The interaction between nanofluids and porous structured heat sinks is not well documented. In addition, the micro-miniaturization of electronic components has led to an increase in the heat dissipation rate of electronic chips. As a result, researchers are now searching for new cooling techniques that allow for the rapid removal of the heat generated from electronic components. The purpose of this study was to investigate the heat transfer characteristics and thermal performance of an ERG aluminum foam heat sink for the Intel core i7 processor. The aluminum foam heat sink was subjected to a steady flow of gamma-Al2O3-water nanofluid covering the entire non-Darcy flow regime (210-631 Reynolds numbers). The gamma-Al2O3 nanopartide concentrations (volume percent) ranged from 0.1 vol% to 0.6 vol%. The experimental results revealed that the average Nusselt number was enhanced by 20% when the ERG aluminum foam was used (compared with the empty channel). The results also revealed that at low volume fractions of 0.1 vol% and 0.2 vol%, the gamma-Al2O3-water nanofluids led to an enhancement in both the local and average Nusselt numbers. The maximum enhancement of the heat transfer rate was achieved at 0.2 vol% and there was a sudden drop in the positive effect at 0.3 vol% (compared with pure water). The positive effect then showed a slight increase along with increases in the volume fraction of the nanoparticles up to 0.6 vol%. This heat transfer enhancement trend due to the presence of nanofluids is related to the mechanism which causes the superior thermal properties of nanofluids. When compared with pure water, optimal removal of the heat generated from the surface was achieved at a 0.2 vol% concentration of gamma-Al2O3-water nanofluid. The average enhancement percentages of the Nusselt number at a 0.2 vol% nanofluid concentration compared with pure water were 37% and 28% at Reynolds numbers of 601.3 and 210, respectively. The numerical results were in good agreement with the experimental data of local Nusselt number and local temperature with a maximum relative error of 3% and 2%, respectively. (C) 2016 Elsevier Ltd. All rights reserved.