Attention

This documentation is under active development, meaning that it can change over time as we refine it. Please email help@massive.org.au if you require assistance.

About M3#

M3 is the third stage of MASSIVE.

MASSIVE is a High Performance Computing (HPC) facility designed specifically to process complex data. Since 2010, MASSIVE has played a key role in driving discoveries across many disciplines including biomedical sciences, materials research, engineering and geosciences.

MASSIVE is pioneering and building high performance computing upon Monash’s specialist Research Cloud fabric. M3 has been supplied by Dell with a Mellanox low latency network and NVIDIA GPUs.

System configuration#

  • Total number of cores: 5673

  • Total number of GPU cores: 1,752,576 CUDA cores

  • Total memory: 63.2 TBytes

  • Storage capacity: 5.6 PBytes Lustre parallel file system

  • Network: RoCE v2 via mixed speed Mellanox ConnectX cards

M3 utilises SLURM scheduler to manage the partitions available to users, with different configurations to suit a variety of computational requirements. Details about the partitions can be found here.

M3 provides the following configurations:

Compute#

Name

Partition

Nodes

CPU Cores per node

Total CPU cores in partition

Memory per node (GB)

CPU Speed

Processor model

Codename

High-Density CPUs

m3d

6

48

288

790

2.20 GHz

Intel Xeon Gold 6150

Cascade Lake

High-Density CPUs

m3i

47

36

1,692

171

2.70 GHz

Intel Xeon Gold 6150

Skylake

High-Density CPUs with High Memory

m3j

11

36

396

357

2.70 GHz

Intel Xeon Gold 6150

Skylake

High-Density CPUs with Extra High Memory

m3m

1

36

36

1,024

2.70 GHz

Intel Xeon Gold 6150

Skylake

Short Jobs

short

3

27

81

128

2.70 GHz

Intel Xeon Gold 6150

Skylake

GPU Compute#

Name

Partition

Nodes

CPU Cores per node

Total CPU cores in partition

Memory per node (GB)

Number of GPUs per node

Total GPUs in partition

GPU model

GPU cores per card

CPU Speed

Processor model

Codename

K80 with High-Density GPUs

m3e

1

24

24

256

8

8

NVIDIA Tesla K80

4,992

2.50GHz

Intel Xeon E5-2680 v3

Haswell

T4

gpu

2

52

104

1055

8

16

NVIDIA Tesla T4

2,560

2.10GHz

Intel Xeon Gold 6230R

Cascade Lake

P100

m3h

7

28

196

256

2

14

NVIDIA Tesla P100

3,584

2.40GHz

Intel Xeon E5-2680 v4

Broadwell

V100

m3g

24

36

864

384

3

72

NVIDIA Tesla V100

5,120

2.70GHz

Intel Xeon Gold 6150

Skylake

A40

gpu

4

52

312

1056

4

12

NVIDIA A40

10,752

2.20GHz

Intel Xeon Gold 5320

Ice Lake

DGX1

dgx

11

40

440

512

8

88

NVIDIA Tesla V100 32GB

5,120

2.20GHz

Intel Xeon E5-2698 v4

Broadwell

Remote desktops through the CVL#

Name

Partition

Desktop

CPU Cores per desktop

Total CPU cores in partition

Memory per desktop (GB)

Number of GPUs per desktop

Total GPUs in partition

GPU model

GPU cores per card

CPU Speed

Processor model

Codename

K1- Visualisation and Job Submission

m3f

32

3

96

13

1

32

NVIDIA Grid K1

192

2.50GHz

Intel Xeon E5-2680 v3

Haswell

P4 - Light Compute

m3p

60

6

396

55

1

60

NVIDIA Tesla P4

2,560

2.70GHz

Intel Xeon Gold 6150

Skylake

Single T4 - Light Compute

desktop

<32

6

192

100

1

32

NVIDIA Tesla T4

2560

2.10GHz

Intel Xeon Gold 6230R

Cascade Lake

Dual T4 - Heavy Compute

desktop

<16

13

192

225

2

32

NVIDIA Tesla T4

2560

2.10GHz

Intel Xeon Gold 6230R

Cascade Lake

K80 - Heavy Compute

m3c

26

12

312

128

2

52

NVIDIA Tesla K80

4,992

2.50GHz

Intel Xeon E5-2680 v3

Haswell

A40 - Heavy Compute

desktop

16

13

208

250

1

16

NVIDIA A40

10,752

2.20GHz

Intel Xeon Gold 5320

Ice Lake