Arm System Characterization Tool

The Arm System Characterization Tool (ASCT) is a standalone command-line utility for running low-level benchmarks, diagnostic scripts, and system tests to analyze and debug performance on Arm-based platforms.

ASCT provides a standardized environment for evaluating key hardware characteristics, such as memory latency and bandwidth, and is especially suited for platform bring-up, system tuning, and architectural comparison tasks. It helps developers and system architects gain early and repeatable insights into performance-critical subsystems.

Current capabilities include:

• System report. Generate a structured inventory of CPU, caches, NUMA topology, operating system, kernel, and tooling state. You can also generate Arm Coherent Mesh Network (CMN) topology and configuration reports on demand. For more information, see System report.

• Memory latency and bandwidth benchmarks. Measure how latency and bandwidth change across cache levels and DRAM. Characterize NUMA effects, including idle latency, loaded latency, core-to-core latency, and peak bandwidth. For more information, see Memory characterization.

• Storage latency and bandwidth benchmarks. Run controlled fio-based parameter sweeps. Vary request size, queue depth, process count, and access pattern. Measure bandwidth, IOPS, latency, and CPU utilization under input and output load. For more information, see Storage characterization.

• System diff. Compare one or more ASCT run output directories. Highlight configuration differences and benchmark differences. Generate CSV and JSON output and plots for supported benchmarks. For more information, see Run diff.

Planned features include:

• Hardware register inspection and configuration for debug and bring-up use

• Input and output performance testing

• Locking and synchronization stress tests

• Floating-point compute performance measurement in FLOPS

Documentation

For detailed guidance, see:

• Installation

• Getting started

• System report

• Memory characterization

• Storage characterization

• Run diff

Requirements

• ASCT is supported on Linux systems
• Python 3.10 or later. You need pip if you want to use a Python virtual environment (options 1 and 2).
• Build tools such as:
  • gcc
  • make
• If you want to install ASCT as a system-wide tool (option 3), you need:
  • uv. Install uv from: https://docs.astral.sh/uv/getting-started/installation/
  • sudo privileges
  • Permission to create:
    • /opt/uv/tools
    • /usr/local/bin

Download ASCT

1. Download the latest ASCT release bundle from: artifacts.tools.arm.com.

   For users within the Arm network, early-access versions might be available at: artifacts.internal.tools.arm.com.

2. Locate the downloaded file on your host machine. Move it to the target machine where you want to install and run the tool.

3. Extract the release bundle:

   tar -xvzf asct-<version>-release.tar.gz

   This process creates a directory:

   asct-<version>/

   The directory contains the following files:

   • asct-<version>.tar.gz: The Python source distribution (the package to install)
   • asct_docs/: Generated HTML documentation. Open asct_docs/index.html in a web browser.
   • install.sh: An optional installer helper (tested on Ubuntu 24.04)
   • README.txt: A release bundle summary and quick-start notes

Install ASCT

You can install ASCT in one of the following ways:

• Option 1 (recommended): Install into a Python virtual environment using the install.sh helper script.

• Option 2: Install manually using pip into a Python environment.

• Option 3: Install system-wide using uv (for shared machines or CI systems).

Option 1: Install ASCT with the helper script

This is the simplest installation method. It creates a dedicated Python virtual environment and installs ASCT into it.

Steps

From inside the extracted asct-<version>/ directory:

    ./install.sh --install-dir /path/to/install/directory

If --install-dir is not specified, the script installs into the directory containing install.sh.

The install.sh script performs the following tasks:

• Creates a Python virtual environment
• Installs ASCT into that environment using pip
• Optionally builds fio if required

By default, the virtual environment is created at <install-dir>/asct_venv.

After installation, follow the instructions printed by the script to activate the virtual environment before running ASCT.

Option 2: Install ASCT manually using pip

Use this method if you prefer to manage Python environments manually, are installing into an existing virtual environment, or are automating installation in your own scripts.

You can install ASCT into any active Python environment.

