Skip to content

Commit 5dd6f47

Browse files
mattmcinnesmattmcinnes
committed
Fixing missing files
1 parent 101eb73 commit 5dd6f47

File tree

8 files changed

+276
-22
lines changed

8 files changed

+276
-22
lines changed

‎articles/virtual-machines/TOC.yml

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -457,7 +457,7 @@
457457
- name: Ncv3 NC6s NC12s NC24s retirement
458458
href: ncv3-nc6s-nc12s-nc24s-retirement.md
459459
- name: NCasT4_v3 series
460-
href: ./sizes/gpu-accelerated/nct4v3-series.md
460+
href: ./sizes/gpu-accelerated/ncast4v3-series.md
461461
- name: NC_A100_v4 series
462462
href: ./sizes/gpu-accelerated/nca100v4-series.md
463463
- name: ND family
@@ -494,15 +494,15 @@
494494
- name: NV series
495495
href: ./sizes/gpu-accelerated/nv-series.md
496496
- name: NV series retirement
497-
href: ./sizes/gpu-accelerated/nv-series-retirement.md
497+
href: nv-series-retirement.md
498498
- name: NV series migration guide
499-
href: ./sizes/gpu-accelerated/nv-series-migration-guide.md
499+
href: nv-series-migration-guide.md
500500
- name: NVv3 series
501501
href: ./sizes/gpu-accelerated/nvv3-series.md
502502
- name: NVv4 series
503503
href: ./sizes/gpu-accelerated/nvv4-series.md
504504
- name: NVadsA10_v5 series
505-
href: ./sizes/gpu-accelerated/nva10v5-series.md
505+
href: ./sizes/gpu-accelerated/nvadsa10v5-series.md
506506
- name: Setup NVIDIA GPU drivers
507507
items:
508508
- name: Linux

‎articles/virtual-machines/sizes/gpu-accelerated/includes/ndh100v5-series-specs.md

Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,20 @@
1+
---
2+
title: ND-H100-v5 series specs include
3+
description: Include file containing specifications of ND-H100-v5-series VM sizes.
4+
author: mattmcinnes
5+
ms.topic: include
6+
ms.service: azure-virtual-machines
7+
ms.subservice: sizes
8+
ms.date: 08/01/2024
9+
ms.author: mattmcinnes
10+
ms.reviewer: mattmcinnes
11+
ms.custom: include file
12+
---
13+
| Part | Quantity <br><sup>Count Units | Specs <br><sup>SKU ID, Performance Units, etc. |
14+
|---|---|---|
15+
| Processor | 96 vCPUs | Intel Xeon (Sapphire Rapids) [x86-64] |
16+
| Memory | 1900 GiB | |
17+
| Local Storage | 1 Disk | 28000 GiB |
18+
| Remote Storage | 32Disks | |
19+
| Network | 8 NICs | |
20+
| Accelerators | 8 GPUs | Nvidia PCIe H100 GPU (80GB) |

‎articles/virtual-machines/sizes/gpu-accelerated/includes/ndh100v5-series-summary.md

Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,19 @@
1+
---
2+
title: ND-H100-v5-series summary include file
3+
description: Include file for ND-H100-v5-series summary
4+
author: mattmcinnes
5+
ms.topic: include
6+
ms.service: virtual-machines
7+
ms.subservice: sizes
8+
ms.date: 08/01/2024
9+
ms.author: mattmcinnes
10+
ms.reviewer: mattmcinnes
11+
ms.custom: include file
12+
---
13+
The ND H100 v5 series virtual machine (VM) is a new flagship addition to the Azure GPU family. It’s designed for high-end Deep Learning training and tightly coupled scale-up and scale-out Generative AI and HPC workloads.
14+
15+
The ND H100 v5 series starts with a single VM and eight NVIDIA H100 Tensor Core GPUs. ND H100 v5-based deployments can scale up to thousands of GPUs with 3.2Tb/s of interconnect bandwidth per VM. Each GPU within the VM is provided with its own dedicated, topology-agnostic 400 Gb/s NVIDIA Quantum-2 CX7 InfiniBand connection. These connections are automatically configured between VMs occupying the same virtual machine scale set, and support GPUDirect RDMA.
16+
17+
Each GPU features NVLINK 4.0 connectivity for communication within the VM, and the instance is backed by 96 physical 4th Gen Intel Xeon Scalable processor cores.
18+
19+
These instances provide excellent performance for many AI, ML, and analytics tools that support GPU acceleration ‘out-of-the-box,’ such as TensorFlow, Pytorch, Caffe, RAPIDS, and other frameworks. Additionally, the scale-out InfiniBand interconnect is supported by a large set of existing AI and HPC tools that are built on NVIDIA’s NCCL communication libraries for seamless clustering of GPUs.

