2025-05-15 - 2026-05-15
Overview
183 issues closed from 1 user
Closed
#207 [PR #203] [MERGED] deps(workflows): bump actions/download-artifact from 7 to 8
Closed
#208 [PR #206] [MERGED] deps(workflows): bump docker/build-push-action from 6 to 7
Closed
#205 [PR #204] [MERGED] deps(workflows): bump docker/setup-buildx-action from 3 to 4
Closed
#206 [PR #205] [MERGED] deps(workflows): bump docker/login-action from 3 to 4
Closed
#204 [PR #202] [MERGED] deps(workflows): bump actions/attest-build-provenance from 3 to 4
Closed
#203 [PR #201] [MERGED] deps(workflows): bump actions/upload-artifact from 6 to 7
Closed
#201 [PR #196] [MERGED] deps(workflows): bump actions/download-artifact from 6 to 7
Closed
#202 [PR #197] [MERGED] deps(workflows): bump actions/upload-artifact from 5 to 6
Closed
#198 [PR #194] [MERGED] deps(workflows): bump actions/checkout from 4 to 6
Closed
#199 [PR #195] [MERGED] feat(api/device): support devices with unified memory
Closed
#196 [PR #191] [MERGED] feat(workflows/docker): add workflow to build docker images
Closed
#195 [PR #190] [MERGED] feat(tui/device): add bar charts for memory bandwidth and power usage
Closed
#192 [PR #182] [CLOSED] add support k8s
Closed
#193 [PR #187] [MERGED] deps(workflows): bump actions/download-artifact from 5 to 6
Closed
#194 [PR #188] [MERGED] deps(workflows): bump actions/upload-artifact from 4 to 5
Closed
#191 [PR #186] [MERGED] deps(nvidia-ml-py): add nvidia-ml-py 13.580.82 to support list
Closed
#190 [PR #184] [MERGED] deps(workflows): bump actions/setup-python from 5 to 6
Closed
#189 [PR #180] [MERGED] deps(workflows): bump actions/checkout from 4 to 5
Closed
#186 [PR #177] [MERGED] deps(workflows): bump actions/download-artifact from 4 to 5
Closed
#188 [PR #179] [MERGED] chore!: remove per-version install extras for nvidia-ml-py and prefer nvitop[cudaXX] instead
Closed
#187 [PR #176] [MERGED] fix(api/device): fix device.pcie_tx_throughput() returns PCIe RX throughput due to a typo in argument
Closed
#185 [PR #178] [MERGED] deps(nvidia-ml-py): add nvidia-ml-py 13.580.65 to support list
Closed
#184 [PR #172] [MERGED] fix(api/process): fix snapshot cache for GPU processes with shared host process
Closed
#182 [PR #164] [MERGED] lint(tui): add doctests and add type annotations in nvitop.tui
Closed
#183 [PR #166] [CLOSED] Adds workflow to publish nvitop-exporter
Closed
#181 [PR #165] [CLOSED] add GH actions build and publish nvitop-exporter
Closed
#180 [PR #163] [MERGED] fix(tui/host): ignore errors when collecting host metrics
Closed
#179 [PR #159] [MERGED] feat(exporter): add Dockerfile and docker-compose template for nvitop-exporter
Closed
#176 [PR #150] [MERGED] deps(python): drop Python 3.7 support
Closed
#178 [PR #157] [MERGED] chore(callbacks): deprecate nvitop.callbacks as officially unmaintained
Closed
#175 [PR #156] [MERGED] feat(tui): show %GMBW in main screen
Closed
#173 [PR #147] [MERGED] deps(cachetools): remove third-party dependency cachetools
Closed
#174 [PR #148] [MERGED] deps(termcolor): vendor third-party dependency termcolor
Closed
#170 [PR #132] [MERGED] fix(collector): fix documentation for ResourceMetricCollector.clear() function
Closed
#171 [PR #146] [MERGED] fix(api/device): fix passing invalid device handle to NVML functions
Closed
#172 [PR #138] [MERGED] feat(exporter): add dashboard example
Closed
#167 [PR #123] [CLOSED] Support for AMD ROCm devices
Closed
#168 [PR #129] [CLOSED] Import getpass in try-except block
Closed
#169 [PR #130] [MERGED] feat(api): handle exceptions for function getpass.getuser()
Closed
#166 [PR #115] [MERGED] deps(workflows): bump actions/download-artifact from 3 to 4
Closed
#165 [PR #116] [MERGED] deps(workflows): bump actions/upload-artifact from 3 to 4
Closed
#163 [PR #113] [CLOSED] Convert pytorch_lighting imports into lightning.pytorch imports
Closed
#164 [PR #114] [MERGED] fix(callbacks/lightning): populate callback for lightning
Closed
#162 [PR #111] [MERGED] deps(workflows): bump actions/setup-python from 4 to 5
Closed
#161 [PR #107] [MERGED] feat(exporter): remove metrics if process is gone
Closed
#160 [PR #101] [MERGED] deps(python): add Python 3.12 classifiers
Closed
#159 [PR #100] [MERGED] fix(libcuda): fix cuDeviceGetUuid() when the UUID contains 0x00
Closed
#156 [PR #94] [MERGED] fix(api/libnvml): fix upstream changes for process info v3 APIs on 535.104.05 driver
Closed
#157 [PR #92] [MERGED] feat(exporter): add Prometheus exporter
Closed
#158 [PR #96] [MERGED] deps(workflows): bump actions/checkout from 3 to 4
Closed
