Skip to content

Release Notes — v0.5.0

Release Date: 2026-03-28 Type: Minor Release

Overview

v0.5.0 delivers high-precision hardware analysis with core-aware CPU normalization, deep RAM generation tracking, optional Gemini AI insights, and full Apple Silicon support.

Added

Smart Core Normalization

A core-aware utilization tracking system properly scales across multi-core processors (1 to 128+ logical cores). This eliminates the utilization inflation effect in multi-threaded Python applications and ensures scientific accuracy for real-world production workloads.

RAM Generation Tracking

EcoTrace now detects RAM type (DDR4, DDR5, LPDDR) and clock speed to apply type-specific watt-factors using RSS-based recursive process tracking. Energy estimation for data-heavy tasks is significantly more accurate as a result.

Gemini AI Insights (Beta)

Optional integration with Google Gemini AI. When enabled, PDF reports include:

  • Hardware-aware optimization recommendations tailored to the specific CPU and GPU model.
  • Pythonic refactoring suggestions (async, vectorization, library-level swaps) to reduce carbon spikes.
  • Actionable sustainability advice embedded directly in the audit report.

Requires pip install ecotrace[ai].

Apple Silicon Support

Full native support for M1, M2, and M3 series architectures with specific power profile mappings.

Architectural Improvements

  • Engine refactored into distinct cpu, gpu, and ram intelligence modules.
  • Monitoring daemon hardened to ensure minimal impact on production performance.
  • Strict input validation for gpu_index and region_code parameters.

How to Upgrade

pip install --upgrade ecotrace