Nvidia Maxwell Quadro Upset

[Mr PotHead]

1) To use the same number of transistors or more? Yes you do. HDL actually shrinks the number of transistors used per component by eliminating redundancy (a safety feature actually in case a signal doesn't make it through, extremely necessary these days with so many quantum effects to deal with) and reduce clock rates (lowering the electrical noise on the wires to improve signaling) while adding more components (like the security chip in Carrizo). Every atomic part of a CPU core uses a defined number of transistors, but the manner in which they are used/organized will determine heat density. In order to pack them closer together and avoid heat problems you must either drop electrical power (reducing heat generated per transistor) or rotate the components to line up the hot and cold regions of each to decrease thermal density or do both.

Thus, like I said, either Nvidia came up with an HDL scheme which only applies to GM 200 (really unlikely that they did that, because it would increase chip yields if you could shrink GM 204 by 30% with it) or some "experts" have some explaining to do as to how they missed 64-bit SPs on the GM 204 chip.

 

2) Chiphell? Hardly. Chiphell gets extremely accurate when you get close to launch. It's only when you're multiple months away that they're generally way off. We're within a month of the Titan X. That's fairly accurate I guarantee you.

 

3) If you think AMD doesn't you have no clue how electrical design processes work. There are less than 100 people in the world that actually have a full understanding of these large chips. If you think they don't work on optimizing for power, space, and heat, you're blind. AMD sucks at making coolers though, no doubt.

 

AMD's dies are smaller than Nvidia's, by far.

that was so fuckin deep dude

[DildorTheDecent]

well i think i know what i must do now to get cool new tech. 

[HippY]

well i think i know what i must do now to get cool new tech. 

give Deadmau5 a blowjob?

[Maslofski]

dafuq does a DJ do with that much power?

[Deletive]

that was so fuckin deep dude

tech new threads usuallyu end up as different sided arguments.

[patrickjp93]

tech new threads usuallyu end up as different sided arguments.

As much as people complain we academics spend too much time in our ivory towers, the reality is few people do research better than a dedicated Master's/PHD student. Few people know how to properly argue and evaluate the world around them empirically, but what do I know? I'm just a college student without his Bachelor's (1.5 semesters left  ).

[Opcode]

AAAAActually AMD sucks, because NVIDIA leads with low TDP GPUs and still has the best GPU most of the time, if NVIDIA would stop its progress and make GPUs that uses as much power as AMD, NVIDIA would rule again so stop comparing GPUs if they are not in the same league

Wat?

 

[Guest Kloaked]

dafuq does a DJ do with that much power?

 

He records a lot of video with his expensive ass RED camera.

[Maslofski]

He records a lot of video with his expensive ass RED camera.

ah, thought he only did music

[CtW]

I swear I have seen this pic on Deadmau5's instagram

That's becasue that is a pic of his computer.

[CtW]

He records a lot of video with his expensive ass RED camera.

and he does a shit load of 3D rendering and shit. I'm pretty sure that he's responsible for all of his album covers and the design of his mouse heads.

[LukaP]

This is still very confusing. Quadros kinda need DP for the rendering workloads, and Teslas need it for scientific computing. 

Actually most rendering (like 99% of it) is based on SP calculations still, because its much faster, and rendering isnt really something that benefits a lot from extra precision, unless you are rendering like 16k+, with lots of tiny polygons and very highres textures

[patrickjp93]

Actually most rendering (like 99% of it) is based on SP calculations still, because its much faster, and rendering isnt really something that benefits a lot from extra precision, unless you are rendering like 16k+, with lots of tiny polygons and very highres textures

Usually what I've seen is renders are done in DP and then carefully are readjusted to SP to maximize quality. Did this practice die out?

[LukaP]

Usually what I've seen is renders are done in DP and then carefully are readjusted to SP to maximize quality. Did this practice die out?

ive never seen anyone do that. sounds like it would take waay more time. you use maximum settings for the production render, but single precision. unless of course you are so damn fancy you need to use double precision (basically what all the noobs do, but in the end, there is no percievable difference)

 

Octane, iray, keyshot, arnold are all single precision renderers. i dont think they even have a DP mode

[patrickjp93]

ive never seen anyone do that. sounds like it would take waay more time. you use maximum settings for the production render, but single precision. unless of course you are so damn fancy you need to use double precision (basically what all the noobs do, but in the end, there is no percievable difference)

Octane, iray, keyshot, arnold are all single precision renderers. i dont think they even have a DP mode

I mean the last time I was at all familiar with rendering tools I spent 2 weeks learning Maya with some 3D game artists from Rochester Institute of Tech, but that was what they did, a precise and then imprecise render and a reconciliation process.

[LukaP]

I mean the last time I was at all familiar with rendering tools I spent 2 weeks learning Maya with some 3D game artists from Rochester Institute of Tech, but that was what they did, a precise and then imprecise render and a reconciliation process.

The precision you use to calculate stuff is not what we call precise renders To do a production (or final) render, you set the renderer to be more precise you use more rays, bigger lighmap buffers, etc. and that stuff has a monumentaly bigger effect on the image than SP/DP.

[patrickjp93]

The precision you use to calculate stuff is not what we call precise renders To do a production (or final) render, you set the renderer to be more precise you use more rays, bigger lighmap buffers, etc. and that stuff has a monumentaly bigger effect on the image than SP/DP.

The curves and line work are still at the mercy of SP/DP.

[LukaP]

The curves and line work are still at the mercy of SP/DP.

what do you mean with that? as in what curves and linework?

[Sakkura]

Wat?

 

 

And it's passively cooled! :blink:

[TheGamingBarrel]

I feel great knowing that this is from Deadmau5's instagram.

[patrickjp93]

what do you mean with that? as in what curves and linework?

 

Given a skeleton model and a texture, how the final lines on the figure end up looking (properly curved?) is at the mercy of SP/DP. Enough errors in SP approximation can cause some strange end effects.

[patrickjp93]

And it's passively cooled! :blink:

 

Most of those "passively cooled" ones are in the presence of 10,000+ RPM fan arrays or are attached to water plates.

 

[Aniallation]

Quadro-M6000-GM200-GPU-635x635.jpg

dat airflow wow 11/10

[LukaP]

Given a skeleton model and a texture, how the final lines on the figure end up looking (properly curved?) is at the mercy of SP/DP. Enough errors in SP approximation can cause some strange end effects.

hence why i said "unless you render super tiny triangles with super highres textures, at 16k and beyond" or something like that. I have never seen a rendering error caused by lack of calculation precision. 99.99% of the time, SP is more than enough. i mean its still 32bits of precision, its not like we are talking 8 or 16 bit here

[Agent181]

snip

snip

why is there always fighting over this damn thing just go settle it on battlefield or csgo and see who's better that way