hipdeviceprop_t Type Reference#
Public Attributes | |
| character(kind=c_char), dimension(256) | name |
| Device name. More... | |
| integer(c_size_t) | totalglobalmem |
| Size of global memory region (in bytes). More... | |
| integer(c_size_t) | sharedmemperblock |
| Size of shared memory region (in bytes). More... | |
| integer(c_int) | regsperblock |
| Registers per block. More... | |
| integer(c_int) | warpsize |
| Warp size. More... | |
| integer(c_int) | maxthreadsperblock |
| Max work items per work group or workgroup max size. More... | |
| integer(c_int), dimension(3) | maxthreadsdim |
| Max number of threads in each dimension (XYZ) of a block. More... | |
| integer(c_int), dimension(3) | maxgridsize |
| Max grid dimensions (XYZ). More... | |
| integer(c_int) | clockrate |
| Max clock frequency of the multiProcessors in khz. More... | |
| integer(c_int) | memoryclockrate |
| Max global memory clock frequency in khz. More... | |
| integer(c_int) | memorybuswidth |
| Global memory bus width in bits. More... | |
| integer(c_size_t) | totalconstmem |
| Size of shared memory region (in bytes). More... | |
| integer(c_int) | major |
| Major compute capability. More... | |
| integer(c_int) | minor |
| Minor compute capability. More... | |
| integer(c_int) | multiprocessorcount |
| Number of multi-processors (compute units). More... | |
| integer(c_int) | l2cachesize |
| L2 cache size. More... | |
| integer(c_int) | maxthreadspermultiprocessor |
| Maximum resident threads per multi-processor. More... | |
| integer(c_int) | computemode |
| Compute mode. More... | |
| integer(c_int) | clockinstructionrate |
| Frequency in khz of the timer used by the device-side "clock*" instructions. More... | |
| integer(c_int) | arch |
| Architectural feature flags. More... | |
| integer(c_int) | concurrentkernels |
| Device can possibly execute multiple kernels concurrently. More... | |
| integer(c_int) | pcidomainid |
| PCI Domain ID. More... | |
| integer(c_int) | pcibusid |
| PCI Bus ID. More... | |
| integer(c_int) | pcideviceid |
| PCI Device ID. More... | |
| integer(c_size_t) | maxsharedmemorypermultiprocessor |
| Maximum Shared Memory Per Multiprocessor. More... | |
| integer(c_int) | ismultigpuboard |
| 1 if device is on a multi-GPU board, 0 if not. More... | |
| integer(c_int) | canmaphostmemory |
| Check whether HIP can map host memory. More... | |
| integer(c_int) | gcnarch |
| DEPRECATED: use gcnArchName instead. More... | |
| character(kind=c_char), dimension(256) | gcnarchname |
| AMD GCN Arch Name. More... | |
| integer(c_int) | integrated |
| APU vs dGPU. More... | |
| integer(c_int) | cooperativelaunch |
| HIP device supports cooperative launch. More... | |
| integer(c_int) | cooperativemultidevicelaunch |
| HIP device supports cooperative launch on multiple devices. More... | |
| integer(c_int) | maxtexture1dlinear |
| Maximum size for 1D textures bound to linear memory. More... | |
| integer(c_int) | maxtexture1d |
| Maximum number of elements in 1D images. More... | |
| integer(c_int), dimension(2) | maxtexture2d |
| Maximum dimensions (width, height) of 2D images, in image elements. More... | |
| integer(c_int), dimension(3) | maxtexture3d |
| Maximum dimensions (width, height, depth) of 3D images, in image elements. More... | |
| type(c_ptr) | hdpmemflushcntl |
| Addres of HDP_MEM_COHERENCY_FLUSH_CNTL register. More... | |
| type(c_ptr) | hdpregflushcntl |
| Addres of HDP_REG_COHERENCY_FLUSH_CNTL register. More... | |
| integer(c_size_t) | mempitch |
| Maximum pitch in bytes allowed by memory copies. More... | |
