|
Work item:
|
L.QA_AI
|
|
Subject/title:
|
Quality assessment and verification of AI computing power resources for reuse
|
|
Status:
|
Under study
|
|
Approval process:
|
AAP
|
|
Type of work item:
|
Recommendation
|
|
Version:
|
New
|
|
Equivalent number:
|
-
|
|
Timing:
|
2028 (Medium priority)
|
|
Liaison:
|
SG13, SG21, SG20
|
|
Supporting members:
|
State Grid Corporation of China, Wuhan University, China Unicom, CAICT, China Mobile, China Telecom
|
|
Summary:
|
This Recommendation specifies assessment framework and procedure for evaluating the performance of AI computing power device for reuse, including GPUs, NPUs and TPUs. An assessment framework is defined, covering baseline operational requirements. The Recommendation provides detailed test procedures and key performance measurement.
By enabling data-driven decisions on whether a device can be reused, should be disassembled for component reuse, or sent for material recovery. This Recommendation directly supports the effective reduction of carbon emissions for AI computing, the reduction of electronic waste, and the extension of device operating lifetime.
|
|
Comment:
|
-
|
|
Reference(s):
|
|
|
Historic references:
|
|
Contact(s):
|
|
| ITU-T A.5 justification(s): |
|
|
|
|
First registration in the WP:
2026-07-01 14:22:54
|
|
Last update:
2026-07-06 18:12:34
|
|
