Skip to main content
. 2015 Dec 15;9:464. doi: 10.3389/fnins.2015.00464

Figure 7.

Figure 7

Benchmark results comparing three hardware systems in terms of their performance when running as many neurons as they are capable of per 0.1 Watt of power consumption. Scatterplots show individual runs (with random jitter on the x-axis to avoid overlap), the shaded area is the mean plus or minus the standard deviation, and the 95% confidence interval of the mean is shown. All differences are statistically significant (p < 0.05; two-tailed t-test with Bonferroni correction; 400 samples per condition).