| integer(c_size_t) | texturealignment |
| Alignment requirement for textures. More... | |
| integer(c_size_t) | texturepitchalignment |
| Pitch alignment requirement for texture references bound to pitched memory. More... | |
| integer(c_int) | kernelexectimeoutenabled |
| Run time limit for kernels executed on the device. More... | |
| integer(c_int) | eccenabled |
| Device has ECC support enabled. More... | |
| integer(c_int) | tccdriver |
| 1:If device is Tesla device using TCC driver, else 0 More... | |
| integer(c_int) | cooperativemultideviceunmatchedfunc |
| HIP device supports cooperative launch on multiple. More... | |
| integer(c_int) | cooperativemultideviceunmatchedgriddim |
| HIP device supports cooperative launch on multiple. More... | |
| integer(c_int) | cooperativemultideviceunmatchedblockdim |
| HIP device supports cooperative launch on multiple. More... | |
| integer(c_int) | cooperativemultideviceunmatchedsharedmem |
| HIP device supports cooperative launch on multiple. More... | |
| integer(c_int) | islargebar |
| 1: if it is a large PCI bar device, else 0 More... | |
| integer(c_int) | asicrevision |
| Revision of the GPU in this device. More... | |
| integer(c_int) | managedmemory |
| Device supports allocating managed memory on this system. More... | |
| integer(c_int) | directmanagedmemaccessfromhost |
| Host can directly access managed memory on the device without migration. More... | |
| integer(c_int) | concurrentmanagedaccess |
| Device can coherently access managed memory concurrently with the CPU. More... | |
| integer(c_int) | pageablememoryaccess |
| Device supports coherently accessing pageable memory without calling hipHostRegister on it. More... | |
| integer(c_int) | pageablememoryaccessuseshostpagetables |
| Device accesses pageable memory via the host's page tables. More... | |
| character(kind=c_char), dimension(256) | gpufort_padding |
| GPUFORT :Some extra bytes to prevent seg faults in newer versions. More... | |
Member Data Documentation
◆ arch
| integer(c_int) hipfort_types::hipdeviceprop_t::arch |
Architectural feature flags.
New for HIP.
◆ asicrevision
| integer(c_int) hipfort_types::hipdeviceprop_t::asicrevision |
Revision of the GPU in this device.
◆ canmaphostmemory
| integer(c_int) hipfort_types::hipdeviceprop_t::canmaphostmemory |
Check whether HIP can map host memory.
◆ clockinstructionrate
| integer(c_int) hipfort_types::hipdeviceprop_t::clockinstructionrate |
Frequency in khz of the timer used by the device-side "clock*" instructions.
New for HIP.
◆ clockrate
| integer(c_int) hipfort_types::hipdeviceprop_t::clockrate |
Max clock frequency of the multiProcessors in khz.
◆ computemode
| integer(c_int) hipfort_types::hipdeviceprop_t::computemode |
Compute mode.
◆ concurrentkernels
| integer(c_int) hipfort_types::hipdeviceprop_t::concurrentkernels |
Device can possibly execute multiple kernels concurrently.
◆ concurrentmanagedaccess
| integer(c_int) hipfort_types::hipdeviceprop_t::concurrentmanagedaccess |
Device can coherently access managed memory concurrently with the CPU.
◆ cooperativelaunch
| integer(c_int) hipfort_types::hipdeviceprop_t::cooperativelaunch |
HIP device supports cooperative launch.
◆ cooperativemultidevicelaunch
| integer(c_int) hipfort_types::hipdeviceprop_t::cooperativemultidevicelaunch |
HIP device supports cooperative launch on multiple devices.
◆ cooperativemultideviceunmatchedblockdim
| integer(c_int) hipfort_types::hipdeviceprop_t::cooperativemultideviceunmatchedblockdim |
HIP device supports cooperative launch on multiple.
