# instinct > Note: ROCm documentation is split across multiple projects. In addition to this file, each project publishes its own `llms.txt` and `llms-full.txt` under `https:///projects//en/latest/`. ## Docs - [Computer Vision](https://instinct.docs.amd.com/latest/vision/index.html) - [Models and applications](https://instinct.docs.amd.com/latest/vision/ai.html) - [Decoding](https://instinct.docs.amd.com/latest/vision/decode.html) - [Image processing](https://instinct.docs.amd.com/latest/vision/preprocess.html) - [Data Science (ROCm-DS)](https://instinct.docs.amd.com/latest/data-science/index.html) - [hipDF](https://instinct.docs.amd.com/latest/data-science/hipDF.html) - [hipGRAPH](https://instinct.docs.amd.com/latest/data-science/hipGRAPH.html) - [hipVS](https://instinct.docs.amd.com/latest/data-science/hipVS.html) - [hipMM](https://instinct.docs.amd.com/latest/data-science/hipMM.html) - [hipRAFT](https://instinct.docs.amd.com/latest/data-science/hipRAFT.html) - [Life Science (ROCm-LS)](https://instinct.docs.amd.com/latest/life-science/index.html) - [hipCIM](https://instinct.docs.amd.com/latest/life-science/hipCIM.html) - [MONAI](https://instinct.docs.amd.com/latest/life-science/MONAI.html) - [Finance (ROCm-Finance)](https://instinct.docs.amd.com/latest/finance/index.html) - [XGBoost](https://instinct.docs.amd.com/latest/finance/xgboost.html) - [LightGBM](https://instinct.docs.amd.com/latest/finance/lightgbm.html) - [ThunderGBM](https://instinct.docs.amd.com/latest/finance/thundergbm.html) - [Simulation & Modeling Apps](https://instinct.docs.amd.com/latest/isv-apps/index.html) - [Fluent](https://instinct.docs.amd.com/latest/isv-apps/ansys-fluent.html) - [Mechanical](https://instinct.docs.amd.com/latest/isv-apps/ansys-mechanical.html) - [Cadence Fidelity](https://instinct.docs.amd.com/latest/isv-apps/cadence-fidelity.html) - [DevitoPRO](https://instinct.docs.amd.com/latest/isv-apps/devito.html) - [Siemens](https://instinct.docs.amd.com/latest/isv-apps/siemens.html) - [Stone Ridge](https://instinct.docs.amd.com/latest/isv-apps/stone-ridge.html) - [GSplat](https://instinct.docs.amd.com/latest/simulation/gsplat.html) - [Bare Metal](https://instinct.docs.amd.com/latest/system-admin/bare-metal.html) - [AMD GPU Driver](https://instinct.docs.amd.com/projects/amdgpu-docs/en/latest/) - [Containers and Orchestration Tools](https://instinct.docs.amd.com/latest/system-admin/co-tools.html) - [GPU-Operator](https://instinct.docs.amd.com/projects/gpu-operator/en/latest/) - [Network-Operator](https://instinct.docs.amd.com/projects/network-operator/en/main/) - [K8s Device Plugin](https://instinct.docs.amd.com/projects/k8s-device-plugin/en/latest/) - [Device Metrics Exporter](https://instinct.docs.amd.com/projects/device-metrics-exporter/en/latest/) - [AMD Container Toolkit](https://instinct.docs.amd.com/projects/container-toolkit/en/latest/) - [Spur](https://instinct.docs.amd.com/projects/spur/en/latest/) - [Cluster](https://instinct.docs.amd.com/latest/system-admin/cluster.html) - [Instinct Customer Acceptance Guide](https://instinct.docs.amd.com/projects/system-acceptance/en/latest/) - [Networking](https://instinct.docs.amd.com/projects/gpu-cluster-networking/en/latest/) - [MI3XX Reference Design](https://instinct.docs.amd.com/projects/MI3XX-reference/latest/overview.html) - [Cloud](https://instinct.docs.amd.com/latest/system-admin/cloud.html) - [Instinct on Azure](https://instinct.docs.amd.com/projects/instinct-azure/latest/) - [Virtualization](https://instinct.docs.amd.com/latest/system-admin/virtualization.html) - [Virtualization Driver](https://instinct.docs.amd.com/projects/virt-drv/en/latest/) - [AMD SMI Documentation](https://instinct.docs.amd.com/projects/amd-smi-virt/en/latest/) - [Tools](https://instinct.docs.amd.com/latest/resources/index.html) - [Enterprise AI](https://account.amd.com/en/forms/registration/enterpriseai-ea.html) - [Omnistat](https://amdresearch.github.io/omnistat/) - [Cluster