Install into a new virtual environment

1. Create a virtual environment:

   python3 -m venv asct_venv

2. Activate it:

   source asct_venv/bin/activate

3. Install ASCT from the bundled source distribution:

   python3 -m pip install ./asct-<version>.tar.gz

   ASCT is now installed inside the virtual environment.

Install into an existing environment

python3 -m pip install ./asct-<version>.tar.gz

ASCT is installed into the currently active Python environment.

Option 3: Install ASCT as a system-wide tool with uv

This method installs ASCT as a system-wide tool available on the system PATH.

Use this option only for shared systems, managed environments, or CI environments.

Steps

1. Create the required directories:

   sudo mkdir -p /opt/uv/tools /usr/local/bin

2. Install ASCT with uv and place the executable in /usr/local/bin:

   sudo UV_TOOL_DIR=/opt/uv/tools UV_TOOL_BIN_DIR=/usr/local/bin uv tool install ./asct-<version>.tar.gz

   After installation, ASCT is available on the system PATH.

Runtime tools

Some workloads require additional system tools, such as:

• numactl
• fio 3.36 or later

If tools are missing, ASCT reports this at runtime.

Run ASCT

• If you installed ASCT in a virtual environment, activate the environment first.
• If you installed ASCT with uv, no activation is required.

To verify the installation, run:

asct help

You can also open the offline documentation at: asct_docs/index.html.

Uninstall ASCT

Option 1: If you installed ASCT using the helper script

To uninstall ASCT, remove that virtual environment directory:

rm -rf <install-dir>/asct_venv

Option 2: If you manually installed ASCT using pip

If you installed ASCT into a dedicated virtual environment, you can remove that environment directory.

Alternatively, activate the relevant Python environment and uninstall ASCT:

python3 -m pip uninstall asct

Option 3: If you installed ASCT system-wide using uv

Remove ASCT using uv:

sudo UV_TOOL_DIR=/opt/uv/tools UV_TOOL_BIN_DIR=/usr/local/bin uv tool uninstall asct

Getting started

ASCT contains a number of separate commands with the following common pattern:

asct [command] [arguments]

For more information on a specific command, run:

asct [command] --help

To display the ASCT version, run:

asct version

To get started, run the default set of benchmarks:

sudo asct run

[!IMPORTANT] Some benchmarks require sudo or root privileges to configure huge pages and access certain system information. You can run ASCT without sudo, but some benchmarks might be unavailable or have limited functionality.

asct run command

The asct run command runs one or more benchmarks.

Default behavior

sudo asct run

Running the asct run command with no arguments executes a default set of benchmarks and displays the results in the terminal.

Each time you run asct run, ASCT collects system information and includes it with the benchmark results. ASCT also saves detailed output in a directory under the current working directory.

The output can include data in multiple formats, such as CSV and JSON. Some benchmarks also generate additional artifacts, such as plots or raw data dumps.

By default, ASCT creates an output directory named: data.<YYYYMMDD_HHMMSS_microseconds>.

You can change the output location using the --output-dir flag.

Select benchmarks

To run all available benchmarks:

sudo asct run all

To run specific benchmarks, pass a list of benchmark names as arguments to the asct run command.

For example, to run the latency-sweep and idle-latency benchmarks:

sudo asct run latency-sweep idle-latency

Alternatively, each benchmark also has associated keywords that you can use to select multiple benchmarks that match those keywords.

For example, to run all benchmarks tagged with the memory keyword:

sudo asct run memory

You can exclude benchmarks by negating their names or keywords. To do this, prepend the name or keyword with a caret (^).

For example, to run all benchmarks but exclude all of those tagged with the bandwidth keyword:

sudo asct run all ^bandwidth

[!NOTE] If a benchmark is a dependency for running another benchmark, negating or excluding it has no effect. For example, if you use ^latency-sweep to exclude the latency-sweep benchmark, but another benchmark depends on latency-sweep, the latency-sweep benchmark still runs.