#155 [PR #89] [MERGED] fix(api/libnvml): fix removal for process info v3 APIs on the upstream 535.98 driver
Closed
#154 [PR #87] [MERGED] feat(api/device): add methods to query PCIe and NVLink throughput
Closed
#153 [PR #85] [MERGED] fix(api/device): use recent timestamp for GPU process utilization query
Closed
#151 [PR #84] [CLOSED] Update pytorch_lightning.py callback
Closed
#152 [PR #80] [MERGED] feat(api/collector): include last snapshot metrics in the log results
Closed
#150 [PR #79] [MERGED] fix(api/libnvml): fix process info support for NVIDIA R535 driver (CUDA 12.2+)
Closed
#149 [PR #73] [MERGED] feat(linter): mypy integration
Closed
#146 [PR #66] [MERGED] refactor(api): move TTLCache usage to CLI-only
Closed
#147 [PR #64] [CLOSED] chore(pre-commit): [pre-commit.ci] autoupdate
Closed
#148 [PR #70] [MERGED] fix(api/device): further isolate the CUDA_VISIBLE_DEVICE parser in a subprocess
Closed
#145 [PR #67] [MERGED] feat(cli): support float number as snapshot interval (>= 0.25s)
Closed
#142 [PR #57] [MERGED] feat(linter): ruff integration
Closed
#143 [PR #62] [CLOSED] chore(pre-commit): [pre-commit.ci] autoupdate
Closed
#144 [PR #63] [CLOSED] simple change to accept float numbers as interval
Closed
#140 [PR #59] [MERGED] feat(gui/host): show more metrics
Closed
#141 [PR #56] [MERGED] deps(python): drop Python 3.6 support
Closed
#137 [PR #52] [MERGED] feat(core/libcudart): add Python bindings for CUDA Runtime APIs
Closed
#138 [PR #53] [MERGED] docs(core): rephrase docstrings
Closed
#139 [PR #54] [MERGED] refactor(api): compact and rename top-level module
Closed
#136 [PR #51] [MERGED] fix(Dockerfile): migrate docker image repo to nvcr.io/nvidia/driver
Closed
#135 [PR #50] [CLOSED] [update] Fix build problem of 'Dockerfile'
Closed
#134 [PR #48] [MERGED] feat(core/collector): add function and method to collect metrics in background thread
Closed
#132 [PR #39] [MERGED] feat(gui/metrics): add process metrics screen
Closed
#133 [PR #42] [MERGED] feat(gui/messagebox): add message box
Closed
#130 [PR #31] [MERGED] feat(gui): add 256 color support for bar charts
Closed
#131 [PR #34] [MERGED] feat(gui): support process tagging
Closed
#129 [PR #30] [MERGED] feat(core/libnvml): add compatibility layers for NVML Python bindings
Closed
#127 [PR #26] [MERGED] feat(core/libcuda): add Python bindings for CUDA driver APIs
Closed
#128 [PR #28] [MERGED] feat(select): add CUDA visible devices selection tool
Closed
#126 [PR #27] [MERGED] feat(pyproject): migrate to pyproject.toml for packaging
Closed
#125 [PR #25] [MERGED] refactor(core/libnvml): use module-based approach rather than singleton
Closed
#124 [PR #24] [MERGED] Add extra options to specify nvidia-ml-py version
Closed
#121 [PR #15] [MERGED] Fix locale encoding when C.UTF-8 is not available
Closed
#122 [PR #22] [MERGED] Add detailed documentation for API references
Closed
#123 [PR #21] [MERGED] Async Status Collector for Logger Integration (e.g. CSV or TensorBoard)
Closed
#120 [PR #8] [MERGED] MIG device support
Closed
#117 [GH-ISSUE #193] [BUG] NVML cannot get device memory info for NVIDIA DGX Spark due to unified memory
Closed
#116 [GH-ISSUE #199] [Feature Request] nvitop compatible from unified memory
Closed
#112 [GH-ISSUE #189] [Feature Request] show POWER in graphs instead of utilization
Closed
#109 [GH-ISSUE #174] [Feature Request] PCIE transmission speed
Closed
#108 [GH-ISSUE #173] [BUG][upstream] used value of host memory is different from the result of free command
Closed
#106 [GH-ISSUE #171] [BUG] Shared GPU Job - listed CPU %
Closed
#104 [GH-ISSUE #168] [Feature Request] support jetson orin nano
Closed
#102 [GH-ISSUE #162] [Question] Can't use nvitop callback in pytorch lightning
Closed
#101 [GH-ISSUE #161] [BUG][Upstream] failed to get SWAP status on Windows with performance counter disabled
Closed
#97 [GH-ISSUE #155] [Feature Request] Add %GMBW in default monitor mode
Closed
#98 [GH-ISSUE #154] [BUG] _curses module not found
Closed
#99 [GH-ISSUE #158] [BUG] Unable to use monitor mode
Closed
#95 [GH-ISSUE #153] [BUG] Can't use pipx to run the latest nvitop as bug reporting page suggests
Closed
#94 [GH-ISSUE #151] [Feature Request] Add working directory of each process to processes table
Closed
#93 [GH-ISSUE #145] [BUG] GPU Unknown Error 会导致 as_snapshot() 调用触发段错误
Closed
#91 [GH-ISSUE #143] [Question] nvitop cannot be installed by uvx on Windows with Python 3.13
Closed
#88 [GH-ISSUE #141] [Question] Create new release for vllm compatibility
Closed
#89 [GH-ISSUE #140] [BUG] Release changelog URLs are incorrect
Closed
#85 [GH-ISSUE #137] [Question] How on earth does the sorting work?
