Update benchmark data

Author: adenzler-nvidia

results/adenzler-asv-runner/fad7055f-virtualenv-py3.13.json

@@ -0,0 +1 @@
+{"commit_hash": "fad7055f6fe8f4e1582bba564e42192517c901d2", "env_name": "virtualenv-py3.13", "date": 1761847185000, "params": {"arch": "x86_64", "cpu": "AMD EPYC-Milan Processor", "machine": "adenzler-asv-runner", "num_cpu": "32", "os": "Linux 5.15.0-131-generic", "ram": "65836440", "python": "3.13"}, "python": "3.13", "requirements": {}, "env_vars": {}, "result_columns": ["result", "params", "version", "started_at", "duration", "stats_ci_99_a", "stats_ci_99_b", "stats_q_25", "stats_q_75", "stats_number", "stats_repeat", "samples", "profile"], "results": {"compilation.bench_example_load.SlowExampleBasicUrdf.time_load": [[24.285476382588968], [], "5212ba785a2d013b8fc4b4fd1f0281ba23f65a10036f337e6abebdcd1784a35a", 1761847880825, 103.23, [19.738], [31.261], [24.104], [25.681], [1], [4]], "compilation.bench_example_load.SlowExampleClothFranka.time_load": [[36.560582337668166], [], "9a1d510bf84d0d2d4d3c68dc0b43f672866854e51df5b70071be0cc9a9aa37ee", 1761847929545, 152.21, [31.296], [44.292], [36.044], [38.31], [1], [4]], "compilation.bench_example_load.SlowExampleClothTwist.time_load": [[19.330012077232823], [], "2944bcd1d41a1d0755a46fab1ccbb2b38418c589a0ebb982702ba9be5bd32471", 1761848001947, 78.268, [18.833], [19.761], [19.185], [19.442], [1], [4]], "compilation.bench_example_load.SlowExampleRobotAnymal.time_load": [[33.51826209318824], [], "30b4e830fd859608ff412e9052e49492eb804327e44f3cbedf4183876c171021", 1761848041004, 137.95, [30.466], [37.955], [33.347], [34.381], [1], [4]], "compilation.bench_example_load.SlowExampleRobotCartpole.time_load": [[35.17323106224649], [], "71e2c9a9b0957497094f25a0b4dff3546059e17c249880ccff2c5a3dc09debf7", 1761848108107, 143.97, [32.664], [38.745], [34.797], [36.081], [1], [4]], "compilation.bench_example_load.SlowExampleRobotHumanoid.time_load": [[40.36822128389031], [], "3d838883c58af639755f7dcfa6d7b6f603f4faefa35577421e70a41be5877518", 1761848179017, 161.87, [37.359], [43.021], [39.447], [41.111], [1], [4]], "simulation.bench_anymal.FastExampleAnymalPretrained.time_simulate": [[0.1705406007822603], [], "29d871d361f5e50619ca9ee4af8ee5f22def523c03a5aa19b0365c23ecb5dd28", 1761848585432, 49.123, [0.1675], [0.17302], [0.16911], [0.17166], [1], [6]], "simulation.bench_cloth.FastExampleClothManipulation.time_simulate": [[0.9730092138051987], [], "b3ceadbc55945ff4c6130a4fbcec6e455edc4e3085dd62f0e02103dbd33101ad", 1761848609431, 77.788, [0.96967], [0.976], [0.97101], [0.97457], [1], [6]], "simulation.bench_cloth.FastExampleClothTwist.time_simulate": [[0.7845400075893849], [], "01dfa4b62b6e71ea72ece4c0c2371d26b1ba72a8ef8c94a900d8e40be3907293", 1761848644400, 17.325, [0.78254], [0.78572], [0.78358], [0.78474], [1], [10]], "simulation.bench_mujoco.FastAllegro.time_simulate": [[0.8867516175378114], [], "13821b334a19fbc7b1a7333e1ef68506404d4e1fa9c7031bc46336537cea9c68", 1761848653061, 102.81, [0.88359], [0.89024], [0.88618], [0.88749], [1], [4]], "simulation.bench_mujoco.FastCartpole.time_simulate": [[0.28532431088387966], [], "13821b334a19fbc7b1a7333e1ef68506404d4e1fa9c7031bc46336537cea9c68", 1761848702486, 35.259, [0.28429], [0.28612], [0.28452], [0.28594], [1], [16]], "simulation.bench_mujoco.FastG1.time_simulate": [[0.3637667130678892], [], "13821b334a19fbc7b1a7333e1ef68506404d4e1fa9c7031bc46336537cea9c68", 1761848720221, 76.158, [0.36277], [0.36483], [0.36345], [0.36412], [1], [4]], "simulation.bench_mujoco.FastHumanoid.time_simulate": [[0.2745674070902169], [], "13821b334a19fbc7b1a7333e1ef68506404d4e1fa9c7031bc46336537cea9c68", 1761848758284, 30.124, [0.27376], [0.275], [0.27432], [0.27482], [1], [16]], "simulation.bench_mujoco.FastKitchenG1.time_simulate": [[0.23215825273655355], [], "13821b334a19fbc7b1a7333e1ef68506404d4e1fa9c7031bc46336537cea9c68", 