Expert routing animation
++ Every layer of this model has 128 small expert networks, and a tiny router picks just 8 of + them for each token. The other 120 sit still. That is the whole trick of a mixture of + experts: you pay for 235B parameters in memory, but only about 22B of arithmetic per token. +
+ ++ GPU {gpu} + experts {gpu * PER_GPU}-{gpu * PER_GPU + PER_GPU - 1} +
+ +2 of 32 firing
+Memory fit animation
++ Every number here came out of a real run. All three bars are drawn to the same scale, and + the dashed line is the 85.51 GiB that vLLM may use on one card at + --gpu-memory-utilization 0.90. A bar reaching past that line means the model does not fit. + Watch it shrink as GPUs are added, and note that it takes 4 before the bar finally lands to + the left of the line. +
+ +Three ways to split animation
++ These are not competing products, they are three different cuts through the same pile of + weights, and you can combine them. Each box below is one GPU. Watch which parts light up, + because that tells you which GPUs are doing work at the same moment. +
+ +{mode.name}
+{mode.flag}
+
+
+
+ + What it does + {mode.plain} +
++ How much it talks + {mode.talks} +
++ When it wins + {mode.good} +
+Tensor parallelism animation
++ This is the part people usually get wrong, so it is worth being precise. The weights get + divided, and the thing flowing through them does not. Every GPU starts each layer holding + an identical copy of the token, does a quarter of the arithmetic on its own slice of the + weights, and ends up with a quarter of an answer. Then they add their quarters together. +
+ ++ GPU {gpu} + heads {gpu * 16}-{gpu * 16 + 15} +
+ +
+ 16 of 64 heads
+
+ 1 of 4 kv heads
+
+ each square above = 4 heads
+
a quarter answer
+