Planning for IBM Visual Insights
You must meet the software and hardware requirements and understand the supported file types before you can install IBM® Visual Insights.
Hardware requirements
- Hardware
- The following devices are supported:
- POWER8 S822LC (8335-GTB) or POWER9 AC922 with at least one NVIDIA NVLink capable GPU
- POWER9 IC922 with at least one NVIDIA T4 GPU
- x86 system with at least one NVIDIA Pascal, Volta, or Turing-architecture GPU
- Required specifications
- Your device must meet these minimum requirements:
- 64 GB of memory
- Ethernet network interface
- 75 GB of storage for installation, and at least 40 GB of storage for runtime. See Disk space requirements for details.
- The Nvidia GPU must be configured in the
Default
compute mode. TheExclusive Process
mode will cause trainings to fail. See nvidia-smi usage for details on compute modes. - There are multiple options for deploying the model for testing. Deploying a model to a Xilinx FPGA requires the Xilinx Alveo U200 Accelerator card.
Software requirements
You must install the following software before you install IBM Visual Insights:- Linux
-
- Red Hat Enterprise Linux® (RHEL) RHEL 7.6 ALT (little endian) for POWER9™
- RHEL 7.7 for x86
- Ubuntu 18.04Note: The Ubuntu Hardware Enablement (HWE) kernel is not supported. Kubernetes services do not function correctly, preventing IBM Visual Insights from starting successfully.
- NVIDIA CUDA
- 10.1 Update 1 (if fix pack 1 is not
installed) or later drivers.
If fix pack 1 is installed, 10.2 or later drivers are required.
For information, see the NVIDIA CUDA Toolkit website.
- Docker
-
- RHEL: docker-1.13.1
- Ubuntu: Docker CE or EE 18.06.01
Networking requirements
- A default route must be specified on the host system.
- For instructions to do this on Ubuntu, refer to the IP addressing section in the Ubuntu Network Configuration. Search for the steps to configure and verify the default gateway.
- For instructions to do this on Red Hat Enterprise Linux (RHEL), refer to 2.2.4 Static Routes and the Default Gateway in the Red Hat Customer Portal.
- For RHEL, docker0 must be in a trusted firewall zone. If it is not in a trusted firewall zone,
modify the RHEL settings as follows:
sudo nmcli device set docker0 managed yes sudo nmcli connection modify docker0 connection.zone trusted sudo systemctl stop NetworkManager.service sudo firewall-cmd --permanent --zone=trusted --change-interface=docker0 sudo systemctl start NetworkManager.service sudo nmcli connection modify docker0 connection.zone trusted sudo systemctl restart docker.service
- IPv4 port forwarding must be enabled.
If IPv4 port forwarding is not enabled, run the /sbin/sysctl -w net.ipv4.conf.all.forwarding=1 command. For more information about port forwarding with Docker, see UCP requires IPv4 IP Forwarding in the Docker success center.
- IPv6 must be enabled.
Disk space requirements
IBM Visual Insights has the following storage requirements for the initial product installation and for the data sets that will be managed by the product.
- Standalone installation
-
- /var - The product installation requires at least 75 GB of
space in the
/var
file system for the product Docker images. IBM Visual Insights also generates log information in this file system. The installation process requires additional space because all Docker images are extracted to disk requiring about 40 GB of space, then the images are loaded by Docker. When the image is loaded, the extracted image is deleted but while images are being extracted and loaded, space is needed for both copies. All application Docker images are extracted and loaded in parallel.Recommendation: If you want to minimize the root (/) file system, make sure that
/var
has its own volume. - /opt - IBM Visual Insights
data sets, models, and runtime data are stored in this file system. This file system must have at
least five GB of free space, in addition to any data sets, models or other
runtime data for IBM Visual Insights to operate successfully. The
storage needs will vary depending on the data sets and the contents. For example, video data can
require large amounts of storage.
Recommendation: If you want to minimize the root (
/
) file system, make sure that/opt
has its own volume. The/opt
file system should have at least 40 GB of space, although this value might be more depending on your data sets.
