Measurement

MEM Score

One number on a log scale. The reference machine scores 50; ten times the reference scores 100.

0.1x1x10x050100throughput relative to the reference machine (log)reference, 502x, 6510x, 100GTX 1080 Ti, measured 77.6
Every doubling of effective throughput adds about 15 points. The scale is open-ended and clamped at zero.

Formula

Score = 50 + 50 · log₁₀( Π (xᵢ ÷ refᵢ)^wᵢ )

Each metric is divided by a fixed reference machine, weighted, and combined as a weighted geometric mean. Latency enters inverted because lower is better. The server recomputes the score from the submitted metrics; the result is rounded to one decimal. Score version: mem-1.

Reference and weights

MetricReferenceWeight
Read100 GB/s0.25
Write100 GB/s0.15
Copy100 GB/s0.20
Random20 GB/s0.15
Sustain100 GB/s0.20
Latency (inverted)800 ns0.05

Reading the scale

Throughput vs reference0.1x0.5x1x2x4x10x30x
MEM Score0.034.950.065.180.1100.0123.9

Score and percentile

The score says how much memory throughput a machine delivers. It is not a verdict on the machine: that is the job of the class percentile, and the reason every board ranks inside a device class.

AI decode ceiling

Single-stream token generation reads every weight once per token, so tokens per second cannot exceed read throughput divided by the model's size in bytes. The result screen shows this upper bound; real runtimes land below it.

Read throughput8B at 4-bit (4 GB)70B at 4-bit (35 GB)
100 GB/s25 tokens/s2.9 tokens/s
430 GB/s108 tokens/s12.3 tokens/s
1000 GB/s250 tokens/s28.6 tokens/s