◆ cooperativemultideviceunmatchedfunc
| integer(c_int) hipfort_types::hipdeviceprop_t::cooperativemultideviceunmatchedfunc |
HIP device supports cooperative launch on multiple.
◆ cooperativemultideviceunmatchedgriddim
| integer(c_int) hipfort_types::hipdeviceprop_t::cooperativemultideviceunmatchedgriddim |
HIP device supports cooperative launch on multiple.
◆ cooperativemultideviceunmatchedsharedmem
| integer(c_int) hipfort_types::hipdeviceprop_t::cooperativemultideviceunmatchedsharedmem |
HIP device supports cooperative launch on multiple.
◆ directmanagedmemaccessfromhost
| integer(c_int) hipfort_types::hipdeviceprop_t::directmanagedmemaccessfromhost |
Host can directly access managed memory on the device without migration.
◆ eccenabled
| integer(c_int) hipfort_types::hipdeviceprop_t::eccenabled |
Device has ECC support enabled.
◆ gcnarch
| integer(c_int) hipfort_types::hipdeviceprop_t::gcnarch |
DEPRECATED: use gcnArchName instead.
◆ gcnarchname
| character(kind=c_char), dimension(256) hipfort_types::hipdeviceprop_t::gcnarchname |
AMD GCN Arch Name.
◆ gpufort_padding
| character(kind=c_char), dimension(256) hipfort_types::hipdeviceprop_t::gpufort_padding |
GPUFORT :Some extra bytes to prevent seg faults in newer versions.
◆ hdpmemflushcntl
| type(c_ptr) hipfort_types::hipdeviceprop_t::hdpmemflushcntl |
Addres of HDP_MEM_COHERENCY_FLUSH_CNTL register.
◆ hdpregflushcntl
| type(c_ptr) hipfort_types::hipdeviceprop_t::hdpregflushcntl |
Addres of HDP_REG_COHERENCY_FLUSH_CNTL register.
◆ integrated
| integer(c_int) hipfort_types::hipdeviceprop_t::integrated |
APU vs dGPU.
◆ islargebar
| integer(c_int) hipfort_types::hipdeviceprop_t::islargebar |
1: if it is a large PCI bar device, else 0
◆ ismultigpuboard
| integer(c_int) hipfort_types::hipdeviceprop_t::ismultigpuboard |
1 if device is on a multi-GPU board, 0 if not.
◆ kernelexectimeoutenabled
| integer(c_int) hipfort_types::hipdeviceprop_t::kernelexectimeoutenabled |
Run time limit for kernels executed on the device.
◆ l2cachesize
| integer(c_int) hipfort_types::hipdeviceprop_t::l2cachesize |
L2 cache size.
◆ major
| integer(c_int) hipfort_types::hipdeviceprop_t::major |
Major compute capability.
On HCC, this is an approximation and features may differ from CUDA CC. See the arch feature flags for portable ways to query feature caps.
◆ managedmemory
| integer(c_int) hipfort_types::hipdeviceprop_t::managedmemory |
Device supports allocating managed memory on this system.
◆ maxgridsize
| integer(c_int), dimension(3) hipfort_types::hipdeviceprop_t::maxgridsize |
Max grid dimensions (XYZ).
◆ maxsharedmemorypermultiprocessor
| integer(c_size_t) hipfort_types::hipdeviceprop_t::maxsharedmemorypermultiprocessor |
Maximum Shared Memory Per Multiprocessor.
◆ maxtexture1d
| integer(c_int) hipfort_types::hipdeviceprop_t::maxtexture1d |
Maximum number of elements in 1D images.
◆ maxtexture1dlinear
| integer(c_int) hipfort_types::hipdeviceprop_t::maxtexture1dlinear |
Maximum size for 1D textures bound to linear memory.
