Dataset: 3825 CPUs

Variables: cores (X) and cpuMark (Y)

\[Y = \beta_0 + \beta_1 X + \epsilon\]

• \(Y\) = cpuMark
• \(X\) = cores
• \(\beta_0\) = intercept
• \(\beta_1\) = slope

model <- lm(cpuMark ~ cores, data = clean_data)
summary(model)

## 
## Call:
## lm(formula = cpuMark ~ cores, data = clean_data)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -82105  -1954   -694   1154  40715 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  -407.42     116.45  -3.499 0.000473 ***
## cores        1266.70      14.58  86.873  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5578 on 3823 degrees of freedom
## Multiple R-squared:  0.6638, Adjusted R-squared:  0.6637 
## F-statistic:  7547 on 1 and 3823 DF,  p-value: < 2.2e-16

\[\hat{Y} = -407.42 + 1266.7 X\]

R² = 0.664

More cores = higher performance

Slope: 1266.7 points per core

Model fits well (R² = 0.664)