1761848773467, 14.744, [0.23111], [0.23311], [0.23188], [0.23239], [1], [4]], "simulation.bench_quadruped_xpbd.FastExampleQuadrupedXPBD.time_simulate": [[0.7366542126983404], [], "8527e31e794cfba679aafd4d750232df62492b0b836ab15cdad57ab6c92db2dc", 1761849102494, 47.676, [0.73569], [0.73803], [0.73615], [0.73754], [1], [20]], "simulation.bench_selection.FastExampleSelectionCartpoleMuJoCo.time_simulate": [[0.3482107550371438], [], "adc93f1285ec796ed3088445a10636624dd2a89b99b5c90bcbb42add8c9087df", 1761849126157, 40.493, [0.34746], [0.34954], [0.34759], [0.34925], [1], [20]], "setup.bench_model.FastInitializeSolver.time_initialize_solver": [[0.02035575918853283, 0.09279550984501839, 0.020447688177227974], [["'humanoid'", "'g1'", "'cartpole'"], ["256"]], "b2d1b17d4e4a010a2441d0d10b9b209c4884c57db6d16b718c8c9b5127f838a1", 1761848259519, 23.815, [0.016151, 0.069332, 0.018372], [0.024972, 0.11911, 0.023045], [0.020033, 0.091185, 0.020403], [0.020987, 0.096547, 0.020884], [1, 1, 1], [3, 3, 3]], "setup.bench_model.FastInitializeViewerGL.time_initialize_renderer": [[0.2675033863633871], [["'g1'"], ["256"]], "d9add5c09316bc748c3e612cc16cbd7f095ec18b01a1b27b1ea8e239cab0b94a", 1761848283338, 10.324, [0.22029], [0.32556], [0.26608], [0.27705], [1], [3]], "setup.bench_model.KpiInitializeModel.time_initialize_model": [[3.0937541387975216, 10.428515096195042, 0.4841427751816809], [["'humanoid'", "'g1'", "'cartpole'"], ["8192"]], "53272b20ee0eacff6fdbc0e21ce8b18aab7e4201a123fc59bc4606b6de60b08d", 1761848293666, 64.117, [3.0046, 9.8869, 0.46675], [3.1835, 10.994, 0.49767], [3.0842, 10.377, 0.48114], [3.1037, 10.498, 0.48426], [1, 1, 1], [3, 3, 3]], "setup.bench_model.KpiInitializeSolver.time_initialize_solver": [[0.12102447636425495, 0.40550001012161374, 0.3072626260109246, 0.05193715961650014], [["'humanoid'", "'g1'", "'cartpole'", "'ant'"], ["8192"]], "55e5ec342e9efdcb797c22fdcf5d25b898c00a14c8edbbaf90a3858b53683d50", 1761848357786, 123.44, [0.10969, 0.39957, 0.23134, 0.010716], [0.13285, 0.41079, 0.39034, 0.10261], [0.11995, 0.40466, 0.30135, 0.050693], [0.12247, 0.40587, 0.31855, 0.06027], [1, 1, 1, 1], [3, 3, 3, 3]], "setup.bench_model.KpiInitializeViewerGL.time_initialize_renderer": [[7.776575423311442], [["'g1'"], ["8192"]], "2d33b796bb69aafb93c3c88f9a12b13808e8a5cdb6df7562ded097818f80b51c", 1761848481228, 104.2, [7.1014], [8.3167], [7.6633], [7.7886], [1], [3]], "simulation.bench_mujoco.FastNewtonOverheadG1.track_simulate": [[4.910042608094523], [["256"]], "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558", 1761848780934, 8.0451], "simulation.bench_mujoco.FastNewtonOverheadHumanoid.track_simulate": [[9.289847390158236], [["256"]], "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558", 1761848788982, 5.2088], "simulation.bench_mujoco.KpiAllegro.track_simulate": [[0.001479922352397504], [["8192"]], "62a2fd707ff553505c8e26bae864c53867d55fc0442b3c6429ccc112ee6b21ad", 1761848794194, 86.981], "simulation.bench_mujoco.KpiCartpole.track_simulate": [[6.819512054789811e-05], [["8192"]], "62a2fd707ff553505c8e26bae864c53867d55fc0442b3c6429ccc112ee6b21ad", 1761848881178, 8.6541], "simulation.bench_mujoco.KpiG1.track_simulate": [[0.00275125793996267], [["8192"]], "62a2fd707ff553505c8e26bae864c53867d55fc0442b3c6429ccc112ee6b21ad", 1761848889835, 50.001], "simulation.bench_mujoco.KpiHumanoid.track_simulate": [[0.00094327828992391], [["8192"]], "62a2fd707ff553505c8e26bae864c53867d55fc0442b3c6429ccc112ee6b21ad", 1761848939838, 48.126], "simulation.bench_mujoco.KpiKitchenG1.track_simulate": [[0.016040524933487177], [["512"]], "62a2fd707ff553505c8e26bae864c53867d55fc0442b3c6429ccc112ee6b21ad", 1761848987966, 