Conference Guide

Profiling: Principles and Practices for Performant Processing - Q&A
Filter mode:
Topics
Media
language
memory
profiling
(null topic)
🗪Audience contribution🟉
🎭
🏟
🔗
Link to: current timestamp
Credits
Host
Abner Coimbre
Presenter
Elizabeth Baumel
Indexer
Matt Mascarenhas
?
?

Keyboard Navigation

Global Keys

[, < / ], > Jump to previous / next episode
W, K, P / S, J, N Jump to previous / next timestamp
t / T Toggle theatre / SUPERtheatre mode
V Revert filter to original state Y Select link (requires manual Ctrl-c)

Menu toggling

q Quotes r References f Filter y Link c Credits

In-Menu and Index Controls

a
w
s
d
h j k l
←
↑
↓
→

Esc Close menu / unfocus timestamp

Quotes and References Menus and Index

Enter Jump to timestamp

Quotes, References and Credits Menus

o Open URL (in new tab)

Filter Menu

x, Space Toggle category and focus next
X, ShiftSpace Toggle category and focus previous
v Invert topics / media as per focus

Filter and Link Menus

z Toggle filter / linking mode

Credits Menu

Enter Open URL (in new tab)
0:00postgoodism Is "Heisenbottleneck" a thing? If not, can I coin it now?
0:00postgoodism Is "Heisenbottleneck" a thing? If not, can I coin it now?
0:00postgoodism Is "Heisenbottleneck" a thing? If not, can I coin it now?
0:09scottlee Pro tip: if you're using C# or a garbage collected language and you determine that's the cause of your spikes, you can at least allocate upfront and then re-use those points in memory to avoid garbage collection. For those of us that actually decided to use Unity, that's a real thing that has to be thought about. And then later you can use a lower level language instead
0:09scottlee Pro tip: if you're using C# or a garbage collected language and you determine that's the cause of your spikes, you can at least allocate upfront and then re-use those points in memory to avoid garbage collection. For those of us that actually decided to use Unity, that's a real thing that has to be thought about. And then later you can use a lower level language instead
0:09scottlee Pro tip: if you're using C# or a garbage collected language and you determine that's the cause of your spikes, you can at least allocate upfront and then re-use those points in memory to avoid garbage collection. For those of us that actually decided to use Unity, that's a real thing that has to be thought about. And then later you can use a lower level language instead
0:38cipherpunk Can we call hot-chocolaty opinions "chocolate whine"?
0:38cipherpunk Can we call hot-chocolaty opinions "chocolate whine"?
0:38cipherpunk Can we call hot-chocolaty opinions "chocolate whine"?
0:52jshbrntt What was the use-case for using 8 GPUs?
0:52jshbrntt What was the use-case for using 8 GPUs?
0:52jshbrntt What was the use-case for using 8 GPUs?
1:35pal When diving into the bottlenecks of a codebase, how much upfront effort is usually spent just getting a handle of what the code is even doing vs profiling it?
1:35pal When diving into the bottlenecks of a codebase, how much upfront effort is usually spent just getting a handle of what the code is even doing vs profiling it?
1:35pal When diving into the bottlenecks of a codebase, how much upfront effort is usually spent just getting a handle of what the code is even doing vs profiling it?
2:58michael Is profiling a full-time job or just part of a broader role?
2:58michael Is profiling a full-time job or just part of a broader role?
2:58michael Is profiling a full-time job or just part of a broader role?
3:46abeyer Is there a good tl;dr for GPU counters to care about if you know you're GPU-bound but know nothing about GPU?
3:46abeyer Is there a good tl;dr for GPU counters to care about if you know you're GPU-bound but know nothing about GPU?
3:46abeyer Is there a good tl;dr for GPU counters to care about if you know you're GPU-bound but know nothing about GPU?
6:24slimsag If you had to pick one metric to replace FPS what would it be?
6:24slimsag If you had to pick one metric to replace FPS what would it be?
6:24slimsag If you had to pick one metric to replace FPS what would it be?
6:29ginger_bill ms
6:29ginger_bill ms
6:29ginger_bill ms
6:32byp Are there any common profiling pitfalls that you see early engineers making?
6:32byp Are there any common profiling pitfalls that you see early engineers making?
6:32byp Are there any common profiling pitfalls that you see early engineers making?
7:54mark.dev Has profiling the same code across different GPU models revealed an interesting difference? How often does that make a relevant difference?
7:54mark.dev Has profiling the same code across different GPU models revealed an interesting difference? How often does that make a relevant difference?
7:54mark.dev Has profiling the same code across different GPU models revealed an interesting difference? How often does that make a relevant difference?
9:05klacansky What tools do you use to systematically track time spent in specific subsystems / passes over time? Do you run these every day?
9:05klacansky What tools do you use to systematically track time spent in specific subsystems / passes over time? Do you run these every day?
9:05klacansky What tools do you use to systematically track time spent in specific subsystems / passes over time? Do you run these every day?