Closed
#87 [GH-ISSUE #139] [BUG] Segmentation Fault when one GPU lost from PCIe bus
Closed
#82 [GH-ISSUE #133] [Feature Request] Nvitop-Exporter: Processes name & args (command)
Closed
#84 [GH-ISSUE #135] [Question] Missing font
Closed
#81 [GH-ISSUE #127] [BUG] UTF-8 Error during decoding device name on R555 driver
Closed
#78 [GH-ISSUE #124] [Feature Request] can you support word wrap for COMMAND information
Closed
#77 [GH-ISSUE #125] Windows download link please..
Closed
#73 [GH-ISSUE #122] [Question] Grafana Dashboard Example
Closed
#74 [GH-ISSUE #121] [BUG] Prometheus connection refused
Closed
#75 [GH-ISSUE #120] Installation: which step to follow?
Closed
#70 [GH-ISSUE #117] [Feature Request] Show real-time bandwidth under monitor mode
Closed
#71 [GH-ISSUE #119] [Question] How snapshot could be used
Closed
#72 [GH-ISSUE #118] [Question] ERROR: Failed to initialize curses (setupterm: could not find terminfo database)
Closed
#67 [GH-ISSUE #110] [Feature Request] add io stat for disk and process like glances
Closed
#68 [GH-ISSUE #112] [BUG] Pytorch lightning callback
Closed
#69 [GH-ISSUE #109] [Question] How to log GPU performance to wandb
Closed
#65 [GH-ISSUE #105] [Feature Request] It is recommended to change the dependency from nvidia-ml-py to pynvml
Closed
#66 [GH-ISSUE #106] [BUG][exporter] Process metrics still exist when the process is gone
Closed
#64 [GH-ISSUE #108] [BUG] curser error init display on ubuntu 22.10
Closed
#61 [GH-ISSUE #102] [Feature Request] Add shortcut that shows parent process's name for selected process
Closed
#62 [GH-ISSUE #103] [Question] Memory bandwidth utilization of GPUs?
Closed
#63 [GH-ISSUE #104] [BUG] Monitor mode displays nothing under Python 3.12 in Windows 11
Closed
#60 [GH-ISSUE #98] [BUG] Display issue when running in Docker environment
Closed
#58 [GH-ISSUE #97] [BUG] ERROR: Failed to initialize curses (curs_set() returned ERR) in tmux environment
Closed
#59 [GH-ISSUE #99] [BUG] nvitop.Device.from_cuda_visible_devices() not detecting GPU
Closed
#55 [GH-ISSUE #95] [Question] How to count the total amount of one user's GPU memory on all graphics cards by using nvitop?
Closed
#56 [GH-ISSUE #93] [BUG] Missing process list on NVIDIA 535.98 driver
Closed
#57 [GH-ISSUE #91] [Question] 建议修改 api/device.py line 2125 0.25s 改为 1s
Closed
#52 [GH-ISSUE #90] [BUG] nvidia-ml-py-12.535.77 兼容性問題
Closed
#53 [GH-ISSUE #86] [BUG] unable to display power limit
Closed
#54 [GH-ISSUE #88] [Bug] Processes information cannot be obtained normally on 535.98 driver
Closed
#49 [GH-ISSUE #82] [Feature Request] Add support for PCIE throughput and NVLink throughput
Closed
#50 [GH-ISSUE #83] [BUG] nvidia-smi pmon 和 nvitop -o 输出的 sm % 不一致且有较大差异
Closed
#48 [GH-ISSUE #78] [Question] Installation issue on NVIDIA Jetson AGX Orin (aarch64)
Closed
#47 [GH-ISSUE #77] [Question] live metrics collector
Closed
#46 [GH-ISSUE #76] [BUG] (Windows) nvitop lists no processes; OverflowError: Python int too large to convert to C long
Closed
#45 [GH-ISSUE #75] [BUG] PIDs are scrambled and No Such Process is printed since update to NVIDIA drivers
Closed
#44 [GH-ISSUE #72] [Feature Request] Add support for prometheus metrics exporter
Closed
#41 [GH-ISSUE #69] [Bug] Execution Freeze while using CudaDevice.count() in global scope
Closed
#40 [GH-ISSUE #68] [Feature Request] 需要一个GUI版本
Closed
#42 [GH-ISSUE #71] [Question] process.cpu_percent() got over 100% value
Closed
#39 [GH-ISSUE #61] [Question] Why fan speed will be set to 0 when nvitop is running
Closed
#37 [GH-ISSUE #60] [Question] Custom criteria for select_devices
Closed
#38 [GH-ISSUE #65] [Feature Request] Refresh rate < 1 sec
Closed
#34 [GH-ISSUE #58] [Feature Request] Show the real memory usage instead of just show the percentage
Closed
#35 [GH-ISSUE #55] [Question] Unable to view GPU memory usage in Windows (N/A memory usage)
Closed
#31 [GH-ISSUE #45] [Enhancement] Skip error gpus and show normal infos automatically
Closed
#32 [GH-ISSUE #47] [Feature Request] Collect metrics in a fixed interval for the lifespan of a training job
Closed
#33 [GH-ISSUE #46] Max memory clock methods point to current memory clock methods
Closed
#29 [GH-ISSUE #44] [Bug] Corrupted dependency of version 0.10.0 with pynvml
Closed
#30 [GH-ISSUE #41] NVML ERROR: RM has detected an NVML/RM version mismatch.
