Ads Top

What is Google TPU? Does it replace GPU and CPU?

We are in a place where the Amazon with AWS and the Microsoft with Azure are ruling the cloud sector. In the recent Google I/O conference Google announced another processing unit called TPU, which replaces current GPU's.

TPU stands for Tensor Processing Unit. A tensor processing unit (TPU) is an application-specific integrated circuit (ASIC) developed specifically for machine learning. Compared to a graphics processing unit, it is designed explicitly for a higher volume of reduced precision computation upto a little as 8-bit precision. The chip has been specifically designed for Google's TensorFlow framework, however Google still uses CPUs and GPUs for other types of machine learning.

Google is a software company and its software needs a compatible hardware to run smoothly, thus Google has been working on its own hardware since 1 year and we could see its development in Pixel devices and upcoming products from Google.



Architecture:

The TPU is an 8-bit matrix multiply engine, driven with CISC instructions by the host processor across a PCIe 3.0 bus. It is manufactured on a 28 nm process with a die size ≤ 331 mm2. The clock speed is 700 MHz and has a thermal design power of 28-40 W. It has 28 MiB of on chip memory, and 4 MiB of 32-bit accumulators taking the results of a 256x256 array of 8-bit multipliers. Instructions transfer data to or from the host, perform matrix multiplies or convolutions, and apply activation functions
 For example, In Google Photos an individual TPU can process over 100 million photos a day. It is also used in RankBrain which Google uses to provide search results. The tensor processing unit was announced in 2017 at Google I/O, although the company stated that the TPU had been used their datacenter for over a year prior.

Google developed a next generation processing unit for AI which can process more datasets than current processing units used by Amazon and Microsoft, which means its gonna partially turning all the users interested in cloud towards Google.

Google is currently improving machine learning through Google assistant and Google Home, where the AI processes the large datasets and takes quick decisions through machine learning and it can influence a lot in the way we are using cloud and many other fields like Medicine, Automobiles, Smartcities etc.,

Presently, Google hasn't given much information regarding TPU but, it given about 1000 TPU's to developers, researchers and students to work on and develop projects using it.

Google's aim is to bring fast processing systems to a normal user where the applications like Google Street View, Map navigation and search results can be faster and more accurate and are processed from large amounts of data.

To know about  my recent posts :

• What is Google Opinion Rewards? How does it help you make money ? click here

• Why Chinese Mobiles are cheap? Are they any Good? click here

•  What is CAPTCHA? Why does it appear?  click here

• What is a VTOL Jet ? Is it a "Flying Car" ? click here

• Bluetooth 5.0 Explained...click here

Top 10 Best Windows 10 Creators Build Features...click here

• HISTORY OF ANDROID (ALPHA 1.0 - NOUGAT 7.0 ) click here

• what is VR (virtual reality)and AR(augmented reality)? click here

•  Is it better to put your PC to Shut down or leave it ON all night ?click here

•  QWERTY v/s DVORAK keyboards explained...  click here

•  32 bit v/s 64 bit processors explained..  click here

 •  DOES MORE RAM MAKES YOUR PC FASTER? click here

 •  DIFFERENT TYPES OF HACKERS EXPLAINED...click here

• 10 HACKING TECHNIQUES YOU SHOULD KNOW...click here

•  10 MYTHS WE BELIEVE WHILE OPERATING ON COMPUTERS.. click here

•   WHAT IS PROJECT FI ? IS GOOGLE COMING UP WITH CELLULAR CARRIER ? click here

•  TOP 10 WAYS TO MAKE YOUR PC RUN FASTER FOR FREE.. click here

• FOG COMPUTING....IS IT BETTER THAN CLOUD..??......click here

• Difference Between LTE and VoLTE in 4G Spectrum click here
  
• INTEL,NVIDIA,AMD which produces better graphics?? click here

• Li-Fi is it faster than Wi-Fi ?? click here

• Difference Between 2G,3G,4G,5G Technologies  click here

• CAN YOU LEAVE YOUR MOBILE TO CHARGE OVERNIGHT?? click here

• DIFFERENCES BETWEEN i3 ,i5 &i7 PROCESSORS..... click here

No comments:

Theme images by Storman. Powered by Blogger.