For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).
Mojo struct
WarpStaggerRule
struct WarpStaggerRule
Declarative warp stagger configuration.
Controls how warp groups are staggered in the prologue to hide latency. The stagger index determines how many prologue ops G0 executes before G1 starts.
Modes:
- enabled=False: no stagger (single-buffer default)
- enabled=True, compute_from_body=True: derive stagger by counting stage-0 prefetch loads in the body (double-buffer default)
- enabled=True, compute_from_body=False: use fixed_amount directly
Fields
- enabled (
Bool): - compute_from_body (
Bool): - fixed_amount (
Int):
Implemented traits
AnyType,
Copyable,
ImplicitlyCopyable,
ImplicitlyDestructible,
Movable
Methods
none
static none() -> Self
No stagger (single-buffer default).
auto
static auto() -> Self
Auto-derive stagger from body (double-buffer default).
fixed
static fixed(amount: Int) -> Self
Fixed stagger amount.
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!