BenchmarkXPRT Blog banner

Tag Archives: precision

The Introduction to AIXPRT white paper is now available!

Today, we published the Introduction to AIXPRT white paper. The paper serves as an overview of the benchmark and a consolidation of AIXPRT-related information that we’ve published in the XPRT blog over the past several months. For folks who are completely new to AIXPRT and veteran testers who need to brush up on pre-test configuration procedures, we hope this paper will be a quick, one-stop reference that helps reduce the learning curve.

The paper describes the AIXPRT toolkits and workloads, adjusting key test parameters (batch size, level of precision, number of concurrent instances, and default number of requests), using alternate test configuration files, understanding and submitting results, and accessing the source code.

We hope that Introduction to AIXPRT will prove to be a valuable resource. Moving forward, readers will be able to access the paper from the Helpful Info box on AIXPRT.com and the AIXPRT section of our XPRT white papers page. If you have any questions about AIXPRT, please let us know!

Justin

How to use alternate configuration files with AIXPRT

In last week’s AIXPRT Community Preview 3 announcement, we mentioned the new public GitHub repository that we’re using to publish AIXPRT-related information and resources. In addition to the installation readmes for each AIXPRT installation package, the repository contains a selection of alternative test config files that testers can use to quickly and easily change a test’s parameters.

As we discussed in previous blog entries about batch size, levels of precision, and number of concurrent instances, AIXPRT testers can adjust each of these key variables by editing the JSON file in the AIXPRT/Config directory. While the process is straightforward, editing each of the variables in a config file can take some time, and testers don’t always know the appropriate values for their system. To address both of these issues, we are offering a selection of alternative config files that testers can download and drop into the AIXPRT/Config directory.

In the GitHub repository, we’ve organized the available config files first by operating system (Linux_Ubuntu and Windows) and then by vendor (All, Intel, and NVIDIA). Within each section, testers will find preconfigured JSON files set up for several scenarios, such as running with multiple concurrent instances on a system’s CPU or GPU, running with FP32 precision instead of FP16, etc. The picture below shows the preconfigured files that are currently available for systems running Ubuntu on Intel hardware.

AIXPRT public repository snip 2

Because potential AIXPRT use cases cut across a wide range of hardware segments, including desktops, edge devices, and servers, not all AIXPRT workloads and configs will be applicable to each segment. As we move towards the AIXPRT GA, we’re working to find the best way to parse out these distinctions and communicate them to end users. In many cases, the ideal combination of test configuration variables remains an open question for ongoing research. However, we hope the alternative configuration files will help by giving testers a starting place.

If you experiment with an alternative test configuration file, please note that it should replace the existing default config file. If more than one config file is present, AIXPRT will run all the configurations and generate a separate result for each. More information about the config files and detailed instructions for how to handle the files are available in the EditConfig.md document in the public repository.

We’ll continue to keep everyone up to date with AIXPRT news here in the blog. If you have any questions or comments, please let us know.

Justin

Understanding concurrent instances in AIXPRT

Over the past few weeks, we’ve discussed several of the key configuration variables in AIXPRT, such as batch size and level of precision. Today, we’re discussing another key variable: number of concurrent instances. In the context of machine learning inference, this refers to how many instances of the network model (ResNet-50, SSD-MobileNet, etc.) the benchmark runs simultaneously.

By default, the toolkits in AIXPRT run one instance at a time and distribute the compute load according to the characteristics of the CPU or GPU under test, as well as any relevant optimizations or accelerators in the toolkit’s reference library. By setting the number of concurrent instances to a number greater than one, a tester can use multiple CPUs or GPUs to run multiple instances of a model at the same time, usually to increase throughput.

With multiple concurrent instances, a tester can leverage additional compute resources to potentially achieve higher throughput without sacrificing latency goals.

