lamp-sten
API
Generated with
lamp-sten/lamp/CudaDevice

CudaDevice

lamp.CudaDevice
case class CudaDevice(i: Int) extends Device

Attributes

Graph
Supertypes
trait Serializable
trait Product
trait Equals
trait Device
class Object
trait Matchable
class Any
Show all

Members list

Value members

Concrete methods

def getCurrentStream: CudaStream

Effective for the current OS thread

Effective for the current OS thread

Attributes

def getStreamFromPool: CudaStream
def measureTime[A](f: => A): (A, Long)
def setCurrentStream(s: CudaStream): Unit

Effective for the current OS thread

Effective for the current OS thread

Attributes

def setSeed(seed: Long): Unit
def to[S : Sc](t: STenOptions): STenOptions
def to(t: Tensor): Tensor
override def withOtherStream[A](synchronizeBefore: Boolean, synchronizeAfter: Boolean)(f: => A): A

Executes f on a new stream

Executes f on a new stream

f must not switch to other threads

Restores the stream to the original stream Optionally synchronizes the host before and/or after f

Attributes

Definition Classes

Inherited methods

def allocateBuffers[S : Sc](size: Long, options: STenOptions): BufferPair

Attributes

Inherited from:
Device
def productElementNames: Iterator[String]

Attributes

Inherited from:
Product
def productIterator: Iterator[Any]

Attributes

Inherited from:
Product
def to[S : Sc](t: STen): STen

Attributes

Inherited from:
Device
def toBatched[S : Sc](tensors: Seq[STen], buffers: BufferPair): Seq[STen]

Copies tensors to this device in a single cross device copy. Data is copied via a buffer pair which consists of a source and a destinatin buffer. srcBuffer is supposed to be on the source device. dstBuffer has to be on this device. Tensors are first copied to the srcBuffer, then the srcBuffer is copied to dstBuffer, then the dstBuffer is split into views.

Copies tensors to this device in a single cross device copy. Data is copied via a buffer pair which consists of a source and a destinatin buffer. srcBuffer is supposed to be on the source device. dstBuffer has to be on this device. Tensors are first copied to the srcBuffer, then the srcBuffer is copied to dstBuffer, then the dstBuffer is split into views.

All tensors must have the same data type.

Might make sense to pin the srcBuffer.

Attributes

Inherited from:
Device
Generated with