A behind-the-scenes look into the latest renderer technology powering the critically acclaimed DOOM. The lecture will cover how technology was designed for balancing a good visual quality and performance ratio. Numerous topics will be covered, among them details about the lighting solution, techniques for decoupling costs frequency and GCN specific approaches.

1. THE DEVIL IS IN THE DETAILS

2. Initial Requirements











3. Anatomy of a Frame
Frame Cost











4. Data Structure for Lighting & Shading










5. Preparing Clustered Structure




 𝑍𝑆𝑙𝑖𝑐𝑒 = 𝑁𝑒𝑎𝑟𝑧 × (
𝐹𝑎𝑟𝑧
𝑁𝑒𝑎𝑟𝑧
)
𝑠𝑙𝑖𝑐𝑒
𝑛𝑢𝑚 𝑠𝑙𝑖𝑐𝑒𝑠





6. Preparing Clustered Structure






//Pseudo-code - 1 job per depth slice ( if any item )
for ( y = MinY; y < MaxY; ++y ) {
for ( x = MinX; x < MaxX; ++x ) {
intersects = N planes vs cell AABB
if ( intersects ) {
Register item
}
}
}

7. Preparing Clustered Structure














8. Preparing Clustered Structure

9. Preparing Clustered Structure

10. Detailing the World










11. Detailing the World













12. Detailing the World

 𝑒0 𝑒1


const float4 albedo = tex2Dgrad( decalsAtlas, uv.xy * scaleBias.xy + scaleBias.zw, uvDDX, uvDDY );
𝑒
𝑒 𝑒
𝑒0 𝑥
𝑒1 𝑥 2 𝑥 𝑥
𝑒0 𝑦
𝑒1 𝑦 2 𝑦 𝑦
𝑒0 𝑧
𝑒1 𝑧 2 𝑧 𝑧
0 0 0 1
0.5
𝑠𝑖𝑧𝑒𝑋
0 0 0.5
0
0.5
𝑠𝑖𝑧𝑒𝑌
0 0.5
0 0
0.5
𝑠𝑖𝑧𝑒𝑍
0.5
0 0 0 1
∙

13. Detailing the World











14. Detailing the World

15. Detailing the World

16. Detailing the World

17. Detailing the World

18. Detailing the World

19. Detailing the World

20. Lighting











21. Lighting
//Pseudocode
ComputeLighting( inputs, outputs ) {
Read & Pack base textures
for each decal in cell {
early out fragment check
Read textures
Blend results
}
for each light in cell {
early out fragment check
Compute BRDF / Apply Shadows
Accumulate lighting
}
}

22. Lighting













23. Lighting











24. Lighting



25. Lighting

 




 




26. Transparents







27. Particle Lighting











28. Decoupled Particle Lighting











29. Decoupled Particle Lighting
//Pseudocode – Particle shading becomes something like this
Particles( inputs, outputs ) {
…
const float3 lighting = tex2D( particleAtlas, inputs.texcoord );
result = lighting * inputs.albedo;
…
}

30. Decoupled Particle Lighting









Particle Light Atlas

31. Decoupled Particle Lighting


32. Post-Process

33. Optimizing Data Fetching ( GCN )









34. Clustered Lighting Access Patterns

35. Clustered Lighting Access Patterns

36. Clustered Lighting Access Patterns

37. Clustered Lighting Access Patterns

38. Analyzing the Data








39. Leveraging Access Patterns










Thread X A B D
Thread Y B C E
Thread Z A C D E
Scalar A B C D E

40. Special Paths












41. Dynamic Resolution Scaling










42. Async Post Processing










43. GCN Wave Limits Tuning








44. GCN Register Usage










51. What’s next ?








52. Special Thanks







53. 

We are Hiring !

54. Thank you






55. References















56. Bonus Slides

57. Lighting
















58. Deferred Passes







