Skip to content

v2.5.0

Latest

Choose a tag to compare

@jnioche jnioche released this 20 Oct 10:59
· 5 commits to main since this release

Highlights

This release changes the numbers for many servers and cloud instances, mainly those with GPUs, local disks or a lot of RAM. Expect different results from 2.4.x for:

  • GPU instances (embedded impacts). GPU impacts are now allocated by GPU count instead of vCPU share (#571). GPUs are now completed from gpu_specs.csv using their name (#591). Before, every GPU fell back to the default H100-sized archetype. Fractional GPUs, such as AWS g6f/gr6f and Azure NVas_v4, are now supported (#568, #571).
  • RAM (use phase). RAM power is modelled per DIMM (3 W + 0.055 W/GB) and depends on load, with an idle ratio of 0.7. It replaces the flat 0.284 W/GB (#581). Power is lower for large-DIMM configurations.
  • Local disks (use phase). SSD and HDD drives now draw power (HDD 6.4 W, SSD 8 W per drive), so storage-heavy instances get higher use impacts (#603).
  • More use-phase criteria. gwppf, odp, pm and lu now return values instead of "not implemented" (#583).

Breaking changes

  • The fe (final energy) impact is now labelled kWh instead of MJ. The values are the same, but anyone who read them as MJ got numbers 3.6× too low (#605).
  • The human toxicity criteria are now lower case: CTUh_c → ctuh_c and CTUh_nc → ctuh_nc (#602). Update any criteria parameters that use the old names.

Features

  • feat(data): support fractional GPU units and add the g6f instances by @jnioche in #568
  • Allocate GPU embodied impacts by GPU count, and support fractional-GPU instances by @jnioche in #571
  • feat(gpu): add more gpu data to API by @jonperron in #575
  • Model RAM power per DIMM and make it load dependent by @jnioche in #581
  • Return use-phase values for gwppf, odp, pm and lu by @simonvanlierde in #583

New cloud instances

  • Missing AWS metal instances by @jnioche in #573
  • Add Xeon 6 (Granite Rapids) AWS c8i, m8i, r8i variants and x8i, t8i by @jnioche in #578
  • Add AWS c7i-flex and p5en instances by @jnioche in #579
  • Add Graviton4 c/m/r8gb, 8gd, 8gn and i8ge AWS families by @jnioche in #577
  • Add AWS g7 and g7e instances by @jnioche in #609

Fixes

  • fix(data): use valid chkcsv types in cloud.fmt so numeric columns are validated by @jnioche in #566
  • fix(data): set GPU.units for NV-series Azure servers by @jnioche in #570
  • Fix spec discrepancies for some AWS instances by @jnioche in #564
  • Fix specs mismatches with AWS API by @jnioche in #572
  • Point server.csv GPU names at exact gpu_specs.csv entries by @jnioche in #576
  • Data cleanup by @jnioche in #590
  • fix(gpu): apply gpu_specs.csv values when completing a GPU from its name by @jnioche in #591
  • fix: correct Azure specs and compare NVMe in decimal GB by @CodeByPeace in #598
  • fix(server): give each RAM entry its own modeled power by @auraseb in #600
  • fix: remove commented code with hardcoded API key by @jonperron in #601
  • fix: use lower case for cthuh impact factors by @jonperron in #602
  • fix(server): record real local drive count for GPU platforms by @jnioche in #604
  • fix: express final energy (fe) in kWh by @jnioche in #605
  • fix: count local disks in use-phase power by @jnioche in #603

Tooling

  • feat(scripts): add check_references.py to validate cross-file data references by @jnioche in #567
  • Add compare_azure_instances.py to audit azure.csv against the Azure API by @jnioche in #589

Docs

  • chore(contributing): add few guidelines about using ai coding assistant by @jonperron in #594
  • Add RICO2 to adopters list by @greenshifuw in #606
  • Adds SDKs.md, a README link and a docs page for the PHP client of the API by @DavidJLN in #607

Chores

  • chore(deps): bump mkdocs-material from 9.7.6 to 9.7.7 by @dependabot[bot] in #569
  • chore(deps): bump anyio from 4.14.0 to 4.14.2 by @dependabot[bot] in #574
  • chore(deps): bump shell-quote from 1.10.0 to 1.12.0 by @dependabot[bot] in #584
  • chore(deps): bump brace-expansion by @dependabot[bot] in #585
  • chore(deps-dev): bump virtualenv from 21.5.0 to 21.7.13 by @dependabot[bot] in #586
  • chore(deps): bump multidict from 6.7.1 to 6.9.1 by @dependabot[bot] in #587
  • chore(deps): bump urllib3 from 2.7.0 to 2.8.0 by @dependabot[bot] in #588
  • chore(deps): bump aiohttp from 3.14.3 to 3.14.4 by @dependabot[bot] in #608

New Contributors

Full Changelog: v2.4.1...v2.5.0