| /external/llvm/docs/Frontend/ |
| PerformanceTips.rst | 164 with the duplicated ones, this can provide a noticeable throughput
|
| /external/llvm/include/llvm/Target/ |
| TargetSchedule.td | 129 // determines the throughput of instructions that require the resource.
|
| /external/llvm/lib/Analysis/ |
| CostModel.cpp | 14 // the throughput of the machine assuming that all loads hit the cache, all
|
| /external/llvm/lib/Target/AArch64/ |
| AArch64SchedM1.td | 92 // TODO: 64-bit multiplication has a throughput of 1/2.
|
| AArch64TargetTransformInfo.cpp | 424 // extra micro-ops can significantly decrease throughput.
|
| /external/llvm/lib/Target/X86/ |
| X86ScheduleAtom.td | 543 // On the Atom, the throughput for taken branches is 2 cycles. For small
|
| /external/mesa3d/src/gallium/auxiliary/draw/ |
| draw_gs.c | 67 * throughput is optimized for the given vector instruction set.
|
| /external/mesa3d/src/gallium/drivers/nouveau/codegen/ |
| nv50_ir_target_nv50.cpp | 525 // These are "inverse" throughput values, i.e. the number of cycles required
|
| nv50_ir_target_nvc0.cpp | 555 // These are "inverse" throughput values, i.e. the number of cycles required
|
| /external/netcat/ |
| netcat.c | 1024 { "throughput", IPTOS_THROUGHPUT },
|
| /external/skia/site/dev/contrib/ |
| simd.md | 121 These details will inevitably change over time. The important takeaway here is, to really work at peak throughput in SIMD fixed point, you need to work with the idiom of the instruction set, and `SkPx` is a design that can present a consistent interface to abstract away backend details for you.
|
| /external/skqp/site/dev/contrib/ |
| simd.md | 121 These details will inevitably change over time. The important takeaway here is, to really work at peak throughput in SIMD fixed point, you need to work with the idiom of the instruction set, and `SkPx` is a design that can present a consistent interface to abstract away backend details for you.
|
| /external/tensorflow/tensorflow/core/kernels/batching_util/ |
| shared_batch_scheduler.h | 90 // For bulk processing jobs and throughput-oriented benchmarks, you may want to
|
| /external/tensorflow/tensorflow/docs_src/performance/ |
| benchmarks.md | 301 above due to our EFS setup not providing enough throughput.
|
| performance_models.md | 210 There are some techniques to improve throughput:
|
| /external/webrtc/webrtc/base/ |
| httpbase.cc | 504 // ideal throughput. The number of reads is limited to prevent starving
|
| /external/webrtc/webrtc/modules/remote_bitrate_estimator/test/ |
| bwe_test_framework.cc | 613 // This plots the max possible throughput of the trace-based delivery filter,
|
| /external/wpa_supplicant_8/src/p2p/ |
| p2p_group.c | 341 wpabuf_put_data(buf, dev_info + 4, 2); /* WFD Device Max Throughput */
|
| /hardware/nxp/nfc/halimpl/ |
| libnfc-brcm.conf | 10 # Change this setting to control how often USERIAL log the performance (throughput)
|
| /libcore/ojluni/src/main/java/java/util/concurrent/locks/ |
| ReentrantLock.java | 60 * may display lower overall throughput (i.e., are slower; often much
|
| /packages/apps/TV/tuner/src/com/android/tv/tuner/exoplayer/buffer/ |
| BufferManager.java | 45 * <p>The buffer manager can be disabled, while running, if the write throughput to the associated
|
| /prebuilts/clang/host/darwin-x86/clang-4053586/prebuilt_include/llvm/include/llvm/Target/ |
| TargetSchedule.td | 129 // determines the throughput of instructions that require the resource.
|
| /prebuilts/clang/host/darwin-x86/clang-4393122/include/llvm/Target/ |
| TargetSchedule.td | 129 // determines the throughput of instructions that require the resource.
|
| /prebuilts/clang/host/darwin-x86/clang-4479392/include/llvm/Target/ |
| TargetSchedule.td | 129 // determines the throughput of instructions that require the resource.
|
| /prebuilts/clang/host/darwin-x86/clang-4579689/include/llvm/Target/ |
| TargetSchedule.td | 129 // determines the throughput of instructions that require the resource.
|