16.415], "simulation.bench_mujoco.KpiNewtonOverheadG1.track_simulate": [[1.2395545460635642], [["8192"]], "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558", 1761849004383, 49.806], "simulation.bench_mujoco.KpiNewtonOverheadHumanoid.track_simulate": [[1.1173206082895175], [["8192"]], "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558", 1761849054193, 48.298], "simulation.bench_viewer.FastViewerGL.time_rendering_frame": [[0.02343240799382329], [["'g1'"], ["256"]], "a7320c2bfd979f20a9efbe3c6bcb6cbe8daf027c03fbeea90667301702873854", 1761849146276, 9.6747, [0.022951], [0.023829], [0.023354], [0.023446], [1], [3]], "simulation.bench_viewer.KpiViewerGL.time_rendering_frame": [[0.03241401771083474], [["'g1'"], ["8192"]], "9705f69f1cdc6900bdaef46555562d3ad19312423237a52028f380687c7c359d", 1761849155955, 107.6, [0.0273], [0.036772], [0.031627], [0.032634], [1], [3]], "setup.bench_model.FastInitializeModel.peakmem_initialize_model_cpu": [[371101696, 666562560, 372604928], [["'humanoid'", "'g1'", "'cartpole'"], ["256"]], "a7473dc081fbb914ac4fa9f1e7e5be2821d5847ec89f7cb912c05555ece87d4d", 1761849268664, 12.109], "setup.bench_model.FastInitializeModel.time_initialize_model": [[0.08889023307710886, 1.0760770118795335, 0.028675859794020653], [["'humanoid'", "'g1'", "'cartpole'"], ["256"]], "c443a475fe79ddf05a471ff6dd41b6902a1dca7225884afad7ab732c240e7119", 1761849280775, 12.721, [0.080741, 1.0367, 0.0099168], [0.095859, 1.1231, 0.043211], [0.087643, 1.0744, 0.02542], [0.089253, 1.0835, 0.028764], [1, 1, 1], [3, 3, 3]]}, "durations": {"<setup_cache setup.bench_model:120>": 5.107653}, "version": 2}

results/benchmarks.json

@@ -398,6 +398,36 @@
         "version": "13821b334a19fbc7b1a7333e1ef68506404d4e1fa9c7031bc46336537cea9c68",
         "warmup_time": -1
     },
+    "simulation.bench_mujoco.FastNewtonOverheadG1.track_simulate": {
+        "code": "class _NewtonOverheadBenchmark:\n    @skip_benchmark_if(wp.get_cuda_device_count() == 0)\n    def track_simulate(self, num_worlds):\n        trace = {}\n        with EventTracer(enabled=True) as tracer:\n            for _iter in range(self.samples):\n                example = Example(\n                    stage_path=None,\n                    robot=self.robot,\n                    randomize=self.random_init,\n                    headless=True,\n                    actuation=\"random\",\n                    num_worlds=num_worlds,\n                    use_cuda_graph=True,\n                    builder=self.builder[num_worlds],\n                    ls_iteration=self.ls_iteration,\n                )\n    \n                for _ in range(self.num_frames):\n                    example.step()\n                    trace = tracer.add_trace(trace, tracer.trace())\n    \n        step_time = trace[\"step\"][0]\n        mujoco_warp_step_time = trace[\"step\"][1][\"mujoco_warp_step\"][0]\n        overhead = 100.0 * (step_time - mujoco_warp_step_time) / step_time\n        return overhead\n\n    def setup(self, num_worlds):\n        if not hasattr(self, \"builder\") or self.builder is None:\n            self.builder = {}\n        if num_worlds not in self.builder:\n            self.builder[num_worlds] = Example.create_model_builder(\n                self.robot, num_worlds, randomize=self.random_init, seed=123\n            )",
+        "name": "simulation.bench_mujoco.FastNewtonOverheadG1.track_simulate",
+        "param_names": [
+            "num_worlds"
+        ],
+        "params": [
+            [
+                "256"
+            ]
+        ],
+        "type": "track",
+        "unit": "%",
+        "version": "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558"
+    },
+    "simulation.bench_mujoco.FastNewtonOverheadHumanoid.track_simulate": {