Validation Suite](https://rocm.docs.amd.com/projects/cvs/en/latest/index.html) - [AMD SMI](https://rocm.docs.amd.com/projects/amdsmi/en/latest/) - [ROCmValidationSuite](https://rocm.docs.amd.com/projects/ROCmValidationSuite/en/latest/) - [Common Reference](https://instinct.docs.amd.com/latest/resources/common-reference.html) - [GPU Partitioning](https://rocm.blogs.amd.com/software-tools-optimization/compute-memory-modes/README.html) - [Instinct Micro-architecture](https://rocm.docs.amd.com/en/latest/reference/gpu-arch/index.html) - [AMD SMI API Doc](https://rocm.docs.amd.com/projects/amdsmi/en/latest/index.html) - [HIP C++](https://rocm.docs.amd.com/projects/HIP/en/latest/index.html) - [OpenMP](https://rocm.docs.amd.com/projects/llvm-project/en/latest/conceptual/openmp.html) - [AMD Technical Information Portal](https://docs.amd.com/v/u/en-US/ug1729-amd-instinct-accelerators) --- Source: https://instinct.docs.amd.com/latest/index.html # AMD Instinct Data Center GPU Documentation The AMD Instinct Documentation site provides comprehensive guides and technical documentation for system administrators and technical users deploying AMD Instinct Data Center GPUs in enterprise environments. This site focuses on large-scale deployment, cluster management, monitoring, and operational best practices for both HPC and AI workloads. For API documentation and core software stack details, visit the [ROCm documentation](https://rocm.docs.amd.com). --- Source: https://instinct.docs.amd.com/latest/vision/index.html # Computer Vision Overview ![](images/models.png)![](images/decoding.png)![](images/processing.png) --- Source: https://instinct.docs.amd.com/latest/vision/ai.html # AI models and applications [MIVisionX](https://github.com/ROCm/MIVisionX) is a comprehensive toolkit consisting of a set of computer vision and machine intelligence libraries, utilities, and applications which enable you to build a wide variety of applications and models. MIVisionX delivers a highly optimized implementation of the [Khronos OpenVX Extensions](https://www.khronos.org/openvx/) along with Convolutional Neural Network model compilers and optimizers supporting [ONNX](https://onnx.ai/) and [NNEF](https://www.khronos.org/nnef) exchange formats. This toolkit enables the rapid prototyping and deployment of optimized computer vision and machine learning inference workloads on AMD Instinct GPUs. The MIVisionX toolkit provides you with all the necessary tools you need throughout the whole neural network life-cycle. Use these tools to design, develop, quantize, prune, retrain, and infer your neural network on the powerful AMD Instinct family of GPUs. ## Documentation For full documentation, refer to the ROCm docs site: - [MIVisionX](https://rocm.docs.amd.com/projects/MIVisionX/en/latest/index.html) --- Source: https://instinct.docs.amd.com/latest/vision/decode.html # Video and Image decoding Leverage the power of AMD Instinct GPUs to quickly decode a wide variety of image and video formats. Eliminate bottlenecks with these libraries and accelerate your training, finetuning, and inference workloads by ensuring quick and easy access to your data. ## Video decoding Decode videos for use in your computer vision workloads with [rocDecode](https://github.com/ROCm/rocDecode) or its Python bindings, [rocPyDecode](https://github.com/ROCm/rocPyDecode). These libraries support the H.264, H.265, AV1, and VP9 video codecs and enable video decoding on AMD Instinct GPUs. ## Image decoding The [rocJPEG](https://github.com/ROCm/rocJPEG) library enables fast and efficient JPEG decoding on AMD GPUs. This library includes the ability to run on batches to quickly decode large numbers of images simultaneously, fully utilizing the advantages offered by AMD Instinct GPUs. ## Documentation For full documentation, refer to the