# FLOPs

FLOP estimation from graph structure is standard, well-established technique applied here to Tensormorph's confirmed graph representation - it is not itself a confirmed, named Tensormorph feature. Treat this page as a plausible capability built on confirmed foundations, not documented UI.

A FLOP (floating-point operation) count estimates the raw computational cost of an operation from its shapes alone, without needing to actually run it - useful for comparing the cost of two architectures, or two [morph](/docs/core-concepts/model-morphing) candidates, before committing to running either.

## Estimating from graph structure

Because an imported model is read into a graph of typed operations (see [Model, Architecture, Layer and Module](/docs/core-concepts/model-architecture-layer-module) and, for graph-native formats, [ONNX](/docs/working-with-models/onnx)), a FLOP estimate for an operation follows directly from its type and its input/output shapes - a `MatMul` of an `[M,K]` matrix against a `[K,N]` matrix costs on the order of `2*M*K*N` FLOPs, for instance. Because [shape inference](/docs/working-with-models/onnx#opset-version-and-shape-inference) can populate every edge's shape before a model ever runs, a whole-graph FLOP estimate is available immediately on import, the same way shape labels are.

## Where this is useful

-   **Comparing morph candidates** - two candidates that are numerically similar can still have very different inference cost if one changed a layer's dimensions; a FLOP comparison surfaces that immediately, without needing a runtime benchmark.
-   **Spotting disproportionate cost** - a [Layer Profile](/docs/editors/signal-and-distribution-editor) of estimated FLOPs per layer can show that one unremarkable-looking layer is responsible for a large share of total compute cost.
-   **Sanity-checking quantization or pruning gains** - a FLOP estimate before and after a compression pass gives an expected speedup figure to check actual measured performance against.

## FLOPs are an estimate, not a measurement

A FLOP count describes theoretical arithmetic cost, not actual wall-clock time - real performance depends on the backend, memory bandwidth, and how well an operation's shapes fit the hardware's execution model. See [GPU Kernels](/docs/profiling/gpu-kernels) and [Performance Baselines](/docs/profiling/performance-baselines) for measured, rather than estimated, performance.

## Related resources

-   [Model Morphing](/docs/core-concepts/model-morphing) - comparing compute cost between morph candidates.
-   [Performance Baselines](/docs/profiling/performance-baselines) - comparing measured, rather than estimated, performance.
-   [ONNX](/docs/working-with-models/onnx) - the graph structure a FLOP estimate is computed over for graph-native formats.