Search found 56 matches

by hobold
Thu May 30, 2013 7:09 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25212

Re: Quadrics kinda sorta work for real time

Do you have a formulation for automatically fitting these to a tri/quad mesh? Also, you say piece-wise quadric. Does this have any implications for the continuity of the overall surface? Short version: this is no better or worse than the Zwart-Powell Element mentioned earlier in this thread. Long v...
by hobold
Thu May 30, 2013 12:13 am
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25212

Re: Quadrics kinda sorta work for real time

After dabbling with CUDA and OpenCL, I have been drawn back to the more theoretical side of things. Been reading a lot about subdivision surfaces, about box splines and simplex splines ... experimented some more. I stumbled on something interesting that might close a gap in our understanding of one ...
by hobold
Fri Apr 19, 2013 12:53 pm
Forum: General Development
Topic: Is GPU computing dead already?
Replies: 20
Views: 30972

Re: Is GPU computing dead already?

This means that in theory, you can take any language that supports LLVM and compile it directly to a GPU binary. The problem is that the GPU target instruction sets have limitations. The LLVM internal representation may contain constructs that cannot directly be mapped to a GPU target ... the respe...
by hobold
Tue Apr 16, 2013 7:30 pm
Forum: General Development
Topic: Ellipsoid/Triangle intersection
Replies: 4
Views: 4804

Re: Ellipsoid/Triangle intersection

What information exactly is your input? - are the ellipsoid's main axes arbitrary? or parallel to some coordinate axis? in world coordinates or in local coordinates? - is the triangle specified by its three vertices or some other way? What information exactly do you want to compute? - a boolean "doe...
by hobold
Sat Apr 06, 2013 10:38 am
Forum: General Development
Topic: Can't pass virtual function objects to CUDA kernel is a pain
Replies: 7
Views: 6500

Re: Can't pass virtual function objects to CUDA kernel is a

It's probably worth noting that GPUs should be able to handle function pointers more efficiently than CPUs for micro-architectural reasons (if you're interested, I can explain), so be sure to give them a try. Are you referring to the GPU's latency hiding strategies with this comment? Or something e...
by hobold
Sat Mar 30, 2013 12:58 am
Forum: General Development
Topic: Is GPU computing dead already?
Replies: 20
Views: 30972

Re: Is GPU computing dead already?

The whole computing landscape is shifting, because the era of exponentially increasing single thread CPU performance is over. IMHO this means two things: 1. Mainstream computing performance will actually shrink, as the mainstream shifts from big bulky personal computers to tiny little gadgets. 2. Pa...
by hobold
Wed Mar 20, 2013 5:39 pm
Forum: Links & Papers
Topic: Haswell benchmarks show up
Replies: 20
Views: 18591

Re: Haswell benchmarks show up

OpenCL on FPGA Do you happen to be one of the people using that kind of setup? If so, are you allowed to talk about your experience with that hardware? (BTW, thanks for your efforts regarding OpenCL. So far, MandelGPU has been the most useful example code for my own attempts to learn and tinker wit...
by hobold
Wed Mar 20, 2013 12:22 am
Forum: Links & Papers
Topic: Haswell benchmarks show up
Replies: 20
Views: 18591

Re: Haswell benchmarks show up

the speed up is quite small: something in the 10%-25% range. Please do not take this as an attack; and not as directed specifically at you. I just want to draw everybody's attention to the awful truth that ... ... the time of exponential performance improvements from new processor generations is ov...
by hobold
Mon Dec 03, 2012 11:09 pm
Forum: General Development
Topic: Quadrics kinda sorta work for real time
Replies: 34
Views: 25212

Re: Quadrics kinda sorta work for real time

Slow progress (no bonus points if you can spot the obvious bugs and flaws): http://vectorizer.org/boar/qRvTorus6x6.png This is based on midedge subdivision of a toroidal control polyhedron of six by six vertices (i.e. a coarse torus with hexagonal cross sections for both ring and tube). It's been su...
by hobold
Wed Nov 21, 2012 8:05 pm
Forum: Visuals, Tools, Demos & Sources
Topic: Intel Xeon Phi Preferred Pricing - Only $400 per Card ?
Replies: 29
Views: 30778

Re: Intel Xeon Phi Preferred Pricing - Only $400 per Card ?

Phi has dozens of megabytes of cache, compared to hundreds of kilobytes on a typical GPU chip. GPUs have to stream potentially hundreds of megabytes of textures per frame, so they did not (need to) evolve large caches. Phi was retargeted from graphics to compute, so maybe it can make good use of its...