ROCm docs site: - [rocDecode](https://rocm.docs.amd.com/projects/rocDecode/en/latest/) - [rocPyDecode](https://rocm.docs.amd.com/projects/rocPyDecode/en/latest/index.html) - [rocJPEG](https://rocm.docs.amd.com/projects/rocJPEG/en/latest/) --- Source: https://instinct.docs.amd.com/latest/vision/preprocess.html # Image Preprocessing Eliminate the bottleneck caused by slow CPU image processing to further accelerate your computer vision workloads on AMD Instinct GPUs with [rocAL](https://github.com/ROCm/rocAL) and [RPP](https://github.com/ROCm/rpp). These libraries allow you to perform a wide variety of well known computer vision operations with GPU acceleration to augment and transform your images, preparing them for your deep learning workloads. ## ROCm Augmentation Library The [ROCm Augmentation Library (rocAL)](https://github.com/ROCm/rocAL) enables efficient loading and data preprocessing for deep learning applications. This allows for the creation of efficient computer vision pipelines that fully utilize the power of AMD Instinct GPUs. rocAL is designed to efficiently decode and process image and video from a variety of storage formats, saving you time that could be better spent training or running your models. ## ROCm Performance Primitives The [ROCm Performance Primitives (RPP)](https://github.com/ROCm/rpp) library contains a collection of high-performance computer vision operations with HIP, OpenCL, or CPU backends. This library enables you to efficiently apply augmentations, statistical functions, geometric and morphological transformations, popular filters, color model conversions, or other popular computer vision operations to preprocess your images for all of your computer vision workloads. ## Documentation For full documentation, refer to the ROCm docs site: - [rocAL](https://rocm.docs.amd.com/projects/rocAL/en/latest/) - [RPP](https://rocm.docs.amd.com/projects/rpp/en/latest/) --- Source: https://instinct.docs.amd.com/latest/data-science/index.html # ROCm-DS: The AMD ROCm Data Science Toolkit Unlock the future of data science with the AMD ROCm™ Data Science Toolkit (ROCm-DS), an innovative open-source toolkit built on the powerful ROCm platform. Tap into the unparalleled speed and efficiency of AMD Instinct™ GPUs, as ROCm-DS provides you with all of the necessary tools to tackle larger datasets and execute complex data science workloads with lightning speed. Transform your data science capabilities to accelerate both new and existing projects, and boost performance and productivity across your business. With ROCm-DS you have access to a rapidly expanding set of tools, empowering you to build, manage, and run entire data science workflows directly on AMD Instinct GPUs. Say goodbye to data-loading bottlenecks in AI workloads, accelerate your traditional data processing and analysis tasks, and manipulate immense datasets quickly and efficiently in GPU memory. Experience the exhilaration of streamlined AI and data science workloads powered by cutting-edge GPU technology. Behind ROCm-DS lies a passionate, driven team dedicated to continuously expanding and improving this toolkit, ensuring you have everything you need to push the boundaries of data science. Get ready to revolutionize your data processing applications and unleash new possibilities with ROCm-DS. ![](images/hipDF.jpg)![](images/hipGRAPH.jpg)![](images/hipVS.jpg)![](images/hipMM.jpg)![](images/hipRAFT.jpg)![](images/ROCm-DS.jpg)![](images/ROCm-DS_Blogs.jpg) --- Source: https://instinct.docs.amd.com/latest/data-science/hipDF.html # hipDF hipDF enables GPU accelerated DataFrames and DataFrame operations. This library, built on top of ROCm™, enables large-scale data processing on AMD Instinct™ GPUs, allowing you to perform many data manipulation operations at breakneck speeds. Built on the familiar [Apache Arrow](https://arrow.apache.org/) memory format and