◆ maxtexture2d
| integer(c_int), dimension(2) hipfort_types::hipdeviceprop_t::maxtexture2d |
Maximum dimensions (width, height) of 2D images, in image elements.
◆ maxtexture3d
| integer(c_int), dimension(3) hipfort_types::hipdeviceprop_t::maxtexture3d |
Maximum dimensions (width, height, depth) of 3D images, in image elements.
◆ maxthreadsdim
| integer(c_int), dimension(3) hipfort_types::hipdeviceprop_t::maxthreadsdim |
Max number of threads in each dimension (XYZ) of a block.
◆ maxthreadsperblock
| integer(c_int) hipfort_types::hipdeviceprop_t::maxthreadsperblock |
Max work items per work group or workgroup max size.
◆ maxthreadspermultiprocessor
| integer(c_int) hipfort_types::hipdeviceprop_t::maxthreadspermultiprocessor |
Maximum resident threads per multi-processor.
◆ memorybuswidth
| integer(c_int) hipfort_types::hipdeviceprop_t::memorybuswidth |
Global memory bus width in bits.
◆ memoryclockrate
| integer(c_int) hipfort_types::hipdeviceprop_t::memoryclockrate |
Max global memory clock frequency in khz.
◆ mempitch
| integer(c_size_t) hipfort_types::hipdeviceprop_t::mempitch |
Maximum pitch in bytes allowed by memory copies.
◆ minor
| integer(c_int) hipfort_types::hipdeviceprop_t::minor |
Minor compute capability.
On HCC, this is an approximation and features may differ from CUDA CC. See the arch feature flags for portable ways to query feature caps.
◆ multiprocessorcount
| integer(c_int) hipfort_types::hipdeviceprop_t::multiprocessorcount |
Number of multi-processors (compute units).
◆ name
| character(kind=c_char), dimension(256) hipfort_types::hipdeviceprop_t::name |
Device name.
◆ pageablememoryaccess
| integer(c_int) hipfort_types::hipdeviceprop_t::pageablememoryaccess |
Device supports coherently accessing pageable memory without calling hipHostRegister on it.
◆ pageablememoryaccessuseshostpagetables
| integer(c_int) hipfort_types::hipdeviceprop_t::pageablememoryaccessuseshostpagetables |
Device accesses pageable memory via the host's page tables.
◆ pcibusid
| integer(c_int) hipfort_types::hipdeviceprop_t::pcibusid |
PCI Bus ID.
◆ pcideviceid
| integer(c_int) hipfort_types::hipdeviceprop_t::pcideviceid |
PCI Device ID.
◆ pcidomainid
| integer(c_int) hipfort_types::hipdeviceprop_t::pcidomainid |
PCI Domain ID.
◆ regsperblock
| integer(c_int) hipfort_types::hipdeviceprop_t::regsperblock |
Registers per block.
◆ sharedmemperblock
| integer(c_size_t) hipfort_types::hipdeviceprop_t::sharedmemperblock |
Size of shared memory region (in bytes).
◆ tccdriver
| integer(c_int) hipfort_types::hipdeviceprop_t::tccdriver |
1:If device is Tesla device using TCC driver, else 0
◆ texturealignment
| integer(c_size_t) hipfort_types::hipdeviceprop_t::texturealignment |
Alignment requirement for textures.
◆ texturepitchalignment
| integer(c_size_t) hipfort_types::hipdeviceprop_t::texturepitchalignment |
Pitch alignment requirement for texture references bound to pitched memory.
◆ totalconstmem
| integer(c_size_t) hipfort_types::hipdeviceprop_t::totalconstmem |
Size of shared memory region (in bytes).
◆ totalglobalmem
| integer(c_size_t) hipfort_types::hipdeviceprop_t::totalglobalmem |
Size of global memory region (in bytes).
◆ warpsize
| integer(c_int) hipfort_types::hipdeviceprop_t::warpsize |
Warp size.
The documentation for this type was generated from the following file: