Revenge of Cell, the DPU Rises *spawn*

[onQ]

You're just making up definitions now, which perhaps explains why no-one can follow your reasoning. A DPU isn't a generic accelerator and you shouldn't be referring to accelerators as DPUs. If you want to speed up raytracing, adding a raytracing accelerator doesn't mean adding a DPU (although if you can add a DPU as a class of Tensilica architecture and have that do the job). Same with AI or video decoding. These accelerators will conform to a processor architecure as defined by whatever taxonomy one uses, for which I don't think there's an officially accepted one. So some accelerators will be DSPs, and others will be ultra wide SIMD processors, etc.

Considering people asked for a definition of DPU earlier to be able to enter into this discussion, and you left it to others of us before offering your definition which none of us could guess, it's no wonder this is another pointless thread that no-one can do anything with.

Stop.

Math has always been the foundational core of all computer anything. Triangles are no exception. Triangles were selected for a variety of reasons over quads for instance. Alongside the GPU we had and still have Larger register extensions. Before that we had math coprocessors. Your explanation sucks. GPUs are good at large matrices math. Even then, there are still particular math functions it will not be good at with certain hardware.

You're all over the place, I don't even need to read anything of what you've written to know that. There are many things we've done that can be done differently but that doesn't necessarily mean it's going to be faster or slower for that matter. Things have become the way they are as a proper form of evolution over time and we can tie a lot of decisions back to costs.

Someone needed a simple example & I gave a simple example.

 

[onQ]

By the way you made this thread off of something that I said in the next gen prediction thread & now you're saying that it is a pointless thread & that I left it up to other people to figure out what a DPU is?

 

[iroboto]

Someone needed a simple example & I gave a simple example.

A link to Wikipedia would have been sufficient. https://en.m.wikipedia.org/wiki/Tensilica

It is not what you described at all. That's like me saying an Xilinx FGPA is a CPU/GPU/DPU waiting to happen.

 

[Shifty Geezer]

A link to Wikipedia would have been sufficient. https://en.m.wikipedia.org/wiki/Tensilica

That links back to the definition I provided authored by Grant Martin, Chief Scientist, Tensilica.

It's actually a frustrating article because it ignores completely the GPU, leaving us guessing whether Mr. Marting considers a GPU to be a DPU or something else.

But this is immaterial to the issue of onQ trying to start a debate without being clear on the points or the terms and, as ever, leaving us guessing what the heck he's actually trying to say between the unexplained images and single liners.

 

[iroboto]

That links back to the definition I provided authored by Grant Martin, Chief Scientist, Tensilica.

It's actually a frustrating article because it ignores completely the GPU, leaving us guessing whether Mr. Marting considers a GPU to be a DPU or something else.

But this is immaterial to the issue of onQ trying to start a debate without being clear on the points or the terms and, as ever, leaving us guessing what the heck he's actually trying to say between the unexplained images and single liners.

That's fair. But I guess where I'm going with this is that few of us are actually truly qualified to talk about next generation rendering techniques programming, even fewer on this board on next generation hardware programming for rendering.

This discussion spinoff is about the hypothetical repurposing of a programmable hardware to speed up X or Y. That's so pie in the sky.

No one here is qualified to talk about it, it's a topic where truthfully have no information on: if DPUs become GPU data flow accelerators then it's a topic warrant discussing because we will have more information to work with. But at the moment it's not, and we are left with trying to figure out how hardware could work with marketing materials??

 

[onQ]

A link to Wikipedia would have been sufficient. https://en.m.wikipedia.org/wiki/Tensilica

It is not what you described at all. That's like me saying an Xilinx FGPA is a CPU/GPU/DPU waiting to happen.

Look through the thread I have posted youtube videos & everything.

 

[Scott_Arm]

A GPU is a DPU that was made for Geometry.

I'll explain: Devs learned how to make triangles so engineers made a special processor for it that was fast at delivering triangles. now think of all the other things that devs have learned to do with a CPU but need a faster way to deliver it. that's what the DPUs are for.

This is clearly not true. A DPU is a product sold by Tensilica. Not all GPUs are DPUs. A DPU appears to be an customizable IC, with a customizable instruction set. I asked earlier in this thread what you thought the relevance of DPUs was to the discussion, and received no answer. If all GPUs are DPUs, then there is no relevance, because everything is a DPU and always has been a DPU. It becomes an all-encompassing term that means nothing.

 

[Scott_Arm]

Look through the thread I have posted youtube videos & everything.

Your posting style is like handing in an appendix and bibliography for a term paper that you never wrote, and expecting the professor to infer what you meant. Links and images are great, but they require context. They're meant to be supplemental to explain,justify what you've written. Part of writing is proving that you understand what you're talking about. In your case, we can not clearly understand what you're trying to say or that you even understand the material.

 

[Shifty Geezer]

onQ's inability to discuss properly has seen his posting rights removed. This thread serves no purpose.