Doom GPU Flame Graphs
brendangregg.com
Doom GPU Flame Graphs
1–10 of 13 posts
Re: Doom GPU Flame Graphs
#2This was an interesting read, along with the companion article about the early development from last year [0].
I feel like this is a very interesting development, but I can't quite put my finger on why. It's also interesting to compare with previous comments about Nvidia have much less available [1]. I wonder if tooling like this (along with Brendan and his team) could prove to be a competitive advantage for Intel?
[0] https://www.brendangregg.com/blog/2024-10-29/ai-flame-graphs... [1] https://news.ycombinator.com/item?id=41994469
Re: Doom GPU Flame Graphs
#3tbh I find magic trace to be more useful
https://github.com/janestreet/magic-trace
Re: Doom GPU Flame Graphs
#4tbh I find magic trace to be more useful https://github.com/janestreet/magic-trace
No, this samples the GPU as well. That's the novelty.
Re: Doom GPU Flame Graphs
#5This was an interesting read, along with the companion article about the early development from last year [0]. I feel like this is a very interesting development, but I can't quite put my finger on why. It's also interesting to compare with previous comments about Nvidia have much less available [1]. I wonder if tooling like this (along with Brendan and his team) could prove to be a competitive advantage for Intel? […
From your zeroth link,
>A flame graph is a visualization I invented in 2011 for showing sampled code stack traces.
Another day, another lie on the internet.
I though Gregg was cool.
Re: Doom GPU Flame Graphs
#6This was an interesting read, along with the companion article about the early development from last year [0]. I feel like this is a very interesting development, but I can't quite put my finger on why. It's also interesting to compare with previous comments about Nvidia have much less available [1]. I wonder if tooling like this (along with Brendan and his team) could prove to be a competitive advantage for Intel? […
From your zeroth link, >A flame graph is a visualization I invented in 2011 for showing sampled code stack traces. Another day, another lie on the internet. I though Gregg was cool.
Who is the real inventor in your opinion?
Re: Doom GPU Flame Graphs
#7This was an interesting read, along with the companion article about the early development from last year [0]. I feel like this is a very interesting development, but I can't quite put my finger on why. It's also interesting to compare with previous comments about Nvidia have much less available [1]. I wonder if tooling like this (along with Brendan and his team) could prove to be a competitive advantage for Intel? […
From your zeroth link, >A flame graph is a visualization I invented in 2011 for showing sampled code stack traces. Another day, another lie on the internet. I though Gregg was cool.
Flame graph, as in a display of sampled call stacks growing upward organized per level to aggregate like call stacks, is almost certainly their invention.
Why they chose that over the much more common visualization of downward call stacks that has been standard for decades is a mystery. That is the form most commonly used to display call stacks in trace visualizers since at least the early 2000s (though likely earlier).
Re: Doom GPU Flame Graphs
#8Earlier quoted context omitted.
From your zeroth link, >A flame graph is a visualization I invented in 2011 for showing sampled code stack traces. Another day, another lie on the internet. I though Gregg was cool.
Flame graph, as in a display of sampled call stacks growing upward organized per level to aggregate like call stacks, is almost certainly their invention. Why they chose that over the much more common visualization of downward call stacks that has been standard for decades is a mystery. That is the form most commonly used to display call stacks in trace visualizers since at least the early 2000s (though likely earlie…
K
See [1], and many others.
1: https://web.archive.org/web/20091124115438/https://graphics....
Re: Doom GPU Flame Graphs
#9Earlier quoted context omitted.
From your zeroth link, >A flame graph is a visualization I invented in 2011 for showing sampled code stack traces. Another day, another lie on the internet. I though Gregg was cool.
Flame graph, as in a display of sampled call stacks growing upward organized per level to aggregate like call stacks, is almost certainly their invention. Why they chose that over the much more common visualization of downward call stacks that has been standard for decades is a mystery. That is the form most commonly used to display call stacks in trace visualizers since at least the early 2000s (though likely earlie…
The answer has always been in the source of flamegraph.pl:
# This was inspired by Neelakanth Nadgir's excellent function_call_graph.rb
# program, which visualized function entry and return trace events. As Neel
# wrote: "The output displayed is inspired by Roch's CallStackAnalyzer which
# was in turn inspired by the work on vftrace by Jan Boerhout".
Neel's was the biggest direct inspiration. I changed it to put the alphabet on the x-axis and reduced the color hues, but they look similar.
Re: Doom GPU Flame Graphs
#10Earlier quoted context omitted.
Flame graph, as in a display of sampled call stacks growing upward organized per level to aggregate like call stacks, is almost certainly their invention. Why they chose that over the much more common visualization of downward call stacks that has been standard for decades is a mystery. That is the form most commonly used to display call stacks in trace visualizers since at least the early 2000s (though likely earlie…
K See [1], and many others. 1: https://web.archive.org/web/20091124115438/https://graphics....
Nice tool. Looks like it uses what we now call a "flame chart" (not a flame graph). I don't know if it ever had an original name, but I've seen these in tools from at least the 1990's.