Package jline.api.cache

Class Cache_miss_fifo_rmf

java.lang.Object

jline.api.cache.Cache_miss_fifo_rmf

public final class Cache_miss_fifo_rmf
extends Object

Position-resolved mean-field miss rates for FIFO(m) caches.

FIFO(m) and RANDOM(m) share the exact stationary distribution (Gast15 Thm 1), so FLD serves FIFO steady state from Cache_miss_rmf; the dedicated value of this class is the FIFO mean-field TRANSIENT (cache_miss_fifo_rmf_tran(jline.util.matrix.Matrix, jline.util.matrix.Matrix, jline.util.matrix.Matrix[], double, int, double[])), which differs from RANDOM(m) even though the fixed points agree. Reduces to RANDOM(m) when m_1 = ... = m_{h-1} = 1.

Reference: N. Gast and B. Van Houdt, "Transient and Steady-state Regime of a Family of List-based Cache Replacement Algorithms", ACM SIGMETRICS 2015.

Method Summary

Modifier and Type	Method	Description
static CacheMissFpiResult	cache_miss_fifo_rmf(Matrix gamma, Matrix m, Matrix[] lambdaCache)	Steady-state FIFO(m) miss rates via the position-resolved mean field.
static CacheMissFpiResult	cache_miss_fifo_rmf(Matrix gamma, Matrix m, Matrix[] lambdaCache, Matrix[][] accost)	FIFO(m) miss rates honouring a custom access graph.
static CacheMissRmfTranResult	cache_miss_fifo_rmf_tran(Matrix gamma, Matrix m, Matrix[] lambdaCache, double time, int nPoints, double[] xinit)	Transient FIFO(m) cache trajectory via the position-resolved mean field.
static CacheMissRmfTranResult	cache_miss_fifo_rmf_tran(Matrix gamma, Matrix m, Matrix[] lambdaCache, double time, int nPoints, double[] xinit, Matrix[][] accost)	Transient FIFO(m) cache trajectory honouring a custom access graph.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

cache_miss_fifo_rmf

public static CacheMissFpiResult cache_miss_fifo_rmf(Matrix gamma,
 Matrix m,
 Matrix[] lambdaCache)

Steady-state FIFO(m) miss rates via the position-resolved mean field. Note FIFO(m) equals RANDOM(m) at steady state (Gast15 Thm 1); prefer Cache_miss_rmf.cache_miss_rmf(jline.util.matrix.Matrix, jline.util.matrix.Matrix, jline.util.matrix.Matrix[]) when only the fixed point is needed.

cache_miss_fifo_rmf

public static CacheMissFpiResult cache_miss_fifo_rmf(Matrix gamma,
 Matrix m,
 Matrix[] lambdaCache,
 Matrix[][] accost)

FIFO(m) miss rates honouring a custom access graph. A non-linear accost uses the general position-resolved drift from a cold cache (see CachePosGraphRMF); the linear default keeps the pre-filled path.

cache_miss_fifo_rmf_tran

public static CacheMissRmfTranResult cache_miss_fifo_rmf_tran(Matrix gamma,
 Matrix m,
 Matrix[] lambdaCache,
 double time,
 int nPoints,
 double[] xinit)

Transient FIFO(m) cache trajectory via the position-resolved mean field. Mirrors Cache_miss_rmf.cache_miss_rmf_tran(jline.util.matrix.Matrix, jline.util.matrix.Matrix, jline.util.matrix.Matrix[], double, int, double[]); the FIFO transient differs from RANDOM(m) even though the steady states agree.

Parameters:

time - end time of the transient window (start is 0).

nPoints - number of uniform grid points (>= 2).

xinit - initial occupancy (dim,), or null for the default warm start.

cache_miss_fifo_rmf_tran

public static CacheMissRmfTranResult cache_miss_fifo_rmf_tran(Matrix gamma,
 Matrix m,
 Matrix[] lambdaCache,
 double time,
 int nPoints,
 double[] xinit,
 Matrix[][] accost)

Transient FIFO(m) cache trajectory honouring a custom access graph. A non-linear accost uses the general position-resolved drift (cold start when xinit is null); the linear default keeps the FIFO drift.