Class MachineProfile

Namespace

EncDotNet.S100.Renderers.Mapsui

Assembly

EncDotNet.S100.Renderers.Mapsui.dll

Computes the default tiled-renderer resource budgets for a PerformanceProfile, including the hardware-derived Auto tier. All members are deterministic given the (cores, RAM) inputs, so the tier mapping is unit-testable without touching the real machine.

public static class MachineProfile

Inheritance

object

MachineProfile

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Methods

AvailableRamGb()

Total available RAM in GB (cross-platform; reflects a container/VM cap).

public static double AvailableRamGb()

Returns

double

Resolve(PerformanceProfile)

Resolves profile against the live host's cores + RAM.

public static PerformanceProfile Resolve(PerformanceProfile profile)

Parameters

profile PerformanceProfile

Returns

PerformanceProfile

Resolve(PerformanceProfile, int, double)

Resolves Auto to a concrete tier from the supplied core count and available RAM. The explicit tiers pass through. LowEnd: ≤4 cores or ≤8 GB; Balanced: ≤8 cores or ≤16 GB; else HighEnd.

public static PerformanceProfile Resolve(PerformanceProfile profile, int cores, double ramGb)

Parameters

profile PerformanceProfile

cores int

ramGb double

Returns

PerformanceProfile

SkiaGpuResourceMb(PerformanceProfile)

Host GPU resource-cache budget, MB, for each resolved tier (see SkiaGpuResourceMb).

public static double SkiaGpuResourceMb(PerformanceProfile tier)

Parameters

tier PerformanceProfile

Returns

double

TileBudgetMb(PerformanceProfile)

Per-layer hot (CPU) tile-cache budget, MB, for each resolved tier.

public static double TileBudgetMb(PerformanceProfile tier)

Parameters

tier PerformanceProfile

Returns

double

TileDiskMb(PerformanceProfile)

Shared warm disk-cache budget, MB, for each resolved tier.

public static double TileDiskMb(PerformanceProfile tier)

Parameters

tier PerformanceProfile

Returns

double

TileGpuBudgetMb(PerformanceProfile)

Per-layer GPU-residency budget, MB, for each resolved tier.

public static double TileGpuBudgetMb(PerformanceProfile tier)

Parameters

tier PerformanceProfile

Returns

double

TileWorkers(PerformanceProfile)

Number of concurrent tile-rasterisation workers per layer for each tier, derived from the live host's logical-core count.

public static int TileWorkers(PerformanceProfile tier)

Parameters

tier PerformanceProfile

Returns

int

TileWorkers(PerformanceProfile, int)

Number of concurrent tile-rasterisation workers per layer for tier, given cores logical cores. LowEnd stays at the original single worker so constrained hosts are never over-subscribed; Balanced gets two; HighEnd scales with cores (one worker per ~4 cores) and is clamped to MaxTileWorkers. A pan that exposes ~14–48 cold tiles is otherwise drained serially by a single worker, so parallelism cuts the cold-tile queue wait roughly proportionally on multi-core hosts.

public static int TileWorkers(PerformanceProfile tier, int cores)

Parameters

tier PerformanceProfile

cores int

Returns

int