Statements (71)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:computer
|
| gptkbp:acceleration |
deep learning tasks
|
| gptkbp:are |
scalable
energy efficient used in research used in scientific computing used in production used in climate modeling used in natural language processing used in reinforcement learning available in various configurations highly parallel used in fraud detection integrated with Google services used in computer vision part of AI ethics discussions used in virtual assistants available for on-premises use cost-effective for large models designed for matrix operations part of AI infrastructure part of AI research community part of Google Cloud AI part of machine learning lifecycle used in anomaly detection used in autonomous driving research used in gaming AI used in model training used in supply chain optimization |
| gptkbp:available_in |
gptkb:cloud_services
|
| gptkbp:can_be_configured_for |
gptkb:Tensor_Flow
|
| gptkbp:can_be_used_for |
language translation
data preprocessing image recognition predictive analytics speech recognition text analysis customer segmentation recommendation systems video analysis time series forecasting image segmentation social media analysis network security analysis robotics applications model inference |
| gptkbp:can_be_used_in |
data centers
financial modeling healthcare applications |
| gptkbp:competes_in |
GPUs
|
| gptkbp:designed_by |
gptkb:Google
|
| gptkbp:enables |
faster training
|
| gptkbp:first_introduced |
gptkb:2016
|
| gptkbp:has_achieved |
high performance per watt
|
| gptkbp:have |
multiple versions
|
| https://www.w3.org/2000/01/rdf-schema#label |
TPUs
|
| gptkbp:include |
gptkb:TPU_v4
gptkb:TPU_v3 gptkb:TPU_v2 |
| gptkbp:is_integrated_with |
gptkb:Kubernetes
|
| gptkbp:is_optimized_for |
gptkb:Tensor_Flow
|
| gptkbp:provides |
high performance
|
| gptkbp:reduces |
training time
|
| gptkbp:support |
high throughput
multi-tenant environments multiple workloads large-scale models |
| gptkbp:type |
application-specific integrated circuit (ASIC)
|
| gptkbp:used_for |
gptkb:machine_learning
|
| gptkbp:bfsParent |
gptkb:Transformers
|
| gptkbp:bfsLayer |
4
|