‎articles/virtual-machines/sizes/gpu-accelerated/includes/ndv2-series-specs.md

Lines changed: 11 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -1,19 +1,20 @@
11
---
2-
title: NDv2-series specs include
2+
title: NDv2 series specs include
33
description: Include file containing specifications of NDv2-series VM sizes.
4-
services: virtual-machines
54
author: mattmcinnes
65
ms.topic: include
7-
ms.service: virtual-machines
6+
ms.service: azure-virtual-machines
87
ms.subservice: sizes
9-
ms.date: 04/18/2024
8+
ms.date: 08/01/2024
109
ms.author: mattmcinnes
10+
ms.reviewer: mattmcinnes
1111
ms.custom: include file
1212
---
13-
| Part | Quantity <br><sup>Count <sup>Units | Specs <br><sup>SKU ID, Performance <sup>Units</sup>, etc. |
13+
| Part | Quantity <br><sup>Count Units | Specs <br><sup>SKU ID, Performance Units, etc. |
1414
|---|---|---|
15-
| Processor | 40<sup>vCores | Intel® Xeon® Platinum 8168 (Skylake) |
16-
| Memory | 672<sup>GiB | |
17-
| Data Disks | 32<sup>Disks | 80000<sup>IOPS</sup> / 800<sup>MBps |
18-
| Network | 8<sup>NICs | 24000<sup>Mbps |
19-
| Accelerators | 8<sup>GPUs</sup> | NVIDIA V100 (NVLink) 32<sup>GiB </sup> <br> 256<sup>GiB</sup> per VM|
15+
| Processor | 40 vCPUs | Intel Xeon Platinum 8168 (Skylake) [x86-64] |
16+
| Memory | 672 GiB | |
17+
| Local Storage | 1 Disk | 2948 GiB |
18+
| Remote Storage | 32 Disks | 80000 IOPS <br>800 MBps |
19+
| Network | 8 NICs | 24000 Mbps |
20+
| Accelerators | None | |

‎articles/virtual-machines/sizes/gpu-accelerated/includes/ndv2-series-summary.md

Lines changed: 11 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -1,13 +1,19 @@
11
---
2-
title: NDv2-series summary include
3-
description: Include file containing a summary of the NDv2-series size family.
4-
services: virtual-machines
2+
title: NDv2-series summary include file
3+
description: Include file for NDv2-series summary
54
author: mattmcinnes
65
ms.topic: include
76
ms.service: virtual-machines
87
ms.subservice: sizes
9-
ms.date: 04/18/2024
8+
ms.date: 08/01/2024
109
ms.author: mattmcinnes
10+
ms.reviewer: mattmcinnes
1111
ms.custom: include file
1212
---
13-
The NDv2-series virtual machine is a new addition to the GPU family designed for the needs of the most demanding GPU-accelerated AI, machine learning, simulation, and HPC workloads. NDv2 is powered by 8 NVIDIA Tesla V100 NVLINK-connected GPUs, each with 32 GB of GPU memory. Each NDv2 VM also has 40 non-HyperThreaded Intel Xeon Platinum 8168 (Skylake) cores and 672 GiB of system memory. NDv2 instances provide excellent performance for HPC and AI workloads utilizing CUDA GPU-optimized computation kernels, and the many AI, ML, and analytics tools that support GPU acceleration 'out-of-box,' such as TensorFlow, Pytorch, Caffe, RAPIDS, and other frameworks. Critically, the NDv2 is built for both computationally intense scale-up (harnessing 8 GPUs per VM) and scale-out (harnessing multiple VMs working together) workloads. The NDv2 series now supports 100-Gigabit InfiniBand EDR backend networking, similar to that available on the HB series of HPC VM, to allow high-performance clustering for parallel scenarios including distributed training for AI and ML. This backend network supports all major InfiniBand protocols, including those employed by NVIDIA’s NCCL2 libraries, allowing for seamless clustering of GPUs.
13+
The NDv2-series virtual machine is a new addition to the GPU family designed for the needs of the most demanding GPU-accelerated AI, machine learning, simulation, and HPC workloads.
14+
15+
NDv2 is powered by 8 NVIDIA Tesla V100 NVLINK-connected GPUs, each with 32 GB of GPU memory. Each NDv2 VM also has 40 non-HyperThreaded Intel Xeon Platinum 8168 (Skylake) cores and 672 GiB of system memory.
16+
17+
NDv2 instances provide excellent performance for HPC and AI workloads utilizing CUDA GPU-optimized computation kernels, and the many AI, ML, and analytics tools that support GPU acceleration 'out-of-box,' such as TensorFlow, Pytorch, Caffe, RAPIDS, and other frameworks.
18+
19+
Critically, the NDv2 is built for both computationally intense scale-up (harnessing 8 GPUs per VM) and scale-out (harnessing multiple VMs working together) workloads. The NDv2 series now supports 100-Gigabit InfiniBand EDR backend networking, similar to that available on the HB series of HPC VM, to allow high-performance clustering for parallel scenarios including distributed training for AI and ML. This backend network supports all major InfiniBand protocols, including those employed by NVIDIA’s NCCL2 libraries, allowing for seamless clustering of GPUs.

