{ "op": "ai.onnx.Mod", "tunableSpace": { "WORKGROUP_SIZE": [64, 128, 256] }, "cases": [ { "name": "8m_u32_scalar_div", "preset": "smoke", "inputs": { "a": { "dtype": "uint32", "shape": [8388608] }, "b": { "dtype": "uint32", "shape": [1], "data": { "kind": "constant", "value": 17 } } }, "outputs": { "c": { "dtype": "uint32", "shape": [8388608] } }, "bench": { "primary": true, "metrics": [{ "type": "bandwidth", "value": "8388608 * 4 * 2" }] } }, { "name": "1m_u32", "inputs": { "a": { "dtype": "uint32", "shape": [1048576] }, "b": { "dtype": "uint32", "shape": [1], "data": { "kind": "constant", "value": 17 } } }, "outputs": { "c": { "dtype": "uint32", "shape": [1048576] } } }, { "name": "1m_f32_fmod_same_shape", "attrs": { "fmod": 1 }, "inputs": { "a": { "dtype": "float32", "shape": [1048576] }, "b": { "dtype": "float32", "shape": [1048576], "data": { "kind": "constant", "value": 3.7 } } }, "outputs": { "c": { "dtype": "float32", "shape": [1048576] } } }, { "name": "f32_broadcast_per_channel_fallback", "preset": "smoke", "vars": { "dtype": "float32" }, "attrs": { "fmod": 1 }, "inputs": { "a": { "dtype": "float32", "shape": [2048, 2049], "dist": "normal", "seed": 711, "scale": 2 }, "b": { "dtype": "float32", "shape": [2049], "data": { "kind": "constant", "value": 3.0 } } }, "outputs": { "c": { "dtype": "float32", "shape": [2048, 2049] } }, "bench": { "metrics": [{ "type": "bandwidth", "value": "2048 * 2049 * 2 * dtypeBytes(args.dtype)" }] } }, { "name": "1m_f16_fmod_same_shape", "preset": "smoke", "vars": { "dtype": "float16", "count": 1048576 }, "attrs": { "fmod": 1 }, "inputs": { "a": { "dtype": "float16", "shape": [1048576], "dist": "normal", "seed": 712, "scale": 8 }, "b": { "dtype": "float16", "shape": [1048576], "data": { "kind": "constant", "value": 3.5 } } }, "outputs": { "c": { "dtype": "float16", "shape": [1048576] } }, "bench": { "metrics": [{ "type": "bandwidth", "value": "args.count * dtypeBytes(args.dtype) * 3" }] } }, { "name": "1m_i32_floor_mod_same_shape", "preset": "smoke", "vars": { "dtype": "int32", "count": 1048576 }, "inputs": { "a": { "dtype": "int32", "shape": [1048576], "dist": "normal", "seed": 713, "scale": 1000 }, "b": { "dtype": "int32", "shape": [1048576], "data": { "kind": "constant", "value": 7 } } }, "outputs": { "c": { "dtype": "int32", "shape": [1048576] } }, "bench": { "metrics": [{ "type": "bandwidth", "value": "args.count * dtypeBytes(args.dtype) * 3" }] } }, { "name": "i32_floor_mod_broadcast_non_vec4_shape", "preset": "stress", "attrs": { "fmod": 0 }, "inputs": { "a": { "dtype": "int32", "shape": [2049, 2051], "dist": "normal", "seed": 801, "scale": 1000 }, "b": { "dtype": "int32", "shape": [2051], "data": { "kind": "constant", "value": 7 } } }, "outputs": { "c": { "dtype": "int32", "shape": [2049, 2051] } }, "bench": { "metrics": [{ "type": "bandwidth", "value": "2049 * 2051 * 4 * 2" }] } } ] }