+        "code": "class _NewtonOverheadBenchmark:\n    @skip_benchmark_if(wp.get_cuda_device_count() == 0)\n    def track_simulate(self, num_worlds):\n        trace = {}\n        with EventTracer(enabled=True) as tracer:\n            for _iter in range(self.samples):\n                example = Example(\n                    stage_path=None,\n                    robot=self.robot,\n                    randomize=self.random_init,\n                    headless=True,\n                    actuation=\"random\",\n                    num_worlds=num_worlds,\n                    use_cuda_graph=True,\n                    builder=self.builder[num_worlds],\n                    ls_iteration=self.ls_iteration,\n                )\n    \n                for _ in range(self.num_frames):\n                    example.step()\n                    trace = tracer.add_trace(trace, tracer.trace())\n    \n        step_time = trace[\"step\"][0]\n        mujoco_warp_step_time = trace[\"step\"][1][\"mujoco_warp_step\"][0]\n        overhead = 100.0 * (step_time - mujoco_warp_step_time) / step_time\n        return overhead\n\n    def setup(self, num_worlds):\n        if not hasattr(self, \"builder\") or self.builder is None:\n            self.builder = {}\n        if num_worlds not in self.builder:\n            self.builder[num_worlds] = Example.create_model_builder(\n                self.robot, num_worlds, randomize=self.random_init, seed=123\n            )",
+        "name": "simulation.bench_mujoco.FastNewtonOverheadHumanoid.track_simulate",
+        "param_names": [
+            "num_worlds"
+        ],
+        "params": [
+            [
+                "256"
+            ]
+        ],
+        "type": "track",
+        "unit": "%",
+        "version": "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558"
+    },
     "simulation.bench_mujoco.KpiAllegro.track_simulate": {
         "code": "class _KpiBenchmark:\n    @skip_benchmark_if(wp.get_cuda_device_count() == 0)\n    def track_simulate(self, num_worlds):\n        total_time = 0.0\n        for _iter in range(self.samples):\n            example = Example(\n                stage_path=None,\n                robot=self.robot,\n                randomize=self.random_init,\n                headless=True,\n                actuation=\"random\",\n                use_cuda_graph=True,\n                builder=self.builder[num_worlds],\n                ls_iteration=self.ls_iteration,\n                environment=self.environment,\n            )\n    \n            wp.synchronize_device()\n            for _ in range(self.num_frames):\n                example.step()\n            wp.synchronize_device()\n            total_time += example.benchmark_time\n    \n        return total_time * 1000 / (self.num_frames * example.sim_substeps * num_worlds * self.samples)\n\n    def setup(self, num_worlds):\n        if not hasattr(self, \"builder\") or self.builder is None:\n            self.builder = {}\n        if num_worlds not in self.builder:\n            self.builder[num_worlds] = Example.create_model_builder(\n                self.robot, num_worlds, randomize=self.random_init, seed=123\n            )",
         "name": "simulation.bench_mujoco.KpiAllegro.track_simulate",
@@ -475,6 +505,37 @@
         "unit": "ms/world-step",
         "version": "62a2fd707ff553505c8e26bae864c53867d55fc0442b3c6429ccc112ee6b21ad"
     },
+    "simulation.bench_mujoco.KpiNewtonOverheadG1.track_simulate": {