- /var - The product installation requires at least 75 GB of
space in the
- IBM Cloud Private installation
- IBM Visual Insights will use the configured persistent storage for the deployment, the requirements are documented in Installing IBM Visual Insights with IBM Cloud Private.
Supported web browsers
The following web browsers are supported:
- Google Chrome Version 60, or later
- Firefox Quantum 59.0, or later
Image support
- The following image formats are supported:
- JPEG
- PNG
- DICOM
- Images with COCO annotations are supported. For details, see Importing images with COCO annotations.
- IBM Visual Insights has limited
support for Pascal VOC annotations. Annotations for multiple files residing in a common XML file are
not supported. In other words, each annotation XML file can only contain annotations for a single
image, identified by the
filename
attribute.If you have a single XML annotation file containing annotations for multiple images in the data set to be imported, the annotations need to be split out into separate XML files before IBM Visual Insights can import the annotations successfully.
- The models used by IBM Visual Insights
have limitations on the size and resolution of images. If the original data is high resolution, then
the user must consider:
- If the images do not need fine detail for classification or object detection, they should be down-sampled to 1-2 megapixels.
- If the images do require fine detail, they should to be divided into smaller images of 1-2 megapixels each.
- There is a 24 GB size limit per upload session. This limit applies to a single .zip file or a set of files. You can, however upload 24 GB of files, then upload more after the original upload completes.
- Large images will be scaled to the appropriate dimensions for the model as follows:
- SSD: 512 x 512 pixels
The original aspect ratio is maintained. If necessary, black bands are added to the image to make it fit.
- tiny YOLO V2: 416 x 416 pixels
The longest edge is scaled to 416 pixels and, if necessary, black bands are added to the shorter side to make it 416 pixels.
YOLO V3: 608 x 608 pixels
- Faster R-CNN: 1000 x 600 pixels
The original aspect ratio is maintained. If necessary, black bands are added to the image to make it fit.
- Detectron: 1333 x 800 pixels
- GoogLeNet: 224 x 224 pixels
- Action detection: 224 x 224 pixels
- SSD: 512 x 512 pixels
Supported video types
The following video formats are supported:- Can be played in the IBM Visual Insights GUI:
-
- Ogg Vorbis (.ogg)
- VP8 or VP9 (.webm)
- H.264 encoded videos with MP4 format (.mp4)
- Supported by API only:
-
- Matroska (.mkv)
- Audio Video Interleave (.avi)
- Moving Picture Experts Group (.mpg or .mpeg2)
- Not supported:
- Videos that are encoded with the H.265 codec.
Deep learning frameworks
The following frameworks are included with IBM Visual Insights.
Framework | Version | Python 2.7 support | Python3.6 support | Notes |
---|---|---|---|---|
Caffe 2 | 1.0.0 (If fix pack 1 is not installed.)
|
Yes | No | Supported for Detectron models |
IBM Caffe | 1.0.0 | Yes | No | Supported for GoogLeNet, Faster R-CNN, and tinyYOLO V2 models |
Keras | 2.2.4 (If fix pack 1 is not installed.)
|
No (If fix pack 1 is not installed.)
|
Yes | Supported for custom models |
TensorFlow | 1.14 (If fix pack 1 is not installed.)
|
No | Yes | Supported for custom models |
TensorRT | 5.1.3 (If fix pack 1 is not installed.)
|
Yes | Yes | Supported for SSD models |
PyTorch | 1.1.0 (If fix pack 1 is not installed.)
|
Yes | Yes | Supported for custom models |
Limitations
- IBM Visual Insights uses an entire GPU when you
are training a dataset. Multiple GoogleNet or Faster R-CNN models can be deployed to a single GPU.
Other types of models take an entire GPU when deployed. For details about other differences between
model types, see Model functionality.
The number of active GPU tasks (model training and deployment) that you can run at the same time depends on the number of GPUs on your server. You must verify that there are enough available GPUs on the system for the desired workload. The number of available GPUs is displayed on the user interface.
- You cannot install IBM Visual Insights
stand-alone on a system that already has any of these products installed:
- IBM Data Science Experience (DSX)
- IBM Cloud Private
- IBM Watson Studio Local Edition
- Any other Kubernetes based applications