The more cores you can use, the better — especially with big data. But the easier a big data framework is to work with, the harder it is for the resulting pipelines, such as TensorFlow plus Apache ...