Closed
#28 [GH-ISSUE #43] 'python setup.py test' fails
Closed
#25 [GH-ISSUE #40] [Enhancement] Add confirm message box before sending signals to processes
Closed
#26 [GH-ISSUE #38] Failed to initialize curses (setupterm: could not find terminfo database)
Closed
#27 [GH-ISSUE #37] interval sec isn't work at linux
Closed
#24 [GH-ISSUE #33] [Question] nvitop runs non-dynamically inside slurm jobs
Closed
#22 [GH-ISSUE #36] [Feature Request] Graphs for per process metrics
Closed
#23 [GH-ISSUE #35] [Enhancement] Monitoring % of tensor cores
Closed
#21 [GH-ISSUE #29] [Enhancement] Backward compatible NVML Python bindings
Closed
#19 [GH-ISSUE #23] nvidia-ml-py version conflicts with other packages (e.g., gpustat)
Closed
#20 [GH-ISSUE #32] Feature request: --interval SEC support milli_second
Closed
#16 [GH-ISSUE #19] [Question] Which processes are using GPU memory?
Closed
#17 [GH-ISSUE #20] [Question] Can nvitop keep a log/record of GPU-Utilization and store in a CSV?
Closed
#18 [GH-ISSUE #18] [Feature Request] torch_geometric support
Closed
#13 [GH-ISSUE #14] [Bug] Usage chart working in centos linux.
Closed
#15 [GH-ISSUE #17] [Feature Request] Make snapshot interval configurable in monitor mode
Closed
#14 [GH-ISSUE #16] [Bug] display issue when running inside tmux
Closed
#11 [GH-ISSUE #12] [Bug] ModuleNotFoundError: No module named 'pwd'
Closed
#12 [GH-ISSUE #13] [Bug] gpu memory-usage not show right in driver 510 version
Closed
#10 [GH-ISSUE #11] nvitop command not found after installation
Closed
#8 [GH-ISSUE #10] How to make the color black and better colorful for nvitop?
Closed
#9 [GH-ISSUE #9] How to show hidden elements in monitor mode?
Closed
#7 [GH-ISSUE #7] Nvidia Driver used in Dockerfile contains multiple CVE's
Closed
#5 [GH-ISSUE #5] [Feature Request] MIG device support (e.g. A100 GPUs)
Closed
#6 [GH-ISSUE #4] Package conflicts with nvidia-ml-py (pynvml.py)
Closed
#4 [GH-ISSUE #6] Great capability, thank you!
Closed
#3 [GH-ISSUE #3] Rendering issues caused by locale settings
Closed
#1 [GH-ISSUE #1] Package conflicts with nvidia-ml-py (pynvml.py)
Closed
#2 [GH-ISSUE #2] nvitop command not found after installation
212 issues created by 1 user
Opened
#1 [GH-ISSUE #1] Package conflicts with nvidia-ml-py (pynvml.py)
Opened
#2 [GH-ISSUE #2] nvitop command not found after installation
Opened
#3 [GH-ISSUE #3] Rendering issues caused by locale settings
Opened
#4 [GH-ISSUE #6] Great capability, thank you!
Opened
#5 [GH-ISSUE #5] [Feature Request] MIG device support (e.g. A100 GPUs)
Opened
#6 [GH-ISSUE #4] Package conflicts with nvidia-ml-py (pynvml.py)
Opened
#7 [GH-ISSUE #7] Nvidia Driver used in Dockerfile contains multiple CVE's
Opened
#8 [GH-ISSUE #10] How to make the color black and better colorful for nvitop?
Opened
#9 [GH-ISSUE #9] How to show hidden elements in monitor mode?
Opened
#10 [GH-ISSUE #11] nvitop command not found after installation
Opened
#11 [GH-ISSUE #12] [Bug] ModuleNotFoundError: No module named 'pwd'
Opened
#12 [GH-ISSUE #13] [Bug] gpu memory-usage not show right in driver 510 version
Opened
#13 [GH-ISSUE #14] [Bug] Usage chart working in centos linux.
Opened
#14 [GH-ISSUE #16] [Bug] display issue when running inside tmux
Opened
#15 [GH-ISSUE #17] [Feature Request] Make snapshot interval configurable in monitor mode
Opened
#16 [GH-ISSUE #19] [Question] Which processes are using GPU memory?
