Conference Guide

Practical DOD - Q&A
Filter mode:
Topics
Media
caching
language
optimisation
performance
(null topic)
🗪Audience contribution🟉
🎭
🏟
🔗
Link to: current timestamp
Credits
Host
Abner Coimbre
Presenter
Andrew Kelley
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)
0:00laurelkeys Did you intend to buy the shirt for Halloween and happened to use it for the talk, or was it the other way around?
0:00laurelkeys Did you intend to buy the shirt for Halloween and happened to use it for the talk, or was it the other way around?
0:00laurelkeys Did you intend to buy the shirt for Halloween and happened to use it for the talk, or was it the other way around?
0:26danbokser Is there any performance cost for using 32 bits for an array index on a 64-bit system, or do compiler optimizations get rid of this issue?
0:26danbokser Is there any performance cost for using 32 bits for an array index on a 64-bit system, or do compiler optimizations get rid of this issue?
0:26danbokser Is there any performance cost for using 32 bits for an array index on a 64-bit system, or do compiler optimizations get rid of this issue?
1:04pal When is the right time to use bit-fields within a struct as opposed to creating custom encoding layouts (e.g. the bee / human enums)? Should custom data encodings be the first tool we reach for?
1:04pal When is the right time to use bit-fields within a struct as opposed to creating custom encoding layouts (e.g. the bee / human enums)? Should custom data encodings be the first tool we reach for?
1:04pal When is the right time to use bit-fields within a struct as opposed to creating custom encoding layouts (e.g. the bee / human enums)? Should custom data encodings be the first tool we reach for?
2:06pulp_user Do you think about DOD as an optimization pass that you apply to existing code, or would you apply it straight away on new projects? It seems like it's most effective once you have explored the problem space pretty well
2:06pulp_user Do you think about DOD as an optimization pass that you apply to existing code, or would you apply it straight away on new projects? It seems like it's most effective once you have explored the problem space pretty well
2:06pulp_user Do you think about DOD as an optimization pass that you apply to existing code, or would you apply it straight away on new projects? It seems like it's most effective once you have explored the problem space pretty well
2:35klltkr If you use MultiArrayList (first MultiArrayList example) and you are accessing both the Animation pointer and the kind during an operation, how does that affect the caching as you switch between accessing those fields? Are you relying on the separate lists being in separate cache lines (to avoid evicting each other)? Not a bad assumption if you have 1024, but I don’t know how the machine decides which cache lines to use for a particular access
2:35klltkr If you use MultiArrayList (first MultiArrayList example) and you are accessing both the Animation pointer and the kind during an operation, how does that affect the caching as you switch between accessing those fields? Are you relying on the separate lists being in separate cache lines (to avoid evicting each other)? Not a bad assumption if you have 1024, but I don’t know how the machine decides which cache lines to use for a particular access
2:35klltkr If you use MultiArrayList (first MultiArrayList example) and you are accessing both the Animation pointer and the kind during an operation, how does that affect the caching as you switch between accessing those fields? Are you relying on the separate lists being in separate cache lines (to avoid evicting each other)? Not a bad assumption if you have 1024, but I don’t know how the machine decides which cache lines to use for a particular access
4:08ilidemi Were the wall clock time and memory savings numbers per pass or compared to full compilation?
4:08ilidemi Were the wall clock time and memory savings numbers per pass or compared to full compilation?
4:08ilidemi Were the wall clock time and memory savings numbers per pass or compared to full compilation?
4:53lf How did you find the structs that are likely the ones that are most beneficial to optimize?
4:53lf How did you find the structs that are likely the ones that are most beneficial to optimize?
4:53lf How did you find the structs that are likely the ones that are most beneficial to optimize?
6:28scottjmaddox Do you think future high-level languages (not Zig) could automate these kinds of data oriented memory layout transformations?
6:28scottjmaddox Do you think future high-level languages (not Zig) could automate these kinds of data oriented memory layout transformations?
6:28scottjmaddox Do you think future high-level languages (not Zig) could automate these kinds of data oriented memory layout transformations?
8:08pulp_user Is changing and adapting DOD code to new requirements more laborious than changing non-DOD code, in your experience?
8:08pulp_user Is changing and adapting DOD code to new requirements more laborious than changing non-DOD code, in your experience?
8:08pulp_user Is changing and adapting DOD code to new requirements more laborious than changing non-DOD code, in your experience?