specific method · filed under inference & serving
MXFP8 activation quantization
Represents intermediate activations in microscaled 8-bit floating point.
Also called Microscaling FP8 (MXFP8) activations, MXFP8 activations.
- sources
- 2
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 2
Documented in
Evidence
2 spans quoted from the sources, strongest treatment first.
MXFP8 activations
usedtraining objectivein Kimi K3Moonshot AI
MXFP8 activations: Intermediate activations use 8-bit floating point, providing higher precision where it matters most
usedmodel architecturein Kimi K3Moonshot AI
Filed alongside
Other methods under inference & serving :: inference quantization.
FP8 KV-cache quantizationFP8FP4 KV-cache quantizationNVFP4 quantizationFour-Over-SixMXFP4 weight quantizationPost-training quantizationQuantizationBlock-scaled INT8 quantization with stochastic roundingFP8 E4M3 quantizationGGUFMax-based scalingMSE-based scalingNVFP4 KV-cache quantizationNVFP4 ModelOpt re-quantizationSSM cache quantizationAWQ INT4 weight quantization (W4A16)BF16 inferenceBlock-wise E4M3 FP8 weight quantizationChannel-wise quantizationDynamic activation scalingEmbedding and KV-cache quantizationEmpirical bits-per-element budget selectionFine-grained FP8 quantization