Opened
#17 [GH-ISSUE #20] [Question] Can nvitop keep a log/record of GPU-Utilization and store in a CSV?
Opened
#18 [GH-ISSUE #18] [Feature Request] torch_geometric support
Opened
#19 [GH-ISSUE #23] nvidia-ml-py version conflicts with other packages (e.g., gpustat)
Opened
#20 [GH-ISSUE #32] Feature request: --interval SEC support milli_second
Opened
#21 [GH-ISSUE #29] [Enhancement] Backward compatible NVML Python bindings
Opened
#22 [GH-ISSUE #36] [Feature Request] Graphs for per process metrics
Opened
#23 [GH-ISSUE #35] [Enhancement] Monitoring % of tensor cores
Opened
#24 [GH-ISSUE #33] [Question] nvitop runs non-dynamically inside slurm jobs
Opened
#25 [GH-ISSUE #40] [Enhancement] Add confirm message box before sending signals to processes
Opened
#26 [GH-ISSUE #38] Failed to initialize curses (setupterm: could not find terminfo database)
Opened
#27 [GH-ISSUE #37] interval sec isn't work at linux
Opened
#28 [GH-ISSUE #43] 'python setup.py test' fails
Opened
#29 [GH-ISSUE #44] [Bug] Corrupted dependency of version 0.10.0 with pynvml
Opened
#30 [GH-ISSUE #41] NVML ERROR: RM has detected an NVML/RM version mismatch.
Opened
#31 [GH-ISSUE #45] [Enhancement] Skip error gpus and show normal infos automatically
Opened
#32 [GH-ISSUE #47] [Feature Request] Collect metrics in a fixed interval for the lifespan of a training job
Opened
#33 [GH-ISSUE #46] Max memory clock methods point to current memory clock methods
Opened
#34 [GH-ISSUE #58] [Feature Request] Show the real memory usage instead of just show the percentage
Opened
#35 [GH-ISSUE #55] [Question] Unable to view GPU memory usage in Windows (N/A memory usage)
Opened
#36 [GH-ISSUE #49] [BUG] Cannot gather infomation of the /XWayland process in WSLg
Opened
#37 [GH-ISSUE #60] [Question] Custom criteria for select_devices
Opened
#38 [GH-ISSUE #65] [Feature Request] Refresh rate < 1 sec
Opened
#39 [GH-ISSUE #61] [Question] Why fan speed will be set to 0 when nvitop is running
Opened
#40 [GH-ISSUE #68] [Feature Request] 需要一个GUI版本
Opened
#41 [GH-ISSUE #69] [Bug] Execution Freeze while using CudaDevice.count() in global scope
Opened
#42 [GH-ISSUE #71] [Question] process.cpu_percent() got over 100% value
Opened
#43 [GH-ISSUE #74] [Feature Request] Add support to AMD's ROCm GPU
Opened
#44 [GH-ISSUE #72] [Feature Request] Add support for prometheus metrics exporter
Opened
#45 [GH-ISSUE #75] [BUG] PIDs are scrambled and No Such Process is printed since update to NVIDIA drivers
Opened
#46 [GH-ISSUE #76] [BUG] (Windows) nvitop lists no processes; OverflowError: Python int too large to convert to C long
Opened
#47 [GH-ISSUE #77] [Question] live metrics collector
Opened
#48 [GH-ISSUE #78] [Question] Installation issue on NVIDIA Jetson AGX Orin (aarch64)
Opened
#49 [GH-ISSUE #82] [Feature Request] Add support for PCIE throughput and NVLink throughput
Opened
#50 [GH-ISSUE #83] [BUG] nvidia-smi pmon 和 nvitop -o 输出的 sm % 不一致且有较大差异
Opened
#51 [GH-ISSUE #81] [Feature Request] Make AppImage format or binary file
Opened
#52 [GH-ISSUE #90] [BUG] nvidia-ml-py-12.535.77 兼容性問題
Opened
#53 [GH-ISSUE #86] [BUG] unable to display power limit
Opened
#54 [GH-ISSUE #88] [Bug] Processes information cannot be obtained normally on 535.98 driver
Opened
#55 [GH-ISSUE #95] [Question] How to count the total amount of one user's GPU memory on all graphics cards by using nvitop?
Opened
#56 [GH-ISSUE #93] [BUG] Missing process list on NVIDIA 535.98 driver
Opened
#57 [GH-ISSUE #91] [Question] 建议修改 api/device.py line 2125 0.25s 改为 1s
Opened
#58 [GH-ISSUE #97] [BUG] ERROR: Failed to initialize curses (curs_set() returned ERR) in tmux environment
Opened
#59 [GH-ISSUE #99] [BUG] nvitop.Device.from_cuda_visible_devices() not detecting GPU
Opened
#60 [GH-ISSUE #98] [BUG] Display issue when running in Docker environment
Opened
#61 [GH-ISSUE #102] [Feature Request] Add shortcut that shows parent process's name for selected process
Opened
#62 [GH-ISSUE #103] [Question] Memory bandwidth utilization of GPUs?