To display the list of available benchmarks and their keywords, run the following command:

asct list

Output options

The run command also supports the following arguments:

• --format, -f [stdout, csv, json]

  • Specify the output format, either to the terminal with stdout (default), individual CSV files (benchmark-name.csv), or a single combined JSON file (report.json)

    For example: asct run idle-latency --format=json

• --log-level, -L [debug,info,warning,error,critical]
  • Specify verbosity. The level is info by default, but you can configure it.
• --log-file LOG_FILE
  • Specify an output file for logging messages as determined by the --log-level parameter. ASCT writes logs to this file and also prints them to standard error stderr.
  • If ASCT does not emit any log messages, it will not create the log file.
• --output-dir, -o OUTPUT_DIR
  • By default, ASCT saves output files in a directory named data.<YYYYMMDD_HHMMSS_microseconds> within the current working directory. Use this flag to override the default and specify a custom output directory, using either an absolute or a relative path.
• --force
  • When the output directory specified by --output-dir already exists, ASCT displays an error and quits to avoid overwriting data. Use the --force flag to overwrite the output directory if it exists.
• --quiet, -q
  • Disable all output to stdout and stderr, including critical errors and log messages. Use --log-file to capture and view logs.
• --no-progress-bar
  • Disable the animated progress bar. Single-line update messages display instead.

asct list command

Run the following command to get the list of available benchmarks and their associated keywords:

asct list

System report

This section describes the system reporting commands in the Arm System Characterization Tool (ASCT). Use these commands to view hardware, software, and platform configuration details.

ASCT collects system information each time you run asct run .... Arm Coherent Mesh Network (CMN) reporting is available as a separate command.

Use these commands:

• asct system-info: View hardware, operating system, and network information
• asct cmn: View CMN topology and configuration when present

View system information

What asct system-info collects

The asct system-info report provides a structured snapshot of your system. It includes the following information:

• CPU architecture, core count, and CPU model
• Cache hierarchy and cache line size
• NUMA topology
• Interconnect identification
• Memory capacity and theoretical peak bandwidth
• Operating system and kernel details
• Performance tooling and kernel feature checks
• Vulnerability status from kernel reporting

Run asct system-info

Run this command:

sudo asct system-info

If you run the command without sudo, ASCT collects only a limited set of information. ASCT cannot access Desktop Management Interface (DMI) platform and memory details without elevated privileges.

Use asct system-info options

asct system-info supports the common ASCT output and logging options.

Example:

sudo asct system-info --format=json --output-dir data --force --log-level=error --log-file data/asct.log --quiet

View asct system-info output

ASCT supports these output formats:

• --format=stdout (default): Prints the report to the terminal
• --format=csv: Writes a key-value CSV to the output directory
• --format=json: Writes a combined JSON file to the output directory

ASCT writes these files to the output directory:

• system_info.csv
• report.json

Example standard output

Example excerpt:

System Information ------------------------------------------------------------

System feature report:
  Collected:           2025-07-17 08:16:18.729527
  ASCT version:        0.4.2
  Running as root:     True

System hardware:
  Architecture:        ARMv8.2
  CPUs:                128
  Interconnect:        CMN-600 x 1

...

Network Information ------------------------------------------------------------

Local IPv4:      192.0.2.10
Local IPv6:      2001:db8::10
Interfaces:
  eth0: UP mtu=1500 ipv4=192.0.2.10

...

View CMN information

What asct cmn collects

Use asct cmn to collect information about the CMN interconnect when it is present on your system. asct cmn requires root access to gather CMN information.

Depending on the options that you select, the output includes:

• A high-level topology summary
• An ASCII topology diagram
• A register dump view

Run asct cmn

Before you run asct cmn for the first time, run:

asct cmn --detect

The --detect option performs CMN discovery and CPU mapping. This step can take several minutes.

Then run:

asct cmn

Use asct cmn options

Use these optional flags:

• --standard: Prints an ASCII topology diagram
• --verbose: Prints additional detail, including registers
• --detect: Performs CMN discovery and CPU mapping before reporting

Examples:

asct cmn --standard
asct cmn --verbose
asct cmn --detect --verbose

View asct cmn output

ASCT supports these output formats:

• --format=stdout (default): Prints the CMN summary or diagram to the terminal
• --format=csv: Writes CMN information as a CSV file to the output directory
• --format=json: Writes CMN information to the combined JSON report

ASCT writes these files to the output directory:

• cmn.csv
• report.json

Example standard output

The examples in this section are truncated to improve readability.