‎articles/virtual-machines/sizes/gpu-accelerated/nd-family.md

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -40,11 +40,11 @@ ms.author: mattmcinnes
4040

4141

4242
### ND_A100_v4-series
43-
[!INCLUDE [nd-a100-v4-series-summary](./includes/nda100v4-series-summary.md)]
43+
[!INCLUDE [nd-a100-v4-series-summary](./includes/ndasra100v4-series-summary.md)]
4444

45-
[View the full ND_A100_v4-series page](./nda100v4-series.md).
45+
[View the full ND_A100_v4-series page](./ndasra100v4-series.md).
4646

47-
[!INCLUDE [nd-a100-v4-series-specs](./includes/nda100v4-series-specs.md)]
47+
[!INCLUDE [nd-a100-v4-series-specs](./includes/ndasra100v4-series-specs.md)]
4848

4949

5050
### NDm_A100_v4-series

‎articles/virtual-machines/sizes/gpu-accelerated/ndh100v5-series.md

Lines changed: 104 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,104 @@
1+
---
2+
title: ND-H100-v5 size series
3+
description: Information on and specifications of the ND-H100-v5-series sizes
4+
author: mattmcinnes
5+
ms.service: azure-virtual-machines
6+
ms.subservice: sizes
7+
ms.topic: conceptual
8+
ms.date: 08/01/2024
9+
ms.author: mattmcinnes
10+
ms.reviewer: mattmcinnes
11+
---
12+
13+
# ND-H100-v5 sizes series
14+
15+
[!INCLUDE [nd-h100-v5-summary](./includes/ndh100v5-series-summary.md)]
16+
17+
## Host specifications
18+
[!INCLUDE [nd-h100-v5-series-specs](./includes/ndh100v5-series-specs.md)]
19+
20+
## Feature support
21+
[Premium Storage](../../premium-storage-performance.md): Supported <br>[Premium Storage caching](../../premium-storage-performance.md): Supported <br>[Live Migration](../../maintenance-and-updates.md): Not Supported <br>[Memory Preserving Updates](../../maintenance-and-updates.md): Not Supported <br>[Generation 2 VMs](../../generation-2.md): Supported <br>[Generation 1 VMs](../../generation-2.md): Not Supported <br>[Accelerated Networking](../../../virtual-network/create-vm-accelerated-networking-cli.md): Supported <br>[Ephemeral OS Disk](../../ephemeral-os-disks.md): Supported <br>[Nested Virtualization](/virtualization/hyper-v-on-windows/user-guide/nested-virtualization): Not Supported <br>
22+
23+
## Sizes in series
24+
25+
### [Basics](#tab/sizebasic)
26+
27+
vCPUs (Qty.) and Memory for each size
28+
29+
| Size Name | vCPUs (Qty.) | Memory (GB) |
30+
| --- | --- | --- |
31+
| Standard_ND96isr_H100_v5 | 96 | 1900 |
32+
33+
#### VM Basics resources
34+
- [Check vCPU quotas](../../../virtual-machines/quotas.md)
35+
36+
### [Local storage](#tab/sizestoragelocal)
37+
38+
Local (temp) storage info for each size
39+
40+
| Size Name | Max Temp Storage Disks (Qty.) | Temp Disk Size (GiB) | Temp Disk Random Read (RR)<sup>1</sup> IOPS | Temp Disk Random Read (RR)<sup>1</sup> Speed (MBps) | Temp Disk Random Write (RW)<sup>1</sup> IOPS | Temp Disk Random Write (RW)<sup>1</sup> Speed (MBps) | Local-Special-Disk-Count | Local-Special-Disk-Size-GB | Local-Special-Disk-RR-IOPS | Local-Special-Disk-RR-MBps |
41+
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
42+
| Standard_ND96isr_H100_v5 | 1 | 28000 | | | | | | | | |
43+
44+
#### Storage resources
45+
- [Introduction to Azure managed disks](../../../virtual-machines/managed-disks-overview.md)
46+
- [Azure managed disk types](../../../virtual-machines/disks-types.md)
47+
- [Share an Azure managed disk](../../../virtual-machines/disks-shared.md)
48+
49+
#### Table definitions
50+
- <sup>1</sup>Temp disk speed often differs between RR (Random Read) and RW (Random Write) operations. RR operations are typically faster than RW operations. The RW speed is usually slower than the RR speed on series where only the RR speed value is listed.
51+
- Storage capacity is shown in units of GiB or 1024^3 bytes. When you compare disks measured in GB (1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller. For example, 1023 GiB = 1098.4 GB.
52+
- Disk throughput is measured in input/output operations per second (IOPS) and MBps where MBps = 10^6 bytes/sec.
53+