using APIs similar to the well known Python [Pandas library](https://pandas.pydata.org/), hipDF allows you to not only build accelerated data processing workloads, but also accelerate your existing Pandas applications with minimal effort. By adopting the well-known cuDF API on AMD hardware, hipDF ensures compatibility and ease of use across various computing environments. This API compatibility enables existing cuDF workloads to be effortlessly transitioned to run on supported AMD devices, allowing you to use the ROCm platform for all of your data processing tasks. ![](images/ROCm-DS_Docs.jpg)![](images/hipDF.jpg)![](images/ROCm-DS_Blogs.jpg) --- Source: https://instinct.docs.amd.com/latest/data-science/hipGRAPH.html # hipGRAPH hipGRAPH enables GPU accelerated complex networks and graphs, and contains a set of well-known graph algorithms. This library, built on top of ROCm™, enables you to build, analyze, and otherwise manipulate complex graphs on AMD Instinct™ GPUs. hipGRAPH is currently in an Early Access state. Running production workloads is not recommended. ![](images/ROCm-DS_Docs.jpg)![](images/hipGRAPH.jpg)![](images/ROCm-DS_Blogs.jpg) --- Source: https://instinct.docs.amd.com/latest/data-science/hipVS.html # hipVS hipVS enables GPU-accelerated vector search operations on AMD Instinct™ GPUs. As a part of the AMD ROCm™ Data Science Toolkit (ROCm-DS), hipVS builds upon the core ROCm libraries to accelerate operations such as approximate and exact nearest neighbor algorithms, as well as a variety of clustering algorithms. hipVS seamlessly integrates with other ROCm-DS libraries such as hipDF, enabling DataFrames to be used for vector search operations. ![](images/ROCm-DS_Docs.jpg)![](images/hipVS.jpg)![](images/ROCm-DS_Blogs.jpg) --- Source: https://instinct.docs.amd.com/latest/data-science/hipMM.html # hipMM HIP Memory Manager (hipMM) provides advanced GPU memory management utilities for a variety of libraries in the AMD ROCm™ Data Science Toolkit (ROCm-DS). hipMM focuses on improving memory usage efficiency for workloads that leverage hipDF, hipGRAPH, hipVS, and hipRAFT. ![](images/ROCm-DS_Docs.jpg)![](images/hipMM.jpg) --- Source: https://instinct.docs.amd.com/latest/data-science/hipRAFT.html # hipRAFT hipRAFT contains a variety of fundamental primitives and algorithms for machine learning and data mining workloads. As part of AMD ROCm™ Data Science Toolkit (ROCm-DS), it provides functionality for other ROCm-DS libraries, including hipDF, hipGRAPH, and hipVS. ![](images/ROCm-DS_Docs.jpg)![](images/hipRAFT.jpg) --- Source: https://instinct.docs.amd.com/latest/life-science/index.html # ROCm-LS: ROCm Toolkit for Life Sciences Unlock the transformative power of the ROCm Life Science Toolkit (ROCm-LS), an innovative and robust open-source toolkit built on top of the powerful ROCm™ platform. Designed with the medical and life sciences fields in mind, ROCm-LS maximizes the computational capabilities of AMD Instinct™ GPUs, enabling you to execute both new and existing workloads with exceptional speed and efficiency. Be a part of the revolution in life sciences research powered by GPUs with the rapidly expanding ROCm-LS suite of tools, empowering researchers and medical professionals to tackle critical tasks and solve real-world challenges. Fully utilize ROCm-LS to push the boundaries of what is possible in medical and life science research and to accelerate critical life-saving tasks. Dive into a future where medical breakthroughs are accelerated with precision, and solutions to humanity’s most pressing health challenges are within arm’s reach. With AMD’s powerful Instinct GPUs backing your workloads, ROCm-LS empowers innovation that has the potential to change lives. Join the forefront of scientific innovation with ROCm-LS and realize the full potential of your research endeavors