Cambricon MLU590 — aggregated analyst and brokerage estimates
Fields
sources/cn-cambricon-mlu590Notes
SECONDARY ONLY: Cambricon has NOT published an official MLU590 datasheet or product page. The flopper.io spec sheet shows all dashes (no data). Analyst estimates from awesomeagents.ai and bestofai.io report: ~192 GB HBM2e at ~2.4 TB/s, TSMC 7nm, 350-550 W TDP, ~512 TOPS INT8, ~80 TFLOPS FP32. These figures are mutually inconsistent across sources and are recorded with vendor_claim: false. The MLU590 is described as Cambricon's training flagship on MLUarch05 architecture, positioned against the NVIDIA A100. Treat all figures as provisional analyst estimates, not vendor specifications.
2026-10-10: set vendor_silence=aggregator-only. The record's own basis/notes already establish the reason; the field only names the case the schema now distinguishes. No number changed.
2026-10-11: Re-read flopper.io/gpu/cambricon-mlu590; confirms aggregator-only mechanism (page explicitly states Flopper has no verified throughput, TDP, or memory figures and Cambricon publishes no official datasheet).
Cross-references
This record references no other records.