Opened
#63 [GH-ISSUE #104] [BUG] Monitor mode displays nothing under Python 3.12 in Windows 11
Opened
#64 [GH-ISSUE #108] [BUG] curser error init display on ubuntu 22.10
Opened
#65 [GH-ISSUE #105] [Feature Request] It is recommended to change the dependency from nvidia-ml-py to pynvml
Opened
#66 [GH-ISSUE #106] [BUG][exporter] Process metrics still exist when the process is gone
Opened
#67 [GH-ISSUE #110] [Feature Request] add io stat for disk and process like glances
Opened
#68 [GH-ISSUE #112] [BUG] Pytorch lightning callback
Opened
#69 [GH-ISSUE #109] [Question] How to log GPU performance to wandb
Opened
#70 [GH-ISSUE #117] [Feature Request] Show real-time bandwidth under monitor mode
Opened
#71 [GH-ISSUE #119] [Question] How snapshot could be used
Opened
#72 [GH-ISSUE #118] [Question] ERROR: Failed to initialize curses (setupterm: could not find terminfo database)
Opened
#73 [GH-ISSUE #122] [Question] Grafana Dashboard Example
Opened
#74 [GH-ISSUE #121] [BUG] Prometheus connection refused
Opened
#75 [GH-ISSUE #120] Installation: which step to follow?
Opened
#76 [GH-ISSUE #126] [Feature Request] Add CPU Processes
Opened
#77 [GH-ISSUE #125] Windows download link please..
Opened
#78 [GH-ISSUE #124] [Feature Request] can you support word wrap for COMMAND information
Opened
#79 [GH-ISSUE #128] [BUG] Memory leaking for Nvitop instances inside docker container
Opened
#80 [GH-ISSUE #131] SSL for nvitop-exporter
Opened
#81 [GH-ISSUE #127] [BUG] UTF-8 Error during decoding device name on R555 driver
Opened
#82 [GH-ISSUE #133] [Feature Request] Nvitop-Exporter: Processes name & args (command)
Opened
#83 [GH-ISSUE #134] [Question] How to kill a procedure named no-such-process
Opened
#84 [GH-ISSUE #135] [Question] Missing font
Opened
#85 [GH-ISSUE #137] [Question] How on earth does the sorting work?
Opened
#86 [GH-ISSUE #136] [BUG] Windows Server Unexpectedly Shuts Down When Using Nvitop to Monitor GPU Usage
Opened
#87 [GH-ISSUE #139] [BUG] Segmentation Fault when one GPU lost from PCIe bus
Opened
#88 [GH-ISSUE #141] [Question] Create new release for vllm compatibility
Opened
#89 [GH-ISSUE #140] [BUG] Release changelog URLs are incorrect
Opened
#90 [GH-ISSUE #142] [Question] 可以打包个PC win的包吗。。
Opened
#91 [GH-ISSUE #143] [Question] nvitop cannot be installed by uvx on Windows with Python 3.13
Opened
#92 [GH-ISSUE #144] [BUG] Missing Windows Build on Conda Forge
Opened
#93 [GH-ISSUE #145] [BUG] GPU Unknown Error 会导致 as_snapshot() 调用触发段错误
Opened
#94 [GH-ISSUE #151] [Feature Request] Add working directory of each process to processes table
Opened
#95 [GH-ISSUE #153] [BUG] Can't use pipx to run the latest nvitop as bug reporting page suggests
Opened
#96 [GH-ISSUE #152] [BUG] Failure to correctly determine GPU-utilizing process details - "No such process"
Opened
#97 [GH-ISSUE #155] [Feature Request] Add %GMBW in default monitor mode
Opened
#98 [GH-ISSUE #154] [BUG] _curses module not found
Opened
#99 [GH-ISSUE #158] [BUG] Unable to use monitor mode
Opened
#100 [GH-ISSUE #160] [Feature Request] Hide CPU usage
Opened
#101 [GH-ISSUE #161] [BUG][Upstream] failed to get SWAP status on Windows with performance counter disabled
Opened
#102 [GH-ISSUE #162] [Question] Can't use nvitop callback in pytorch lightning
Opened
#103 [GH-ISSUE #167] [Feature Request] Webpage for history visualization
Opened
#104 [GH-ISSUE #168] [Feature Request] support jetson orin nano
Opened
#105 [GH-ISSUE #169] [Feature Request] 是否考虑出支持K8S的版本？
Opened
#106 [GH-ISSUE #171] [BUG] Shared GPU Job - listed CPU %
Opened
#107 [GH-ISSUE #170] [Feature Request] Hide CPU Section in Terminal
Opened
#108 [GH-ISSUE #173] [BUG][upstream] used value of host memory is different from the result of free command
Opened
#109 [GH-ISSUE #174] [Feature Request] PCIE transmission speed
Opened
#110 [GH-ISSUE #181] [BUG] PID out of range due to API change of NVIDIA R525 driver
Opened
#111 [GH-ISSUE #175] [BUG] The conda distribution has TERM-related problems
Opened
#112 [GH-ISSUE #189] [Feature Request] show POWER in graphs instead of utilization
Opened
#113 [GH-ISSUE #183] Windows object has no attribute '_win' do you mean mvwin?