Default output

In the CPU grid layout, each parenthesized group lists the Linux logical CPU IDs that map to a CMN crosspoint (XP) at that grid position. An empty () indicates that no CPUs map to that XP.

Output with –standard

Output with –verbose

Memory characterization

Latency

Latency sweep

• Use this benchmark to measure memory latency across data sizes from 128 bytes to 1 GiB. The results include the average access latency for each size.

• The results show how memory access latency shifts across cache levels and dynamic random-access memory (DRAM), revealing transitions in the cache hierarchy.

• Randomized linked lists prevent hardware prefetching from influencing the latency measurements.

• The benchmark uses 1 GiB huge pages (also known as large pages on some systems) to reduce the impact of page table walks and translation lookaside buffer (TLB) lookups on latency measurements.

• The benchmark calculates:

  • Lower and upper bounds for each cache level

  • Average latencies

  • The optimal data size for L1, L2, last-level cache (LLC), and DRAM. Other memory benchmarks like Idle Latency and Loaded Latency reuse these sizes to improve precision.

    Latencies at different levels of cache
    --------------------------------------
    Level Lower bound Upper bound Optimum data size Latency [ns]
    L1           128         64K         32.0625K          1.5
    L2           64K        512K             288K          5.2
    LLC           1M          8M             4.5M         48.1
    DRAM         32M          1G             528M        107.6

• The benchmark also generates a line graph and saves it as latency-sweep.png in the output directory.

Idle latency

• Use this benchmark to measure memory access latency from the last core on each node to its local and remote memory in a non-uniform memory access (NUMA) system. In NUMA systems, memory access times vary by location.

• To characterize the system accurately, ensure it is idle. Close all applications and background processes except the test itself. Arm System Characterization Tool (ASCT) imposes only the minimal load needed for measurement, but it cannot control other background activity that might affect test results.

• The benchmark produces a matrix of size n by n, with n equal to the number of NUMA nodes.

  Latencies of random memory access at idle (in nanoseconds)
  ----------------------------------------------------------
          Node 0 Node 1
  Node 0  113.3  266.9
  Node 1  266.5  115.7

Note: ASCT derives the data size used to target DRAM from the latency-sweep benchmark. If not selected manually, ASCT automatically includes latency-sweep as a dependency.

Loaded latency

• Use this benchmark to measure the memory latency of the last core on the first NUMA node to its local memory, while other cores generate increasing memory traffic.

• To vary memory pressure, ASCT interleaves memory reads with different numbers of no-operation (NOP) instructions.

    Loaded latency with background memory activity
    ------------------------------------------------
     Injected NOPs  Loaded latency [ns]  Bandwidth [GB/s]
              3000                115.5              10.7
               900                115.8              35.2
               500                117.6              61.4
               180                122.2             175.3
               100                129.2             306.5
                80                134.1             385.0
                70                138.3             441.8
                50                158.8             603.9
                40                193.9             753.7
                30                250.7             912.7
                20                266.5             918.3
                10                281.0             918.3
                 0                305.4             920.7

Note: ASCT derives the data size used to target DRAM from the latency-sweep benchmark. If not selected manually, ASCT automatically includes latency-sweep as a dependency.

Core-to-core latency

• Use this benchmark to measure the latency of cache line transfers between pairs of CPU cores across the system.

• The benchmark uses a ping-pong microbenchmark that alternates shared memory access and modification between 2 threads pinned to different cores.

• The shared memory region is a randomized linked list. The benchmark pointer-chases this list so that each memory access depends on the result of the previous one. This approach prevents hardware prefetchers from interfering with results.

