8GPU: A6000 の nvidia-smi と deviceQuery
July 25, 2021
41
HPCDIY-ERMGPU8R4S（こちら）に A6000 を 8 枚実装して、nvidia-smi と deviceQuery を実行してみました。
nvidia-smi
dl@dl-machine:~$ nvidia-smi
Mon Jul 5 10:22:47 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.80 Driver Version: 460.80 CUDA Version: 11.2 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 RTX A6000 Off | 00000000:01:00.0 Off | Off |
| 30% 34C P8 18W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 1 RTX A6000 Off | 00000000:25:00.0 Off | Off |
| 30% 33C P8 19W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 2 RTX A6000 Off | 00000000:41:00.0 Off | Off |
| 30% 34C P8 21W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 3 RTX A6000 Off | 00000000:61:00.0 Off | Off |
| 30% 32C P8 17W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 4 RTX A6000 Off | 00000000:81:00.0 Off | Off |
| 30% 34C P8 20W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 5 RTX A6000 Off | 00000000:A1:00.0 Off | Off |
| 30% 32C P8 20W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 6 RTX A6000 Off | 00000000:C1:00.0 Off | Off |
| 30% 33C P8 23W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 7 RTX A6000 Off | 00000000:E1:00.0 Off | Off |
| 30% 32C P8 21W / 300W | 5MiB / 48685MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 1 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 2 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 3 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 4 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 5 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 6 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
| 7 N/A N/A 1880 G /usr/lib/xorg/Xorg 4MiB |
+-----------------------------------------------------------------------------+
deviceQuery
最後に表示される Peer access に注目してください。これが No だと、GPUが多い場合（>4くらい）並列性能が出ないと言われています。
ちなみに GeForce RTX 2080 ti, Titan RTX は No
Quadro, Tesla は Yes
dl@dl-machine:~$ deviceQuery
deviceQuery Starting...
CUDA Device Query (Runtime API) version (CUDART static linking)
Detected 8 CUDA Capable device(s)
Device 0: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 1 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 1: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 37 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 2: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 65 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 3: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 97 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 4: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 129 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 5: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 161 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 6: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 193 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
Device 7: "RTX A6000"
CUDA Driver Version / Runtime Version 11.2 / 11.1
CUDA Capability Major/Minor version number: 8.6
Total amount of global memory: 48685 MBytes (51050315776 bytes)
(84) Multiprocessors, (128) CUDA Cores/MP: 10752 CUDA Cores
GPU Max Clock rate: 1800 MHz (1.80 GHz)
Memory Clock rate: 8001 Mhz
Memory Bus Width: 384-bit
L2 Cache Size: 6291456 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 102400 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1536
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 225 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU6) : Yes
> Peer access from RTX A6000 (GPU0) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU6) : Yes
> Peer access from RTX A6000 (GPU1) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU6) : Yes
> Peer access from RTX A6000 (GPU2) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU6) : Yes
> Peer access from RTX A6000 (GPU3) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU6) : Yes
> Peer access from RTX A6000 (GPU4) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU6) : Yes
> Peer access from RTX A6000 (GPU5) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU6) -> RTX A6000 (GPU7) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU0) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU1) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU2) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU3) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU4) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU5) : Yes
> Peer access from RTX A6000 (GPU7) -> RTX A6000 (GPU6) : Yes
deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.2, CUDA Runtime Version = 11.1, NumDevs = 8
Result = PASS
Share on Facebook
Share on X
Share on Pinterest
0 Comment(s)
Recent Posts