Opened
#114 [GH-ISSUE #185] [Feature Request] Provide Helm Chart / Kubernetes Manifests for nvitop-exporter
Opened
#115 [GH-ISSUE #200] [Feature Request] Option to display MIG UUID of MIG GPUs in main monitor display
Opened
#116 [GH-ISSUE #199] [Feature Request] nvitop compatible from unified memory
Opened
#117 [GH-ISSUE #193] [BUG] NVML cannot get device memory info for NVIDIA DGX Spark due to unified memory
Opened
#118 [GH-ISSUE #212] [BUG] Device methods not throwing exceptions
Opened
#119 [GH-ISSUE #211] [RFC] AMD GPU Support
Opened
#120 [PR #8] [MERGED] MIG device support
Opened
#121 [PR #15] [MERGED] Fix locale encoding when C.UTF-8 is not available
Opened
#122 [PR #22] [MERGED] Add detailed documentation for API references
Opened
#123 [PR #21] [MERGED] Async Status Collector for Logger Integration (e.g. CSV or TensorBoard)
Opened
#124 [PR #24] [MERGED] Add extra options to specify nvidia-ml-py version
Opened
#125 [PR #25] [MERGED] refactor(core/libnvml): use module-based approach rather than singleton
Opened
#126 [PR #27] [MERGED] feat(pyproject): migrate to pyproject.toml for packaging
Opened
#127 [PR #26] [MERGED] feat(core/libcuda): add Python bindings for CUDA driver APIs
Opened
#128 [PR #28] [MERGED] feat(select): add CUDA visible devices selection tool
Opened
#129 [PR #30] [MERGED] feat(core/libnvml): add compatibility layers for NVML Python bindings
Opened
#130 [PR #31] [MERGED] feat(gui): add 256 color support for bar charts
Opened
#131 [PR #34] [MERGED] feat(gui): support process tagging
Opened
#132 [PR #39] [MERGED] feat(gui/metrics): add process metrics screen
Opened
#133 [PR #42] [MERGED] feat(gui/messagebox): add message box
Opened
#134 [PR #48] [MERGED] feat(core/collector): add function and method to collect metrics in background thread
Opened
#135 [PR #50] [CLOSED] [update] Fix build problem of 'Dockerfile'
Opened
#136 [PR #51] [MERGED] fix(Dockerfile): migrate docker image repo to nvcr.io/nvidia/driver
Opened
#137 [PR #52] [MERGED] feat(core/libcudart): add Python bindings for CUDA Runtime APIs
Opened
#138 [PR #53] [MERGED] docs(core): rephrase docstrings
Opened
#139 [PR #54] [MERGED] refactor(api): compact and rename top-level module
Opened
#140 [PR #59] [MERGED] feat(gui/host): show more metrics
Opened
#141 [PR #56] [MERGED] deps(python): drop Python 3.6 support
Opened
#142 [PR #57] [MERGED] feat(linter): ruff integration
Opened
#143 [PR #62] [CLOSED] chore(pre-commit): [pre-commit.ci] autoupdate
Opened
#144 [PR #63] [CLOSED] simple change to accept float numbers as interval
Opened
#145 [PR #67] [MERGED] feat(cli): support float number as snapshot interval (>= 0.25s)
Opened
#146 [PR #66] [MERGED] refactor(api): move TTLCache usage to CLI-only
Opened
#147 [PR #64] [CLOSED] chore(pre-commit): [pre-commit.ci] autoupdate
Opened
#148 [PR #70] [MERGED] fix(api/device): further isolate the CUDA_VISIBLE_DEVICE parser in a subprocess
Opened
#149 [PR #73] [MERGED] feat(linter): mypy integration
Opened
#150 [PR #79] [MERGED] fix(api/libnvml): fix process info support for NVIDIA R535 driver (CUDA 12.2+)
Opened
#151 [PR #84] [CLOSED] Update pytorch_lightning.py callback
Opened
#152 [PR #80] [MERGED] feat(api/collector): include last snapshot metrics in the log results
Opened
#153 [PR #85] [MERGED] fix(api/device): use recent timestamp for GPU process utilization query
Opened
#154 [PR #87] [MERGED] feat(api/device): add methods to query PCIe and NVLink throughput
Opened
#155 [PR #89] [MERGED] fix(api/libnvml): fix removal for process info v3 APIs on the upstream 535.98 driver
Opened
#156 [PR #94] [MERGED] fix(api/libnvml): fix upstream changes for process info v3 APIs on 535.104.05 driver
Opened
#157 [PR #92] [MERGED] feat(exporter): add Prometheus exporter
Opened
#158 [PR #96] [MERGED] deps(workflows): bump actions/checkout from 3 to 4
Opened
#159 [PR #100] [MERGED] fix(libcuda): fix cuDeviceGetUuid() when the UUID contains 0x00
Opened
#160 [PR #101] [MERGED] deps(python): add Python 3.12 classifiers
Opened
#161 [PR #107] [MERGED] feat(exporter): remove metrics if process is gone
Opened
#162 [PR #111] [MERGED] deps(workflows): bump actions/setup-python from 4 to 5
Opened
#163 [PR #113] [CLOSED] Convert pytorch_lighting imports into lightning.pytorch imports
Opened
