Slashdot Mirror
Google's Tensor Processing Unit Could Advance Moore's Law 7 Years Into The Future (pcworld.com)
An anonymous reader writes from a report via PCWorld: Google says its Tensor Processing Unit (TPU) advances machine learning capability by a factor of three generations. "TPUs deliver an order of magnitude higher performance per watt than all commercially available GPUs and FPGA," said Google CEO Sundar Pichai during the company's I/O developer conference on Wednesday. The chips powered the AlphaGo computer that beat Lee Sedol, world champion of the game called Go. "We've been running TPUs inside our data centers for more than a year, and have found them to deliver an order of magnitude better-optimized performance per watt for machine learning. This is roughly equivalent to fast-forwarding technology about seven years into the future (three generations of Moore's Law)," said Google's blog post. "TPU is tailored to machine learning applications, allowing the chip to be more tolerant of reduced computational precision, which means it requires fewer transistors per operation. Because of this, we can squeeze more operations per second into the silicon, use more sophisticated and powerful machine learning models, and apply these models more quickly, so users get more intelligent results more rapidly." The chip is called the Tensor Processing Unit because it underpins TensorFlow, the software engine that powers its deep learning services under an open-source license.
so that they'll end up on the pile of broken dreams just like all those other amazing Google projects that we're going to save the world.
What's supposed to be amazing about this? Isn't decades-old knowledge that you can make computations faster by reducing the precision and using faster approximation methods?
I only see four or five stories from them, so maybe not...
Moore's law relates to the number of components in an integrated circuit.
I really doubt these things put more transistors onto a piece of silicon.
of Teh G and ye shall live for eternity -- in HELL, in the mouth of The Beast.
It's because today was Google I/O, Google's developer conference, so a lot of projects were announced. Still, Slashdot could have combined all these into story with links to details rather than spam us with a ton of Google stories. However, this TPU project might be the most interesting thing to come out of the conference. Not because the chips are novel (it's just the same principles as GPUs but taken to a further extreme), but because it sounds like Google's getting into low level chip manufacture. We'll have to wait and see if Google can deliver more FLOPS per dollar/watt than the leading co-processor manufacturer, NVIDIA.
Specialized processing chips have been several 'generations' ahead in terms of processing per dollar for many decades. In the 90's at least, DSPs were doing audio/video processing much cheaper by performing many machine-level steps simultaneously in one 'cycle' with less power than a general processor, by leaving out the features and cost of a general processor. And all you had to do to use them was test them on a hardware emulator, flash them, then pop them into production test run until you were good enough to deploy. Depending on the chip, they could run on a trickle of power, without active cooling, and match a much most costly general chip for pennies.
I mean, it's how we got cell phones, and LOTS of other things, including most things in a computer that aren't the CPU.
But isn't Moore's law more about transistors per unit cost, rather than performance per cost? Seems like a fundamental misunderstanding in the headline... which seems about as common as specialized chips in modern technology.
Ryan Fenton
I will trust Intel, AMD and NVidia to sustain Moore's law as it pertains to general purpose computing, not Google. Google gets plaudits for advancing neural net hardware, if indeed they didn't just buy the tech and slap the Google brand on it, which is likely. Thw hyperbole just erodes credibility, in other words, makes me wonder how many other exaggerations will turn up in this department. Yes, it's a fact it plays Go well, and no doubt does a lot of other things well. Let's stick to the facts please. It didn't advance rendering by 7 years, it didn't advance compiling, in fact it didn't advance just about anything I do with my own computer. Putting my own jaded spin on it, it mainly advances Google's ability to invade my privacy and monetize me.
When all you have is a hammer, every problem starts to look like a thumb.
End of message.
This is the FIFTH Google story today.
If you want news from today, you have to come back tomorrow.
Tensor Processing Units are not new. SGI used to offer that for their Octane, aimed pretty heavily at the satellite image analysis crowd.
Not sure anyone has said who is "manufacturing" these units for google. I doubt google is producing them themselves.
Would love to know who it is though.
Because usually it's combined with "green" "eco" and "slower than the last generation of hardware". I don't care about performance per watt; I care about performance. Granted in a data center setting wattage(and heat) is important, however, that hardly translates to talking about Moore's law.
Perhaps when it isn't something that is likely to take multiple lines of work out in one shot, I'll be for it.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
says "Samaritan" (ref to the TV series 'Person of Interest')
sorry but no. Google is a no-go zone for me these days.
They are bent of boing evil to humanity in order to get a few $$$$ more Ad revenue.
When that dries up, they will attempt to rule all our lives 24/7/52
I could flap my arms and fly to the moon. Well, I *could*. Why does everyone like to gag on the cock of google?
Less precision computing, machine learning and neural networks. Mm.. trying too hard to emulate the brain here. Tell me when you reach 20W power draw and 40 petaflops.
Was that not something introduces about 20 years ago by Silicon Graphics? Or am I getting old
http://manx.classiccmp.org/mir...
Well, Moores Law is about hardware. Never heard of Denning.
In the early days, hardware speed and transistor density determined the capabilities of a computer prety well.
Plain old logic and linear programmng. This was advanced somewhat by specialized hardware when array processng
and GPus were developed. Capability was still linked mostly to hardware. Software contributed to the capabilities
by being written to use the hardware and parallel programming techniques. All of these can be measured in terms
of FLOPS.
The advance of AI is different. Measurement here should be based on the standard math/logic ( FLOPS ) and some set of measure
of the information manipulation and processing capabilities. Memory and data handling come into play.
Pattern recognition and analysis of data are integral to this also. And - to be sure that all is accounted for - advances will
require more database capability, more disk storage, and more computing hardware....
Ai capabilities are still using hardware, so anyone who uses moores law for Ai description is only mildly retarded.
( little i in Ai since intelligence is still specific - idiot savant...)
I propose several alternatves, based on animal and human development.
Peak law ( what is the peak capabilty of the Ai?). All intelligences peak at some level... how high?
Stupid Law ( how stupid can it potentially be?). Self-explanatory.
CS Law ( common-sense measurement - humans have this law also...). Self-explanatory.
Humor law ( recogniton and use of humor IS an intelligence measure ).
Moore's law was about feature size and advances continue to slow.
This product is about parallelism. Nothing new.
At least.
Not related to the wizard Tensor of Tensor's Floating Disc.
Remember kids, if you're not paying for the service, YOU ARE THE PRODUCT THAT IS BEING SOLD.
Am I reading this right, the basic gist of the article is that custom, purposebuilt chips are faster at very specific tasks than general all-purpose chips?
Such amaze, much fast, wow.
>...Could Advance Moore's Law 7 Years Into The Future
No, they built an ASIC that accelerates a SPECIFIC APPLICATION - doesn't do shit to Moore's law.
"Google's Law" states that whenever you need to make something sound cool and innovative, just misuse and existing term like "Moore's Law", because reporters are stupid.
~Any apparent grammatical or typographic errors are caused by defects in your display device.
See, this is what's wrong with polls.
"Do you care about transistor density?", will get a lot of "Huh?", which will be interpreted as "no, does not care".
"Transistor density is one of the biggest factors in computer speed. Do you care about making transistors more dense in the future to improve computer speed?" will get a lot of "Yes", because people want the computer speed, even if they don't understand transistor density.
Polling has gotten a bad name because the loaded questions spew the results, and make headlines.
Oh, I didn't give this the right headline -- "Slashdot Readers Care More About The Headline", because of the poll question, "Will a bad, uninformative headline that doesn't interest you result in your not clicking on the story?"
The real question on advances over GPU's: Can it mine bitcoin faster and for less electricity?