today! ![](images/hipCIM.jpg)![](images/MONAI.jpg)![](images/ROCm-LS.jpg)![](images/ROCm-LS_Docs.jpg)![](images/ROCm-LS_Blogs.jpg) --- Source: https://instinct.docs.amd.com/latest/life-science/hipCIM.html # hipCIM hipCIM enables GPU accelerated image processing and computer vision operations for medical and life science images. This library, built on top of ROCm™, enables the processing and analysis of multidimensional images used in life sciences research and medical contexts. hipCIM is currently in an Early Access state and is intended as a preview to the production ready release planned for later this year. ![](images/ROCm-LS_Docs.jpg)![](images/hipCIM.jpg) --- Source: https://instinct.docs.amd.com/latest/life-science/MONAI.html # MONAI The Medical Open Network for Artificial Intelligence ([MONAI](https://monai.io/)) is a domain-optimized, open-source framework based on PyTorch designed to facilitate deep learning for medical images. The [MONAI Model Zoo](https://monai.io/model-zoo.html#/) contains a variety of biomedical imaging models that may be deployed or fine-tuned to suit your purposes. MONAI provides out-of-the-box integration with hipCIM, enabling researcher and healthcare professionals to streamline scientific imaging pipelines, boost computation performance, and speed up innovation across a wide array of healthcare use cases. MONAI support for AMD is currently in an Early Access state and is intended as a preview to the upcoming production ready release. ![](images/ROCm-LS_Docs.jpg)![](images/MONAI.jpg) --- Source: https://instinct.docs.amd.com/latest/finance/index.html # ROCm-Finance: ROCm toolkit for finance ROCm-Finance pulls the trajectory of tomorrow into today: an open toolkit on the [ROCm](https://rocm.docs.amd.com/) stack that delivers GPU-native gradient-boosting stacks that the industry already trusts. XGBoost, LightGBM, and ThunderGBM, tuned for [AMD Instinct](https://www.amd.com/en/products/accelerators/instinct.html) accelerators, so training, scoring, and simulation work land closer to real time than the CPU-era baselines could achieve. ROCm-Finance collapses the distance between signal and decision. The same workloads that once queued overnight now run in minutes. Risk, fraud detection, forecasting, and simulation pipelines step into the high-bandwidth GPU computing ROCm was built to serve. ROCm-Finance provides production-oriented kernels, memory paths, and scaling behavior so your boosting jobs feel like they arrived from the next generation, even on this week’s cluster. For more information on ROCm-Finance, including comparisons, prerequisites, installation, and deep API reference, see the [ROCm-Finance documentation](https://rocm.docs.amd.com/projects/rocm-finance/en/latest/index.html). ![](images/finance-1.png)![](images/finance-2.png)![](images/finance-3.png)![](images/finance-4.png)![](images/finance-5.png)![](images/finance-6.png) --- Source: https://instinct.docs.amd.com/latest/finance/xgboost.html # XGBoost (ROCm-Finance) XGBoost is the general-purpose engine in ROCm-Finance: tabular risk, fraud detection, pricing-side features, and trading-adjacent workloads. Use it when you want a familiar level-wise boosting path with broad finance coverage and a straightforward on-ramp from yesterday’s pipelines to tomorrow’s throughput. ![](images/finance-6.png)![](images/finance-1.png) --- Source: https://instinct.docs.amd.com/latest/finance/lightgbm.html # LightGBM (ROCm-Finance) LightGBM is how ROCm-Finance answers scale: leaf-wise training that shines when dataset size—wide feature stores, long histories, dense microstructure matrices—would otherwise push decisions into the next shift. On Instinct, that wall between overnight queues and minutes thins out; the same boosting idiom, routed through ROCm’s high-bandwidth, multi-GPU environment. Reach for LightGBM when volume is the bottleneck, and you still want gradient boosting semantics with