Sustainable Computing and Green Data Centers

Module: Sustainable Tech | Difficulty: Premium

Power Usage Effectiveness

Carbon Usage Effectiveness

Comparison

Python Implementation

import numpy as np

class GreenCloudOptimizer:
    def __init__(self, pue=1.5, renewable_fraction=0.3):
        self.pue = pue
        self.renewable = renewable_fraction

    def carbon_intensity(self, grid_ci):
        return grid_ci * (1 - self.renewable) * self.pue

    def workload_scheduling(self, workloads, carbon_forecast, deadline):
        n = len(workloads)
        scheduled = np.zeros(n)
        for i in range(n):
            if carbon_forecast[i] < np.percentile(carbon_forecast, 30):
                scheduled[i] = workloads[i]
            elif carbon_forecast[i] < np.percentile(carbon_forecast, 70):
                scheduled[i] = workloads[i] * 0.5
        return scheduled

    def server_consolidation(self, utilization, threshold=0.6):
        active = utilization > threshold
        migrated = []
        for i, u in enumerate(utilization):
            if not active[i]:
                target = np.argmax(utilization[active]) if np.any(active) else 0
                migrated.append({'from': i, 'to': int(target), 'load': u})
        return active, migrated

Research Insight: Carbon-aware workload scheduling reduces cloud computing emissions by 20-30% with minimal performance impact.