Thermal load of a dual-fan graphic processing unit (GPU) system on a server rack

Kamarudin, Muhammad Hazim (2022) Thermal load of a dual-fan graphic processing unit (GPU) system on a server rack. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Thermal load of a dual-fan graphic processing unit (GPU) system on a server rack.pdf - Submitted Version Download (403kB)
	Text (Full text) Thermal load of a dual-fan graphic processing unit (GPU) system on a server rack.pdf - Submitted Version Restricted to Repository staff only Download (3MB)

Abstract

Graphics Processing Unit (GPU) has gone through many revolutionary changes throughout the decade. GPU’s processing power can challenge the existing Central Processing Unit (CPU) in the task of running high-profile software. However, any electronics that flow current into it will generate heat. The thermal problem is the major problem for GPUs. The optimal operating GPU requires minimum fan speed and core temperature with the highest efficiency (hashing power). This research focused on obtaining the optimal response at a certain core clock and memory clock via the optimization tool of Design-Expert software. The response was recorded while applying constant amount of power supplied to the GPU. The GPU used is in the category of dual fan which is Gigabyte GeForce GTX 1070 WINDFORCE OC 8G. The relationship between the clocking (core and memory) and GPU responses (fan speed, core temperature, hash rate) were observed. By that, Central Composite Design (CCD) generated one equation for each fan speed, core temperature, and hash rate. Next, the optimization process suggests several new clocks settings that give the best performance than the current setting. For the validation and confirmation process, the best one core and memory clock was selected. The results from the validation process prove that the predicted response was precise with less than 2% deviation.

Actions (login required)

View Item