• The benchmark binds the memory region to a specific NUMA node, so you can measure both local core-to-core latency on the same node and remote core-to-core latency across nodes. This distinction helps you understand the impact of NUMA topology on inter-core communication.

• The benchmark calculates:

  • Node-to-node median latency matrices that summarize inter-node and intra-node communication costs.

  • Highest-latency core pairs that highlight outliers and potential bottlenecks.

  • Asymmetry in latency between different directions, for example A to B versus B to A.

• The benchmark presents results as:

  • Tables that show node-to-node and core-to-core latencies.

  • Heatmaps that show latency patterns across all core pairs.

  • CSV and JSON files that support further analysis.

    Core-to-Core Latency Summary (ns): Data Address @ Local Numa Node
    =================================================================
    Node-to-Node Median Latency Matrix (ns):
    ----------------------------------------
            Node0   Node1
    Node0   31.91  152.86
    Node1  153.10   32.00

    Latency Statistics (ns):
    ------------------------
    Min       : 23.81
    Max       : 161.65
    Mean      : 92.87
    Median    : 140.66

    Top Latency Core Pairs with Median Latency
    ------------------------------------------
    CPUA   CPUB    Latency (ns)
    ---------------------------
    186       91         161.65
    187       83         161.37
     95      186         161.29
     91      190         161.12
    178       90         161.02
    186       83         160.50
     82      157         160.44
     83      176         160.42
     90      187         160.42
     90      159         160.41

    Node-to-Node Latency Statistics (ns):
    -------------------------------------

    Node0 → Node0:
    Min:    24.68 ns
    Max:    42.24 ns
    Mean:   32.16 ns
    Median: 31.91 ns

    Node0 → Node1:
    Min:    112.43 ns
    Max:    161.29 ns
    Mean:   152.67 ns
    Median: 152.86 ns

    Node1 → Node0:
    Min:    141.72 ns
    Max:    161.65 ns
    Mean:   153.14 ns
    Median: 153.10 ns

    Node1 → Node1:
    Min:    23.81 ns
    Max:    41.44 ns
    Mean:   32.26 ns
    Median: 32.00 ns

Ping-pong microbenchmark

• Use this benchmark to measure latency between 2 CPU cores by forcing cache line transfers between them.

• The benchmark pins 2 threads on the target cores.

• Each thread alternates between accessing and modifying the shared data structure. During its turn, a thread performs a complete pointer chase on the structure.

• The benchmark uses cache invalidation and data dependency chains to estimate latency.

• When one thread writes to the shared data structure, the corresponding cache line in the other core is evicted.

NUMA and memory binding

• The benchmark distinguishes between local and remote cache line transfers by binding the memory region to a specific NUMA node. Local transfers, where the cores and memory are on the same node, typically show lower latency. Remote transfers, where the cores and memory are on different nodes, show the cost of traversing the NUMA interconnect.

Bandwidth

Bandwidth sweep

• Use this benchmark to measure the average memory bandwidth achieved by a single core at each level of the memory hierarchy (L1, L2, LLC, and DRAM).

    Bandwidth at different levels of cache
    --------------------------------------
    Data size used Level Bandwidth [GB/s]
         32.0625K    L1            126.5
             288K    L2             74.5
             4.5M   LLC             35.1
             528M  DRAM             15.8

• The benchmark also generates a bandwidth-by-size graph and saves it as bandwidth.png in the output directory.

Note: ASCT derives the data size used to target each cache level from the results of the latency-sweep benchmark. If not selected manually, ASCT automatically includes latency-sweep as a dependency.

Cross-NUMA bandwidth

• This benchmark measures the maximum aggregate memory bandwidth achieved by all cores in one NUMA node when accessing either their local NUMA memory, or memory from a remote NUMA node.

• The benchmark produces an n × n matrix, where n is the number of NUMA nodes.

    Cross-NUMA bandwidths for the system (in GB/s)
    ----------------------------------------------
           Node 0 Node 1
    Node 0  459.1   78.3
    Node 1   78.8  459.2

Note: ASCT derives the data size used to target DRAM from the latency-sweep benchmark. If not selected manually, ASCT automatically includes latency-sweep as a dependency.