GPU-native backing. ![](images/finance-6.png)![](images/finance-2.png) --- Source: https://instinct.docs.amd.com/latest/finance/thundergbm.html # ThunderGBM (ROCm-Finance) Use ThunderGBM when parallelism and raw throughput dominate the story: massively parallel trees, simulation-scale batches, and scenario grids that want the accelerator to do the heavy lifting today—not in some speculative later hardware generation. ThunderGBM collapses the signal-to-decision distance, optimized for highly parallel, GPU-intensive training runs on [AMD Instinct](https://www.amd.com/en/products/accelerators/instinct.html) silicon. ![](images/finance-6.png)![](images/finance-3.png) --- Source: https://instinct.docs.amd.com/latest/isv-apps/index.html # Simulation & Modeling Apps ![](isv-apps/images/AnsysFluent-tile.png)![](isv-apps/images/AnsysMech-tile.png)![](isv-apps/images/FCharLES-tile.png)![](isv-apps/images/DevitoPRO-tile.png)![](isv-apps/images/Starccm-tile.png)![](isv-apps/images/ECHELON-tile.png) --- Source: https://instinct.docs.amd.com/latest/isv-apps/ansys-fluent.html # Ansys Fluent Ansys Fluent is the industry-leading CFD software tool widely used across aerospace, automotive, energy, high tech, and biomedical industries to simulate complex fluids phenomena and optimize product design. [Get Fluent here.](https://www.ansys.com/products/fluids/ansys-fluent) ## Key Features - Advanced Physics Models and High Accuracy - Supports multiple industries - GA support for MI210, MI250, MI300X, MI325X ## Supported Versions - 2025 R1 & R2 - 2024 R1 & R2 ## Installation and Licensing Information - [AMD Infinity Hub Recipe](https://github.com/amd/InfinityHub-CI/tree/main/ansys-fluent) ## Related News - [Groundbreaking Achievement with CFD Simulation on AMD GPUs](https://www.ansys.com/blog/ansys-baker-hughes-groundbreaking-cfd-simulation) - [AMD Ansys Partnership](https://www.ansys.com/partner-ecosystem/high-performance-computing-partners/amd) - [Boosting Computational Fluid Dynamics Performance with AMD Instinct™ MI300X](https://rocm.blogs.amd.com/ecosystems-and-partners/ansys-fluent-performance/README.html) - [Ansys Fluent® Adds AMD Instinct™ MI200 and MI300 Acceleration to Power CFD Simulations](https://www.hpcwire.com/2024/09/23/ansys-fluent-adds-amd-instinct-mi200-and-mi300-acceleration-to-power-cfd-simulations/) ## Technical Documentation - [Installing AMD GPU Drivers](https://www.amd.com/en/support/download/drivers.html) - [Ansys Fluent GPU Acceleration Guide](https://www.ansys.com/resources/documentation) --- Source: https://instinct.docs.amd.com/latest/isv-apps/ansys-mechanical.html # Ansys Mechanical ## Overview [Ansys Mechanical](https://www.ansys.com/products/structures/ansys-mechanical) is a leading finite element analysis (FEA) platform used for structural engineering. [Get Mechanical here.](https://www.ansys.com/products/structures/ansys-mechanical) ## Key Features - One-stop-shop for multi physics structural analysis - Industry-leading advanced simulation capabilities - Interoperability Advantage through seamless integration with industry-standard tools - GA support on MI210 & MI250. Work in progress support for MI300A, MI300X & MI325X ## Supported Versions - 2025 R1 & R2 - 2024 R1 & R2 - 2023 R2 ## Installation and Licensing Information - [AMD Infinity Hub Recipe](https://github.com/amd/InfinityHub-CI/tree/main/ansys-mechanical) ## Related News - [Ansys and AMD Collaborate to Speed Simulation of Large Structural Mechanical Models Up to 6x Faster](https://www.ansys.com/news-center/press-releases/8-24-22-ansys-and-amd-collaborate-to-speed-simulation-of-large-structural-mechanical-models-up-to-6x-faster) - [Powering Mechanical Simulations: AMD Vs. Intel](https://semiengineering.com/powering-mechanical-simulations-amd-vs-intel) ## Technical Documentation - [Installing AMD GPU Drivers](https://www.amd.com/en/support/download/drivers.html) --- Source: