Tensor Processing Unit
Tensor Processing Unit (TPU) is an AI accelerator application-specific integrated circuit (ASIC) developed by Google for neural network machine learning, using Google's own TensorFlow software. Google began using TPUs internally in 2015, and in 2018 made them available for third party use, both as part of its cloud infrastructure and by offering a smaller version of the chip for sale.
- Caption
- enTensor Processing Unit 3.0
- Comment
- enTensor Processing Unit (TPU) is an AI accelerator application-specific integrated circuit (ASIC) developed by Google for neural network machine learning, using Google's own TensorFlow software. Google began using TPUs internally in 2015, and in 2018 made them available for third party use, both as part of its cloud infrastructure and by offering a smaller version of the chip for sale.
- Depiction
- Designer
- Has abstract
- enTensor Processing Unit (TPU) is an AI accelerator application-specific integrated circuit (ASIC) developed by Google for neural network machine learning, using Google's own TensorFlow software. Google began using TPUs internally in 2015, and in 2018 made them available for third party use, both as part of its cloud infrastructure and by offering a smaller version of the chip for sale.
- Introduced
- enMay 2016
- Is primary topic of
- Tensor Processing Unit
- Label
- enTensor Processing Unit
- Link from a Wikipage to an external page
- cloud.google.com/images/products/tpu/cloud-tpu-v2.png
- cloud.google.com/images/products/tpu/cloud-tpu-v2-pod-alpha.png
- cloud.google.com/images/products/tpu/cloud-tpu-v3-alpha.png
- images.anandtech.com/doci/12195/google-tpu-board-2.png
- cloud.google.com/tpu/docs/tpus
- Link from a Wikipage to another Wikipage
- 19-inch rack
- 32-bit
- 8-bit
- Accumulator (computing)
- Activation function
- AI
- AI accelerator
- AlphaGo versus Lee Sedol
- AlphaZero
- Android (operating system)
- Application-specific integrated circuit
- Artificial neural network
- Asus
- Asus Tinker Board
- Binary multiplier
- Category:AI accelerators
- Category:Application-specific integrated circuits
- Category:Computer-related introductions in 2016
- Category:Google hardware
- Category:Microprocessors
- Central processing unit
- Chess
- CISC instruction
- Clock speed
- Cognitive computer
- Colaboratory
- Consumer Electronics Show
- Convolution
- Convolutional neural network
- Coral (AI)
- DDR3 SDRAM
- Debian
- Dual-channel
- Edge computing
- Embedded computing
- Floating-point arithmetic
- FLOPS
- Gibibyte
- Gigabyte
- Go (game)
- Google Cloud Platform
- Google Compute Engine
- Google Data Centers
- Google Photos
- Google Street View
- Graphics processing unit
- High Bandwidth Memory
- Internet of things
- Joule
- Kaggle
- M.2
- Machine learning
- Matrix multiplication
- MediaTek
- Memory bandwidth
- MHz
- MiB
- Millimetre
- Multi-chip module
- Nanometer
- Neural networks
- Norman Jouppi
- Nvidia
- O
- Operating system
- PCI-e
- PCI Express
- Pixel 4
- Pixel Neural Core
- Project Jupyter
- RankBrain
- Raspberry Pi
- Recurrent neural network
- Robotics
- Semiconductor device fabrication
- Shogi
- Single-board computer
- Spiking neural network
- Structure tensor
- System on module
- Systolic array
- Tensor Core
- TensorFlow
- Texture mapping
- Thermal design power
- TrueNorth
- USB
- Vision processing unit
- Watt
- Name
- enTensor Processing Unit
- SameAs
- 2N7kN
- Q25106376
- Tensorisuoritin
- Tensor Processing Unit
- Tensor Processing Unit
- Tensor Processing Unit
- Tensor Processing Unit
- Tenzorinis procesorius
- TPU
- Unidad de procesamiento tensorial
- Unitat de processament de tensors
- Тензорний блок обробки
- Тензорный процессор Google
- واحد پردازشی تنسور
- テンソル・プロセッシング・ユニット
- 张量处理单元
- 텐서 처리 장치
- Subject
- Category:AI accelerators
- Category:Application-specific integrated circuits
- Category:Computer-related introductions in 2016
- Category:Google hardware
- Category:Microprocessors
- Thumbnail
- Type
- enMachine learning
- Artificial neural network
- WasDerivedFrom
- Tensor Processing Unit?oldid=1123363965&ns=0
- WikiPageLength
- 22003
- Wikipage page ID
- 50673241
- Wikipage revision ID
- 1123363965
- WikiPageUsesTemplate
- Template:About
- Template:Authority control
- Template:Differentiable computing
- Template:Digital electronics
- Template:Google LLC
- Template:Infobox CPU architecture
- Template:Reflist
- Template:Short description