Peak bandwidth

• This benchmark makes full use of all cores on all NUMA nodes to measure the maximum achievable memory bandwidth of the system.

• To examine how different memory access patterns affect maximum usage, the benchmark tests multiple traffic types, for example, all reads or mixed reads-writes ratios. Each pattern yields a corresponding peak bandwidth value.

• If available, compare these results with the theoretical peak bandwidth reported in the system information output.

    Peak memory bandwidth
    ---------------------
                Traffic type  Peak BW [GB/s]
                   All Reads           918.9
            3:1 Reads-Writes           868.8
            2:1 Reads-Writes           859.5
            1:1 Reads-Writes           842.7
    2:1 Rd-Wr (Non-Temporal)           652.1

Note: ASCT derives the data size used to target DRAM from the latency-sweep benchmark. If not selected manually, ASCT automatically includes latency-sweep as a dependency.

Storage characterization

This section describes the storage benchmarks in the Arm System Characterization Tool (ASCT). These tests evaluate performance characteristics of block devices and filesystems using fio under controlled sweep parameters.

What ASCT storage benchmarking measures

ASCT evaluates storage performance by sweeping parameters and measuring their impact on input/output (I/O) performance. - Sustained bandwidth - Input/Output Operations Per Second (IOPS) and latency - Host CPU utilization during I/O

CPU utilization is reported as a breakdown of time spent in user space, kernel code, I/O wait, and hardware and software interrupt handling. This information helps identify whether performance is limited by the application, the kernel, or the storage device.

Sweeps included

ASCT automates multiple types of parameter sweeps, each designed to isolate one dimension of I/O behavior: - Request Size Sweep (4 KiB → 128 KiB) Reveals throughput scaling as well as controller or device saturation points - I/O Queue Depth Sweep (QD=1 → QD=128, powers of two) Capture parallelism gains and concurrent access behavior - Concurrent Process Count Sweep (1 → 16 concurrent processes, powers of two) Show scaling across CPU cores - Access Pattern Sweep Sequential vs random, and read/write/mixed ratios (default: 70% reads). Demonstrates workload-specific differences. Compare sequential and random I/O, as well as read-only, write-only, and mixed workloads (default 70% reads). Use this sweep to show how performance varies by workload. ## How ASCT runs parameter sweeps

ASCT uses fio to perform controlled parameter sweeps. For each sweep, ASCT changes one parameter at a time while keeping all other parameters fixed. It then repeats the sweep for each configured parameter set.

For each sweep point, ASCT records the following key metrics: - From fio: bandwidth, I/O rate, average latency, latency distribution - From mpstat: CPU utilization breakdowns (user/system/iowait/hard irq/soft irq)

Run storage benchmarking

Use the run command with storage keywords and optional --user-config parameters to select a device or create a temporary file.

# Run a specific sweep (in order of Request Size Sweep, I/O Depth Sweep, Concurrent Process Count Sweep, Access Pattern Sweep)
sudo asct run storage-request-size-sweep ...
sudo asct run storage-io-depth-sweep ...
sudo asct run storage-process-count-sweep ...
sudo asct run storage-access-pattern-sweep ...

Alternatively, use the short-form alias:

# Run a specific sweep (short form in order of Request Size Sweep, I/O Depth Sweep, Concurrent Process Count Sweep, Access Pattern Sweep)
sudo asct run srss ...
sudo asct run sids ...
sudo asct run spcs ...
sudo asct run saps ...

ASCT requires a target device or file. CAUTION: tests might overwrite file content.