#164 [PR #114] [MERGED] fix(callbacks/lightning): populate callback for lightning
Opened
#165 [PR #116] [MERGED] deps(workflows): bump actions/upload-artifact from 3 to 4
Opened
#166 [PR #115] [MERGED] deps(workflows): bump actions/download-artifact from 3 to 4
Opened
#167 [PR #123] [CLOSED] Support for AMD ROCm devices
Opened
#168 [PR #129] [CLOSED] Import getpass in try-except block
Opened
#169 [PR #130] [MERGED] feat(api): handle exceptions for function getpass.getuser()
Opened
#170 [PR #132] [MERGED] fix(collector): fix documentation for ResourceMetricCollector.clear() function
Opened
#171 [PR #146] [MERGED] fix(api/device): fix passing invalid device handle to NVML functions
Opened
#172 [PR #138] [MERGED] feat(exporter): add dashboard example
Opened
#173 [PR #147] [MERGED] deps(cachetools): remove third-party dependency cachetools
Opened
#174 [PR #148] [MERGED] deps(termcolor): vendor third-party dependency termcolor
Opened
#175 [PR #156] [MERGED] feat(tui): show %GMBW in main screen
Opened
#176 [PR #150] [MERGED] deps(python): drop Python 3.7 support
Opened
#177 [PR #149] refactor(tui): drop dependency windows-curses
Opened
#178 [PR #157] [MERGED] chore(callbacks): deprecate nvitop.callbacks as officially unmaintained
Opened
#179 [PR #159] [MERGED] feat(exporter): add Dockerfile and docker-compose template for nvitop-exporter
Opened
#180 [PR #163] [MERGED] fix(tui/host): ignore errors when collecting host metrics
Opened
#181 [PR #165] [CLOSED] add GH actions build and publish nvitop-exporter
Opened
#182 [PR #164] [MERGED] lint(tui): add doctests and add type annotations in nvitop.tui
Opened
#183 [PR #166] [CLOSED] Adds workflow to publish nvitop-exporter
Opened
#184 [PR #172] [MERGED] fix(api/process): fix snapshot cache for GPU processes with shared host process
Opened
#185 [PR #178] [MERGED] deps(nvidia-ml-py): add nvidia-ml-py 13.580.65 to support list
Opened
#186 [PR #177] [MERGED] deps(workflows): bump actions/download-artifact from 4 to 5
Opened
#187 [PR #176] [MERGED] fix(api/device): fix device.pcie_tx_throughput() returns PCIe RX throughput due to a typo in argument
Opened
#188 [PR #179] [MERGED] chore!: remove per-version install extras for nvidia-ml-py and prefer nvitop[cudaXX] instead
Opened
#189 [PR #180] [MERGED] deps(workflows): bump actions/checkout from 4 to 5
Opened
#190 [PR #184] [MERGED] deps(workflows): bump actions/setup-python from 5 to 6
Opened
#191 [PR #186] [MERGED] deps(nvidia-ml-py): add nvidia-ml-py 13.580.82 to support list
Opened
#192 [PR #182] [CLOSED] add support k8s
Opened
#193 [PR #187] [MERGED] deps(workflows): bump actions/download-artifact from 5 to 6
Opened
#194 [PR #188] [MERGED] deps(workflows): bump actions/upload-artifact from 4 to 5
Opened
#195 [PR #190] [MERGED] feat(tui/device): add bar charts for memory bandwidth and power usage
Opened
#196 [PR #191] [MERGED] feat(workflows/docker): add workflow to build docker images
Opened
#197 [PR #192] feat: add Kubernetes integration with monitoring POD resources
Opened
#198 [PR #194] [MERGED] deps(workflows): bump actions/checkout from 4 to 6
Opened
#199 [PR #195] [MERGED] feat(api/device): support devices with unified memory
Opened
#200 [PR #198] support mthreads gpu monitoring
Opened
#201 [PR #196] [MERGED] deps(workflows): bump actions/download-artifact from 6 to 7
Opened
#202 [PR #197] [MERGED] deps(workflows): bump actions/upload-artifact from 5 to 6
Opened
#203 [PR #201] [MERGED] deps(workflows): bump actions/upload-artifact from 6 to 7
Opened
#204 [PR #202] [MERGED] deps(workflows): bump actions/attest-build-provenance from 3 to 4
Opened
#205 [PR #204] [MERGED] deps(workflows): bump docker/setup-buildx-action from 3 to 4
Opened
#206 [PR #205] [MERGED] deps(workflows): bump docker/login-action from 3 to 4
Opened
#207 [PR #203] [MERGED] deps(workflows): bump actions/download-artifact from 7 to 8
Opened
#208 [PR #206] [MERGED] deps(workflows): bump docker/build-push-action from 6 to 7
Opened
#209 [PR #207] Add gpu-stats.py nvitop logging script
Opened
#210 [PR #209] feat(api): add GDDR6/GDDR6X VRAM temperature reading via BAR0 registers
Opened
#211 [PR #210] Add mx-smi backend support for MetaX GPUs
Opened
#212 [PR #208] Fix incorrect memory reporting on coherent UMA platforms (GB10 / DGX …