|
gptkbp:instance_of
|
gptkb:computer
|
|
gptkbp:bfsLayer
|
5
|
|
gptkbp:bfsParent
|
gptkb:Google_Tensor_G2
|
|
gptkbp:architectural_style
|
application-specific integrated circuit (ASIC)
|
|
gptkbp:designed_by
|
gptkb:Job_Search_Engine
|
|
gptkbp:developed_by
|
gptkb:Google_Brain_team
|
|
gptkbp:features
|
low power consumption
matrix multiplication
high efficiency
|
|
gptkbp:first_generation
|
gptkb:TPU_v4
gptkb:TPU_v3
gptkb:TPU_v1
gptkb:TPU_v2
|
|
gptkbp:has
|
multiple cores
|
|
gptkbp:has_impact_on
|
AI development
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Tensor Processing Unit (TPU)
|
|
gptkbp:introduced
|
gptkb:2016
|
|
gptkbp:is_available_in
|
cloud-based solutions
various configurations
on-premises solutions
|
|
gptkbp:is_available_on
|
gptkb:Cloud_Computing_Service
|
|
gptkbp:is_compared_to
|
gptkb:CEO
|
|
gptkbp:is_integrated_with
|
other Google services
|
|
gptkbp:is_known_for
|
gptkb:military_unit
energy efficiency
cost-effectiveness
|
|
gptkbp:is_optimized_for
|
gptkb:microprocessor
deep learning
|
|
gptkbp:is_part_of
|
gptkb:Google's_AI_ecosystem
AI infrastructure
Google's AI research initiatives
|
|
gptkbp:is_scalable
|
gptkb:battle
|
|
gptkbp:is_supported_by
|
gptkb:Keras
gptkb:Py_Torch
gptkb:Tensor_Flow_Extended_(TFX)
gptkb:Tensor_Flow_Lite
|
|
gptkbp:is_used_by
|
gptkb:physicist
gptkb:software
companies
|
|
gptkbp:is_used_for
|
gptkb:software_framework
computer vision
image recognition
natural language processing
speech recognition
recommendation systems
|
|
gptkbp:is_used_in
|
gptkb:Research_Institute
data centers
production environments
|
|
gptkbp:performance
|
high throughput
|
|
gptkbp:provides
|
accelerated training
accelerated inference
|
|
gptkbp:ram
|
high bandwidth memory (HBM)
|
|
gptkbp:supports
|
gptkb:Graphics_Processing_Unit
large-scale machine learning
|
|
gptkbp:uses
|
custom hardware
|