- To learn how to get the best storage performance for your VMs, see [Virtual machine and disk performance](../../../virtual-machines/disks-performance.md).
54+
55+
### [Remote storage](#tab/sizestorageremote)
56+
57+
Remote (uncached) storage info for each size
58+
59+
| Size Name | Max Remote Storage Disks (Qty.) | Uncached Disk IOPS | Uncached Disk Speed (MBps) | Uncached Disk Burst<sup>1</sup> IOPS | Uncached Disk Burst<sup>1</sup> Speed (MBps) | Uncached Special<sup>2</sup> Disk IOPS | Uncached Special<sup>2</sup> Disk Speed (MBps) | Uncached Burst<sup>1</sup> Special<sup>2</sup> Disk IOPS | Uncached Burst<sup>1</sup> Special<sup>2</sup> Disk Speed (MBps) |
60+
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
61+
| Standard_ND96isr_H100_v5 | 32 | 40800 | 612 | | | | | | |
62+
63+
#### Storage resources
64+
- [Introduction to Azure managed disks](../../../virtual-machines/managed-disks-overview.md)
65+
- [Azure managed disk types](../../../virtual-machines/disks-types.md)
66+
- [Share an Azure managed disk](../../../virtual-machines/disks-shared.md)
67+
68+
#### Table definitions
69+
- <sup>1</sup>Some sizes support [bursting](../../disk-bursting.md) to temporarily increase disk performance. Burst speeds can be maintained for up to 30 minutes at a time.
70+
- <sup>2</sup>Special Storage refers to either [Ultra Disk](../../../virtual-machines/disks-enable-ultra-ssd.md) or [Premium SSD v2](../../../virtual-machines/disks-deploy-premium-v2.md) storage.
71+
- Storage capacity is shown in units of GiB or 1024^3 bytes. When you compare disks measured in GB (1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller. For example, 1023 GiB = 1098.4 GB.
72+
- Disk throughput is measured in input/output operations per second (IOPS) and MBps where MBps = 10^6 bytes/sec.
73+
- Data disks can operate in cached or uncached modes. For cached data disk operation, the host cache mode is set to ReadOnly or ReadWrite. For uncached data disk operation, the host cache mode is set to None.
74+
- To learn how to get the best storage performance for your VMs, see [Virtual machine and disk performance](../../../virtual-machines/disks-performance.md).
75+
76+
77+
### [Network](#tab/sizenetwork)
78+
79+
Network interface info for each size
80+
81+
| Size Name | Max NICs (Qty.) | Max Bandwidth (Mbps) |
82+
| --- | --- | --- |
83+
| Standard_ND96isr_H100_v5 | 8 | 80000 |
84+
85+
#### Networking resources
86+
- [Virtual networks and virtual machines in Azure](../../../virtual-network/network-overview.md)
87+
- [Virtual machine network bandwidth](../../../virtual-network/virtual-machine-network-throughput.md)
88+
89+
#### Table definitions
90+
- Expected network bandwidth is the maximum aggregated bandwidth allocated per VM type across all NICs, for all destinations. For more information, see [Virtual machine network bandwidth](../../../virtual-network/virtual-machine-network-throughput.md)
91+
- Upper limits aren't guaranteed. Limits offer guidance for selecting the right VM type for the intended application. Actual network performance will depend on several factors including network congestion, application loads, and network settings. For information on optimizing network throughput, see [Optimize network throughput for Azure virtual machines](../../../virtual-network/virtual-network-optimize-network-bandwidth.md).
92+
- To achieve the expected network performance on Linux or Windows, you may need to select a specific version or optimize your VM. For more information, see [Bandwidth/Throughput testing (NTTTCP)](../../../virtual-network/virtual-network-bandwidth-testing.md).
93+
94+
### [Accelerators](#tab/sizeaccelerators)
95+
96+
Accelerator (GPUs, FPGAs, etc.) info for each size
97+
98+
| Size Name | Accelerators (Qty.) | Accelerator-Memory (GB) |
99+
| --- | --- | --- |
100+
| Standard_ND96isr_H100_v5 | 8 | 640 |
101+
102+
---
103+
104+
[!INCLUDE [sizes-footer](../includes/sizes-footer.md)]

0 commit comments

Comments
 (0)