How to pronounce “tensor” in English

TensorFlow is an incredible tool.

and TensorFlow. These are the kind of thing that a lot of people in the computer industry learned of us from.

and tensors just generalize that a little bit more.

A simple kind of tensor is a vector.

Like the Tensor G5 is their latest chip and for years the Pixel has had, you know, the A-series comes out a little later,

And then the Tensor G5 is decently efficient, like I said, and using the phone has been smooth and I have no complaints with regular everyday use,

The metric tensor has two indices.

But the Riemann tensor is a 4 by 4 matrix of 4 by 4 matrices.

of the energy momentum tensor.

But the Riemann tensor isn't separate and independent from the metric.

And that Riemann tensor is responsible for spacetime curvature.

give it to TensorFlow, and maybe get the kind of 15-ish percent step change improvement that the Google data center saw when

It's all in TensorFlow.

He really developed tensors.

You're not doing tensors and general relativity, but you say, I know a really good search system plus maybe ads to monetize it.

And the same is true for tensors, only now instead of having just two components, a general tensor can have any number of them

And of course inside is the new Tensor G6 with 12 gigs of RAM, which I will have to benchmark and put through its

They also have talked about using Tensor to do faster Night Sight and smoother zoom while recording 8K video.

honestly probably says more about the Tensor than anything.

I am once again hoping for Tensor to take a massive leap forward with the next generation of Pixel phones, but I've been hurt before...

Because these different parts of the metric tensor are going to become physically relevant.

So it turns out there's a tensor that does this.

And the reason why you need a tensor to do it is because there's a huge amount of information involved in specifying the curvature of a four-dimensional spacetime

That's what the Riemann tensor provides you.

Plug it into the Riemann tensor, and it spits out how these lines will change in their relationship to each other as time goes on as you move down the path.

Unsurprising that it would be a tensor.

I do know it's a tensor equation with four different variables, and I figure there's somebody here who can explain this to me.

the work that's being done with TensorFlow and machine learning, and just all the advances that that's bringing about, my goal is to--

And it's also kind of close to the Tensor G2, which...

And you can use the metric to calculate the Riemann tensor.

There is a formula that lets you calculate the Riemann tensor once you know the metric.

So there's no more information in the Riemann tensor than there is in the metric.

The obvious strategy is to somehow cut the Riemann tensor down to size, to boil it down to a tensor with two indices

For example, tensor processing unit, TPU, is powered by Google.

Now IT band is basically a muscle called the tensor fasciae latae muscle that runs on the outside part of your hip across your hip joint and your knee joint.

And they're based on Caffe or on TensorFlow, now we're moving towards TensorFlow Lite.

You know, is it in the stress-energy tensor?

But Google also has TPUs, tensor processing units.

the Tensor G4...

It's just pointing out that this very compact notation of the Riemann tensor conveys an enormous amount of information.

And these are all grouped together to make something called the energy momentum tensor.

The point is, the energy momentum tensor is the right way of talking about energy and momentum and mass and all those other things

the curvature of spacetime, to the energy momentum tensor, representing stuff, matter and energy and all that stuff.

The hardest thing I ever did was I had to take courses in tensor calculus because I wanted to understand the math

underlying general relativity, which is a tensor calculus, and how the fabric of space time bends and folds

And that was a year of struggling to learn the fundamentals of tensor calculus.

Another might have the tensor behind G equals T.

And the AI initiatives that are allowing people to contribute in different ways-- use TensorFlow, and other things, to tap into that crowd knowledge.

Two simple tensors, G equals T. Again, I'm summarizing, but basically, that.