# Provide a target file (ASCT will use /tmp/mytest.dat for test, user needs to create file upfront)
sudo asct run srss --user-config srss.file_names=/tmp/mytest.dat
# Provide a target device (ASCT will use the device /dev/nvme07, user needs to ensure device exists)
sudo asct run srss --user-config srss.file_names=/dev/nvme07

ASCT exposes several user overrides, including the following examples: - Create a temporary file for file benchmarking so that you do not need to provide the device name:

# Auto-create temporary file for file I/O
sudo asct run srss --user-config srss.create_temp_file=1

• Change the read-write mix ratio:

# Change read write mix ratio for access pattern sweep to use 20% read instead of default of 70%
sudo asct run saps --user-config saps.rwmixread=20

• Change the sweep steps:

# Use linear iodepth sweep step between 4 and 8 for more details
sudo asct run sids --user-config sids.iodepth_sweep_steps=4,5,6,7,8

See the ASTC help for a full list of user overrides.

Outputs generated

Storage benchmarking in ASCT produces output similar to memory benchmarking:

• Text summary A text summary either printed to the console or exported as CSV/JSON.
  Each sweep point generates a row with the following metrics:
  • Bandwidth (MB/s)
  • I/O rate (kops)
  • Average latency (µs)
  • CPU utilization (%)
• Graphical plots:
  • Bandwidth vs sweep parameter → storage_io_bandwidth.png
  • I/O rate vs sweep parameter → storage_io_io_rate.png
  • CPU utilization vs sweep parameter → storage_io_cpu_utilization.png
  • Latency cumulative distribution (CDF) line per sweep parameter →
    • storage_io_read_latency_distribution.png
    • storage_io_write_latency_distribution.png

Per-sweep sample outputs

Request Size Sweep

This sweep varies the I/O request size while keeping the rest of parameter constant. It highlights how throughput scales with larger transfers and shows when the device reaches saturation.

Console summary

Plots for details

Bandwidth	I/O Rate

CPU Utilization	Read Latency CDF

I/O Depth Sweep

This sweep varies the I/O queue depth (the number of outstanding requests) while keeping the request size and process count constant.
It measures how effectively the device exploits parallelism.

Console summary

Plots for details

Bandwidth	I/O Rate

CPU Utilization	Read Latency CDF

Concurrent Process Count Sweep

This sweep varies the number of concurrent processes issuing I/O.
It evaluates scaling across CPU cores and submission queues, and highlights the point at which performance no longer improves.

Console summary

Plots for details

Bandwidth	I/O Rate

CPU Utilization	Read Latency CDF

Access Pattern Sweep

This sweep evaluates different workload profiles, including sequential and random access patterns, and the mix of read and write operations. It highlights the workload sensitivity of the device.

Console summary

Plots for details

Bandwidth	I/O Rate	CPU Utilization

Read Latency CDF	Write Latency CDF

Additional notes

• Run with elevated privileges if direct I/O or device binding requires it.
• Ensure that results directories are on a stable path. ASCT saves data files under the run-specific data.* directory.

Run diff

Use asct diff to compare results from previous Arm System Characterization Tool (ASCT) runs. Each ASCT run produces an output directory that contains benchmark results and system information.

You can use asct diff to carry out the following tasks:

• Compare the same system before and after a change. For example, compare results before and after a kernel update, firmware update, tuning change, or BIOS or UEFI configuration change.
• Compare 2 different systems that use the same benchmark set

What asct diff compares

Each ASCT run output directory contains a structured summary of the following information:

• Benchmark measurements
• System configuration information
• Metadata

When you run asct diff, ASCT performs the following tasks:

• Loads results from each run directory
• Groups fields by benchmark or measurement name
• Produces a table in standard output or machine-readable output in CSV or JSON format

How asct diff shows values

For benchmark measurements, ASCT prints the percentage difference between the baseline and each comparison run.

For system configuration reports such as system-info, cmn, and metadata, ASCT prints raw values instead of percentage differences.

If a field is missing in one run, ASCT displays <missing>. It reports the change as added or removed.

