Skip to content

Commit 4772a23

Browse files
Merge pull request #241750 from jvenka/docs-editor/hbv4-performance-1686872352
Update HBv4 performance page
2 parents 0a374f9 + 2f7cbef commit 4772a23

File tree

1 file changed

+3
-2
lines changed

1 file changed

+3
-2
lines changed

articles/virtual-machines/hbv4-performance.md

Lines changed: 3 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -20,7 +20,7 @@ Performance expectations using common HPC microbenchmarks are as follows:
2020

2121
| Workload | HBv4 |
2222
|-------------------------------------------------|-------------------------------------------------------------------|
23-
| STREAM Triad | 750-780 GB/s |
23+
| STREAM Triad | 750-780 GB/s of DDR5, up to 5.7 TB/s of 3D V-Cache bandwidth |
2424
| High-Performance Linpack (HPL) | Up to 7.6 TF (Rpeak, FP64) for 144-core VM size |
2525
| RDMA latency & bandwidth | < 2 microseconds (1 byte), 400 Gb/s (one-way) |
2626
| FIO on local NVMe SSDs (RAID0) | 12 GB/s reads, 7 GB/s writes; 186k IOPS reads, 201k IOPS writes |
@@ -38,7 +38,6 @@ cd woc-benchmarking/apps/hpc/stream/
3838
sh build_stream.sh
3939
sh stream_run_script.sh $PWD “hbrs_v4”
4040
```
41-
4241
## Compute performance test
4342

4443
The HPL benchmark can be run using the script in this github repository.
@@ -83,3 +82,5 @@ numactl --physcpubind=[INSERT CORE #] ib_send_lat -a
8382
- Review the performance and scalability results of HPC applications on the HBv4 VMs at the [TechCommunity article](https://techcommunity.microsoft.com/t5/azure-compute/hpc-performance-and-scalability-results-with-azure-hbv4-vms/bc-p/2235843).
8483
- Read about the latest announcements, HPC workload examples, and performance results at the [Azure HPC Microsoft Community Hub](https://techcommunity.microsoft.com/t5/azure-high-performance-computing/bg-p/AzureHighPerformanceComputingBlog).
8584
- For a higher-level architectural view of running HPC workloads, see [High Performance Computing (HPC) on Azure](/azure/architecture/topics/high-performance-computing/).
85+
86+

0 commit comments

Comments
 (0)