+        "code": "class _NewtonOverheadBenchmark:\n    @skip_benchmark_if(wp.get_cuda_device_count() == 0)\n    def track_simulate(self, num_worlds):\n        trace = {}\n        with EventTracer(enabled=True) as tracer:\n            for _iter in range(self.samples):\n                example = Example(\n                    stage_path=None,\n                    robot=self.robot,\n                    randomize=self.random_init,\n                    headless=True,\n                    actuation=\"random\",\n                    num_worlds=num_worlds,\n                    use_cuda_graph=True,\n                    builder=self.builder[num_worlds],\n                    ls_iteration=self.ls_iteration,\n                )\n    \n                for _ in range(self.num_frames):\n                    example.step()\n                    trace = tracer.add_trace(trace, tracer.trace())\n    \n        step_time = trace[\"step\"][0]\n        mujoco_warp_step_time = trace[\"step\"][1][\"mujoco_warp_step\"][0]\n        overhead = 100.0 * (step_time - mujoco_warp_step_time) / step_time\n        return overhead\n\n    def setup(self, num_worlds):\n        if not hasattr(self, \"builder\") or self.builder is None:\n            self.builder = {}\n        if num_worlds not in self.builder:\n            self.builder[num_worlds] = Example.create_model_builder(\n                self.robot, num_worlds, randomize=self.random_init, seed=123\n            )",
+        "name": "simulation.bench_mujoco.KpiNewtonOverheadG1.track_simulate",
+        "param_names": [
+            "num_worlds"
+        ],
+        "params": [
+            [
+                "8192"
+            ]
+        ],
+        "timeout": 900,
+        "type": "track",
+        "unit": "%",
+        "version": "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558"
+    },
+    "simulation.bench_mujoco.KpiNewtonOverheadHumanoid.track_simulate": {
+        "code": "class _NewtonOverheadBenchmark:\n    @skip_benchmark_if(wp.get_cuda_device_count() == 0)\n    def track_simulate(self, num_worlds):\n        trace = {}\n        with EventTracer(enabled=True) as tracer:\n            for _iter in range(self.samples):\n                example = Example(\n                    stage_path=None,\n                    robot=self.robot,\n                    randomize=self.random_init,\n                    headless=True,\n                    actuation=\"random\",\n                    num_worlds=num_worlds,\n                    use_cuda_graph=True,\n                    builder=self.builder[num_worlds],\n                    ls_iteration=self.ls_iteration,\n                )\n    \n                for _ in range(self.num_frames):\n                    example.step()\n                    trace = tracer.add_trace(trace, tracer.trace())\n    \n        step_time = trace[\"step\"][0]\n        mujoco_warp_step_time = trace[\"step\"][1][\"mujoco_warp_step\"][0]\n        overhead = 100.0 * (step_time - mujoco_warp_step_time) / step_time\n        return overhead\n\n    def setup(self, num_worlds):\n        if not hasattr(self, \"builder\") or self.builder is None:\n            self.builder = {}\n        if num_worlds not in self.builder:\n            self.builder[num_worlds] = Example.create_model_builder(\n                self.robot, num_worlds, randomize=self.random_init, seed=123\n            )",
+        "name": "simulation.bench_mujoco.KpiNewtonOverheadHumanoid.track_simulate",
+        "param_names": [
+            "num_worlds"
+        ],
+        "params": [
+            [
+                "8192"
+            ]
+        ],
+        "type": "track",
+        "unit": "%",
+        "version": "f190c2aa963b9499b253447d6300eeff96be1d81b89c9e388d8ed69e4ae3d558"
+    },
     "simulation.bench_quadruped_xpbd.FastExampleQuadrupedXPBD.time_simulate": {
         "code": "class FastExampleQuadrupedXPBD:\n    @skip_benchmark_if(wp.get_cuda_device_count() == 0)\n    def time_simulate(self):\n        for _ in range(self.num_frames):\n            self.example.step()\n        wp.synchronize_device()\n\n    def setup(self):\n        self.num_frames = 1000\n        self.example = Example(newton.viewer.ViewerNull(num_frames=self.num_frames), 200)",
         "min_run_count": 2,