Verified65% confidenceFactExact time
MXFP4 mitigates numerical overflow and precision degradation at 4-bit low precision by sharing a single scaling factor within small blocks of data (typically 32 elements), theoretically reducing memory footprint and bandwidth requirements by approximately 4× compared to FP16.
3
Sources
65%
Confidence
Long-term
Relevance
7/17/2026
First Seen
Sources
Related Claims
UnverifiedThe MXFP4 format has a group of elements (typically 32) sharing a single 8-bit scaling factor, with each element represented by a 4-bit micro floating-point number (1 sign bit + 2 exponent bits + 1 mantissa bit).79% similarVerifiedQuantization technology compresses model weights from FP32/FP16 to low-precision formats such as INT8/INT4, which can reduce model size by 50%–75%.71% similarVerifiedQuantization reduces FP32 (32-bit) weights to INT4 (4-bit), achieving a theoretical compression ratio of 8×, but introduces precision loss.70% similarUnverifiedTaking the Q4_0 format as an example, every 32 FP32 weights are packed into one quantization block containing one scale factor and 32 4-bit integers, averaging approximately 4.5 bits per parameter, reducing memory usage to approximately 1/8 of the original.70% similarUnverifiedA 70B model can be compressed to approximately 35GB at INT4 precision.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/544941API
curl https://kongchang.com/api/v1/knowledge/claims/544941MCP
get_claim(id=544941)