Run asct diff

Compare runs

To compare 2 runs, use:

# ASCT uses the first run directory as the baseline
asct diff runA runB

To compare 3 runs against the first run as the baseline, use:

# Compare 3 runs against the first run as the baseline
asct diff runA runB runC

Specify a baseline

To specify a baseline explicitly, use --baseline:

# Use runB as the baseline and compare runA against it
asct diff runA --baseline runB

# Use runA as the baseline and compare runB and runC against it
asct diff runB runC --baseline runA

Use asct diff options

The asct diff command supports the common ASCT output and logging options:

• --format with values stdout, csv, or json
• --output-dir, --force
• --log-level, --log-file, --quiet

In addition, asct diff supports the following options:

• --baseline, -b BASELINE_DIR

  • Specify the baseline run directory.
  • When you set this option, you can compare 1 or more run directories against the baseline. You do not need to pass 2 run directories to establish a baseline.

• --benchmarks, -k BENCH [BENCH ...]

  • Limit the comparison to selected benchmarks or reports.
  • Prefix a benchmark with ^ to exclude it from the comparison.

  Examples:

  # Compare only these reports
  asct diff runA runB --benchmarks system-info peak-bandwidth
  
  # Compare everything except system-info
  asct diff runA runB --benchmarks ^system-info

• --sort-by, -s RUN_DIR

  • Sort the main delta table by a specific comparison column.
  • RUN_DIR must match one of the run directory arguments. It must not be the baseline.

  Example:

  asct diff runA runB runC --sort-by runC

View asct diff output

Use --output-dir to specify an output directory.

When you use --format=csv, ASCT writes diff.csv to the output directory.

When you use --format=json, ASCT writes diff.json to the output directory.

In addition to tabular output, asct diff can generate comparison plots for supported benchmark types. ASCT writes these plot files to the diff output directory with the other generated artifacts.

Diff graphs

asct diff can generate PNG graphs in the diff output directory.

The images below show example graphs that compare 3 ASCT runs:

Example output

The following example shows abbreviated standard output from asct diff:

Summary
========================================
Stat                  | Value             
========================================
Baseline              | runA           
Total # of Benchmarks | 9                 
========================================

Benchmarks with Most Differences
========================================
Benchmark            | Differences       
========================================            
system-info          | 3                              
peak-bandwidth       | 2                                
metadata             | 2                 
========================================


Differences In System-Info
============================================================================================================================
field                                                   | runA                 | runB                 | runC             
============================================================================================================================
system-info.data.kern_cfg.build_dir                     | /lib/modules/6.14.0- | /lib/modules/6.11.0- | /lib/modules/6.2.0- 
                                                        | 28-generic/build     | rc1-atp+/build       | 36-generic/build    
----------------------------------------------------------------------------------------------------------------------------
system-info.data.kern_cfg.cfg_file                      | /boot/config-6.14.0- | /boot/config-6.11.0- | /boot/config-6.2.0- 
                                                        | 28-generic           | rc1-atp+             | 36-generic          
----------------------------------------------------------------------------------------------------------------------------
system-info.data.kern_cfg.distro                        | Ubuntu 24.04.3 LTS   | Ubuntu 22.04.5 LTS   | Ubuntu 22.04.3 LTS  
----------------------------------------------------------------------------------------------------------------------------


Measurement Delta Percentage Between runA And Comparison Runs
============================================================================================================================
field                                                   | runA                 | runB                 | runC             
============================================================================================================================
peak-bandwidth.data.All Reads.Peak BW [GB/s]            | 186.31               | 155.73%              | 11.86%              
----------------------------------------------------------------------------------------------------------------------------
peak-bandwidth.data.All Reads.% of Peak Theoretical     | 90.97                | -2.57%               | -10.5%              
----------------------------------------------------------------------------------------------------------------------------
peak-bandwidth.data.3:1 Reads-Writes.Peak BW [GB/s]     | 161.98               | 158.14%              | 20.11%              
----------------------------------------------------------------------------------------------------------------------------


Differences In Metadata
============================================================================================================================
field                                                   | runA                 | runB                 | runC             
============================================================================================================================
metadata.version                                        | 0.3.0.               | 0.3.2                | 0.3.1         
----------------------------------------------------------------------------------------------------------------------------

This output shows the following information:

• The selected baseline
• The number of benchmarks compared
• The benchmarks with the most differences
• Configuration differences across runs
• Percentage differences for benchmark measurements