3.8. List of queues¶
Available queues are as follows.
Note
The numbers in () represent the default values.
3.8.1. HPE Cray XD220v¶
3.8.1.1. Node Configuration¶
Node type |
Number of nodes |
CPU |
GPU |
Memory |
|---|---|---|---|---|
XD220v general |
120 |
Xeon Platinum 8480+ x 2 = 112 Cores |
- |
512 GiB |
XD220v large memory |
7 |
Xeon Platinum 8480+ x 2 = 112 Cores |
- |
2.0 TiB |
3.8.1.2. Exclusive Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
P_030 |
30(1) |
13800(460) |
72(24) |
3360(112) |
0 |
For batch job |
TP_002 |
2(1) |
920(460) |
168(24) |
224(112) |
0 |
For long time job |
MP_001 |
1(1) |
1840(1840) |
72(24) |
112(112) |
0 |
For large memory |
Attention
Consumed node time＝Run time(H)×Number of used nodes×1.0
3.8.1.4. Application Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
TP_030 |
30(1) |
13800(460) |
168(24) |
3360(112) |
0 |
For application |
MP_007 |
7(1) |
12880(1840) |
72(24) |
784(112) |
0 |
For application |
HP_127 |
127(127) |
68080(68080) |
24(24) |
14224(14224) |
0 |
For ultra-large-scale HPC challenge |
HP_129 |
129(129) |
75760(75760) |
24(24) |
16144(16144) |
0 |
For ultra-large-scale HPC challenge |
Attention
Consumed node time＝Run time(H)×Number of used nodes×1.0
3.8.2. HPE Scale-up Server 3200¶
3.8.2.1. Node Configuration¶
Node type |
number of nodes |
CPU |
GPU |
Memory |
|---|---|---|---|---|
Scale-up Server 3200 |
2 |
Xeon Platinum 8490H x 16 = 960 Cores |
- |
4.0 TiB |
3.8.2.2. Exclusive Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
S_001 |
1(1) |
3840(3840) |
72(24) |
960(960) |
0 |
For large job |
Attention
Consumed node time＝Run time(H)×Number of used nodes×8.0
3.8.2.3. Shared Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
CS_001 |
1(1) |
1920(960) |
72(24) |
480(240) |
0 |
For large job |
Attention
Consumed node time＝Run time(H)×Number of used cores/960×8.0
3.8.3. HPE Cray XD670¶
3.8.3.1. Node Configuration¶
Node type |
Number of nodes |
CPU |
GPU |
Memory |
|---|---|---|---|---|
XD670 |
11 |
Xeon Platinum 8480+ x 2 = 112 Cores |
NVIDIA H100 x 8 |
1.0 TiB |
3.8.3.2. Exclusive Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
A_002 |
2(1) |
1840(920) |
24(2) |
224(112) |
16(8) |
For batch job |
Attention
Consumed node time＝Run time(H)×Number of used nodes×8.0
3.8.3.3. Shared Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
CA_001 |
1(1) |
460(8) |
72(24) |
56(1) |
4(1) |
For small job |
DA_002 |
2(1) |
32(8) |
2(2) |
4(1) |
4(1) |
For debug |
Attention
Consumed node time＝Run time(H)×Number of used GPUs×1.0
3.8.3.4. Application Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
TA_004 |
4(1) |
3680(920) |
168(24) |
448(112) |
32(8) |
For application |
HA_011 |
11(1) |
10120(920) |
24(24) |
1232(112) |
88(88) |
For ultra-large-scale HPC challenge |
Attention
Consumed node time＝Run time(H)×Number of used nodes×1.0
3.8.4. Front end node¶
3.8.4.1. Node Configuration¶
Node type |
Number of nodes |
CPU |
GPU |
Memory |
|---|---|---|---|---|
super |
3 |
Xeon Gold 6430 x 2 = 64 Cores |
- |
256 GiB |
gpu |
2 |
Xeon Gold 6430 x 2 = 64 Cores |
NVIDIA H100 x 1 |
256 GiB |
3.8.4.2. Shared Queue¶
Queue name |
Limit of number of nodes |
Memory Limit [GB] |
Walltime limit |
Limit of concurrent execution |
Upper Limit of using GPU |
Notes |
|---|---|---|---|---|---|---|
IP_001 |
1(1) |
8(4) |
8(1) |
4(1) |
0 |
For interactive mode |
IA_001 |
1(1) |
8(4) |
8(1) |
4(1) |
1 |
For interactive mode (GPU reference available) |
Attention
Consumed node time＝Run time(H)×Number of used cores/64×1.0
3.8.5. Usage of Application Queue¶
We support researchers who want to develop application software for ultra-large-scale calculations on MASAMUNE-II and a new methodology for Computational Materials Science. We also offer job queues for these calculations and you can use them by submitting your application. As for the application, please contact us .