Why does it need to be divided by 9 when calculating UPI bandwidth on the Intel platform using UNC_UPI_TxL_FLITS.ALL_DATA event?

Question

In Intel perfmon, the calculation of UPI bandwidth is based on the following formula: [UNC_UPI_TxL_FLITS.ALL_DATA] * (64 / 9) / 1000000

According to Intel Perfmon:

UNC_UPI_TxL_FLITS.ALL_DATA: Counts valid data FLITs (80 bit FLow control unITs: 64bits of data) transmitted (TxL) via any of the 3 Intel® Ultra Path Interconnect (UPI) slots on this UPI unit.

Why is the bandwidth not divided by 8 (8 bits per Byte) but by 9 ?