In the current version of AIXPRT, testers can run multiple concurrent instances in the OpenVINO, TensorFlow, and TensorRT toolkits. When AIXPRT Community Preview 3 becomes available, this option will extend to the MXNet toolkit. OpenVINO and TensorRT automatically allocate hardware for each instance and don’t let users make manual adjustments. TensorFlow and MXNet require users to manually bind instances to specific hardware. (Manual hardware allocation for multiple instances is more complicated than we can cover today, so we may devote a future blog entry to that topic.)

Setting the number of concurrent instances in AIXPRT

The screenshot below shows part of a sample config file (the same one we used when we discussed batch size and precision). The value in the “concurrent instances” row indicates how many concurrent instances will be operating during the test. In this example, the number is one. To change that value, a tester simply replaces it with the desired number and saves the changes.

Config_snip

If you have any questions or comments (about concurrent instances or anything else), please feel free to contact us.

Justin

Understanding the basics of AIXPRT precision settings

A few weeks ago, we discussed one of AIXPRT’s key configuration variables, batch size. Today, we’re discussing another key variable: the level of precision. In the context of machine learning (ML) inference, the level of precision refers to the computer number format (FP32, FP16, or INT8) representing the weights (parameters) a network model uses when performing the calculations necessary for inference tasks.

Higher levels of precision for inference tasks help decrease the number of false positives and false negatives, but they can increase the amount of time, memory bandwidth, and computational power necessary to achieve accurate results. Lower levels of precision typically (but not always) enable the model to process inputs more quickly while using less memory and processing power, but they can allow a degree of inaccuracy that is unacceptable for certain real-world applications.

For example, a high level of precision may be appropriate for computer vision applications in the medical field, where the benefits of hyper-accurate object detection and classification far outweigh the benefit of saving a few milliseconds. On the other hand, a low level of precision may work well for vision-based sensors in the security industry, where alert time is critical and monitors simply need to know if an animal or a human triggered a motion-activated camera.

FP32, FP16, and INT8

In AIXPRT, we can instruct the network models to use FP32, FP16, or INT8 levels of precision:

  • FP32 refers to single-precision (32-bit) floating point format, a number format that can represent an enormous range of values with a high degree of mathematical precision. Most CPUs and GPUs handle 32-bit floating point operations very efficiently, and many programs that use neural networks, including AIXPRT, use FP32 precision by default.
  • FP16 refers to half-precision (16-bit) floating point format, a number format that uses half the number of bits as FP32 to represent a model’s parameters. FP16 is a lower level of precision than FP32, but it still provides a great enough numerical range to successfully perform many inference tasks. FP16 often requires less time than FP32, and uses less memory.
  • INT8 refers to the 8-bit integer data type. INT8 data is better suited for certain types of calculations than floating point data, but it has a relatively small numeric range compared to FP16 or FP32. Depending on the model, INT8 precision can significantly improve latency and throughput, but there may be a loss of accuracy. INT8 precision does not always trade accuracy for speed, however. Researchers have shown that a process called quantization (i.e., approximating continuous values with discrete counterparts) can enable some networks, such as ResNet-50, to run INT8 precision without any significant loss of accuracy.

Configuring precision in AIXPRT

The screenshot below shows part of a sample config file, the same sample file we used for our batch size discussion. The value in the “precision” row indicates the precision setting. This test configuration would run tests using INT8. To change the precision, a tester simply replaces that value with “fp32” or “fp16” and saves the changes.

Config_snip

Note that while decreasing the precision from FP32 to FP16 or INT8 often results in larger throughput numbers and faster inference speeds overall, this is not always the case. Many other factors can affect ML performance, including (but not limited to) the complexity of the model, the presence of specific ML optimizations for the hardware under test, and any inherent limitations of the target CPU or GPU.

As with most AI-related topics, the details of model precision are extremely complex, and it’s a hot topic in cutting edge AI research. You don’t have to be an expert, however, to understand how changing the level of precision can affect AIXPRT test results. We hope that today’s discussion helped to make the basics of precision a little clearer. If you have any questions or comments, please feel free to contact us.

Justin

Check out the other XPRTs:

Forgot your password?