amount of power that is demanded. Overproduction results in wasted resources, and underproduction is of course not acceptable. We investigated if an optimal controller would result in a production that followed the demand more closely. The result was a simulated 99 % optimization.

Cookies