As outlined above, the biggest breakthroughs for artificial intelligence research in recent years have been in the field of machine learning, especially in the field of deep learning.
This was achieved in part by the easy accessibility of data, but even more so by an explosion in parallel computing power, during this time the use of graphics processing units (GPUs) to train machine learning systems became more common.
These clusters are much more powerful only for training machine learning models.
not only do they offer mler, but they are now widely available as cloud services over the internet. Over time, major technology firms such as Google, Microsoft and Tesla have moved to using custom chips adapted to both employee and, more recently, educational, machine learning models.
An example of one of these special chips is Google's Tensor Processing Unit (TPU), which accelerates the latest version, as well as the speed at which useful machine learning models created using Google's TensorFlow software library extract information from data.
These chips are used in services such as DeepMind and Google Brain models and models that support Google translation, as well as image recognition in Google Photos and allowing the public to create machine learning models using Google's TensorFlow Research Cloud. The third generation of these chips was unveiled at Google's I/O conference in May 2018 and has since been packaged into machine learning power centers called pods that can perform more than a hundred thousand trillion floating-point operations (100 petaflops) per second. These ongoing TPU upgrades allowed Google to improve its services based on machine learning models; for example, Google halved the time it took to train the models used in translation.
댓글