x86 - How to compile Tensorflow with SSE4.2 and AVX instructions? -


this message received running script check if tensorflow working:

i tensorflow/stream_executor/dso_loader.cc:125] opened cuda library libcublas.so.8.0 locally tensorflow/stream_executor/dso_loader.cc:125] opened cuda library libcudnn.so.5 locally tensorflow/stream_executor/dso_loader.cc:125] opened cuda library libcufft.so.8.0 locally tensorflow/stream_executor/dso_loader.cc:125] opened cuda library libcuda.so.1 locally tensorflow/stream_executor/dso_loader.cc:125] opened cuda library libcurand.so.8.0 locally w tensorflow/core/platform/cpu_feature_guard.cc:95] tensorflow library wasn't compiled use sse4.2 instructions, these available on machine , speed cpu computations. w tensorflow/core/platform/cpu_feature_guard.cc:95] tensorflow library wasn't compiled use avx instructions, these available on machine , speed cpu computations. tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:910] successful numa node read sysfs had negative value (-1), there must @ least 1 numa node, returning numa node 0

i noticed has mentioned sse4.2 , avx,

1) sse4.2 , avx?

2) how these sse4.2 , avx improve cpu computations tensorflow tasks.

3) how make tensorflow compile using 2 libraries?

i ran same problem, seems yaroslav bulatov's suggestion doesn't cover sse4.2 support, adding --copt=-msse4.2 suffice. in end, built with

bazel build -c opt --copt=-mavx --copt=-mavx2 --copt=-mfma --copt=-mfpmath=both --copt=-msse4.2 --config=cuda -k //tensorflow/tools/pip_package:build_pip_package

without getting warning or errors.

probably best choice system is:

bazel build -c opt --copt=-march=native --copt=-mfpmath=both --config=cuda -k //tensorflow/tools/pip_package:build_pip_package

-mfpmath=both works gcc, not clang. i'm not sure default -o2 or -o3 setting is. gcc -o3 enables auto-vectorization, won't help, , make slower.

what does: --copt bazel build passes option directly gcc compiling c , c++ files (but not linking, need different option cross-file link-time-optimization)

x86-64 gcc defaults using sse2 or older simd instructions, can run binaries on any x86-64 system. (see https://gcc.gnu.org/onlinedocs/gcc/x86-options.html). that's not want. want make binary takes advantage of instructions cpu can run, because you're running binary on system built it.

-march=native enables options cpu supports, makes -mavx512f -mavx2 -mavx -mfma -msse4.2 redundant. (also, -mavx2 enables -mavx , -msse4.2, yaroslav's command should have been fine). if you're using cpu doesn't support 1 of these options (like fma), using -mfma make binary faults illegal instructions.

tensorflow's ./configure defaults enabling -march=native, using should avoid needing specify compiler options manually.

-march=native enables -mtune=native, it optimizes cpu things sequence of avx instructions best unaligned loads.

this applies gcc, clang, or icc.


-

Comments

Post a Comment

Popular posts from this blog

python - Error importing VideoFileClip from moviepy : AttributeError: 'PermissionError' object has no attribute 'message' -

i'm using jupyter notebook. have tried anaconda console well. tried importing both ways shown below from moviepy.editor import videofileclip moviepy.video.io.videofileclip import videofileclip both of them gave me same error. full trace below attributeerror traceback (most recent call last) <ipython-input-10-9afa9d6e87c4> in <module>() 6 import glob 7 import math ----> 8 moviepy.editor import videofileclip 9 moviepy.video.io.videofileclip import videofileclip c:\program files\anaconda3\lib\site-packages\moviepy\editor.py in <module>() 20 # clips 21 ---> 22 .video.io.videofileclip import videofileclip 23 .video.io.imagesequenceclip import imagesequenceclip 24 .video.io.downloader import download_webfile c:\program files\anaconda3\lib\site-packages\moviepy\video\io\videofileclip.py in <module>() 1 import os 2 ----> 3 moviepy.video.videoclip import videoclip ...
Read more

java - is not an enclosing class / new Intent Cannot Resolve Constructor -

thanks checking out question! i have relativly ismple loop add onclicklistener each imageview add layout, when trying add new intent it, gives me 1 of following errors: string[] imageurls = imageurlsstring.split("/"); (int = 0; < imageurls.length; i++){ imageview image = new imageview(this); image.setlayoutparams(new android.view.viewgroup.layoutparams(getpx(182),getpx(256))); image.setpadding(getpx(3),getpx(3),getpx(3),getpx(3)); final string imageurl = ".../images/" + imageurls[i]; picasso.with(this).load(imageurl).into(image); image.setonclicklistener(new view.onclicklistener() { @override public void onclick(view view) { intent intent; intent = new intent(this, imageactivity.class); intent.putextra("url", imageurl); startactivity(intent); } }); linearlayout.addview(image)...
Read more

qt - QML MouseArea onWheel event not working properly when inside QML Scrollview -

i have mousearea inside scrollview inside rectangle . implemented zoom feature zooms in/out when pressing ctrl , scrolling mouse wheel. however, zooms in when scrollview way @ top, , zooms out when scrollview way @ bottom. there additional logic handle external drag , drop of files. issue should able replicated long text inside textedit big enough scrollview . apparently bug before, can't work properly. tried solution in following link: qtquick2: handle onwheel event inside of scrollview rectangle { id: palgenrectangle layout.minimumwidth: 50 property string display //width:800 color: "white" scrollview { id: palgentextscrollview anchors.fill: parent mousearea { id: mousearea anchors.fill: parent onwheel: { if (wheel.modifiers & qt.controlmodifier){ if (wheel.angledelta.y > 0) { mainte...
Read more