https://instinct.docs.amd.com/latest/isv-apps/cadence-fidelity.html # Cadence Fidelity LES Solver Cadence Fidelity LES Solver, formerly Cascade CharLES, is the industry’s first high-fidelity computational fluid dynamics (CFD) analysis engine that expands the applicability of large eddy simulations (LES) into the mainstream aerospace, automotive, and turbomachinery domains. [Get Fidelity LES Solver here.](https://www.cadence.com/en_US/home/resources/technical-briefs/fidelity-les-solver-tb.html) ## Key Features - Designed to scale, Fidelity LES Solver addresses the most demanding fluid dynamics challenges - Accurate predictions of complex problems for CFD in aeroacoustics, aerodynamics, combustion, heat transfer, as well as multiphase applications - GA support on MI210 & MI250 ## Related News - [Fidelity LES Solver](https://www.cadence.com/en_US/home/resources/technical-briefs/fidelity-les-solver-tb.html) ## Technical Documentation - [Installing AMD GPU Drivers](https://www.amd.com/en/support/download/drivers.html) --- Source: https://instinct.docs.amd.com/latest/isv-apps/devito.html # DevitoPRO DevitoPRO is a domain-specific language (DSL) and code generation framework designed for highly optimized finite-difference kernels for solutions for seismic imaging and exploration, geophysical research, engineering and environmental sciences. [Get Devito here](https://github.com/devitocodes/devito) ## Key Features - HPC Optimization: based on advanced compiler technology - Symbolic Computation: It allows the definition of operators from high-level symbolic equations, facilitating complex mathematical modeling - GA support for MI210, MI250, MI300A, MI300X, MI325X ## Supported Versions - Devito & DevitoPRO 4.8.2 though 4.8.14 ## Related News - [AMD Drives Leadership Performance and Energy Efficiency in Supercomputing](https://www.amd.com/en/newsroom/press-releases/2022-11-15-amd-drives-leadership-performance-and-energy-effic.html) - [Devito revolutionizes high-performance computing for the oil and gas industry with AMD](https://community.amd.com/t5/instinct-accelerators/devito-revolutionizes-high-performance-computing-for-the-oil-and/ba-p/625392) - [DevitoPRO getting HIP with AMD Instinct™ | Devito Codes](https://www.devitocodes.com/instinct) ## Technical Documentation - [Installing AMD GPU Drivers](https://www.amd.com/en/support/download/drivers.html) - [Build Recipe](https://github.com/amd/InfinityHub-CI/tree/main/devitopro) --- Source: https://instinct.docs.amd.com/latest/isv-apps/siemens.html # Siemens Simcenter STAR-CCM+ Simcenter STAR-CCM+ is a best-in-class computational fluid dynamics (CFD) software to enable engineers and analysts to drive accelerated innovation on aerodynamics, turbulence, reacting flows, fluid-structure interaction, and multiphase flows. ## Key Features - Accelerated Product Development to help reduce time-to-market - Optimized design development thru multi-virtual design alternatives for Cost Savings - Multi-physics simulation capabilities including fluid dynamics, heat transfer, electromagnetics and structural characteristics for complex product interaction analysis - GA support for MI210 & MI250 ## Supported Versions - Simcenter STAR-CCM+ 2402 and above ## Related News - [Siemens and AMD Partnership](https://rocm.blogs.amd.com/ecosystems-and-partners/Siemens/README.html) - [Supercharge your CFD simulations with GPUs](https://blogs.sw.siemens.com/simcenter/cfd-simulations-with-gpus/) - [FD on GPU. A seamless disruption with Simcenter STAR-CCM+](https://blogs.sw.siemens.com/simcenter/cfd-on-gpu-a-seamless-disruption/) ## Technical Documentation - [Installing AMD GPU Drivers](https://www.amd.com/en/support/download/drivers.html) - [Build Recipe](https://github.com/amd/InfinityHub-CI/tree/main/siemens-star-ccm) --- Source: https://instinct.docs.amd.com/latest/isv-apps/stone-ridge.html # Stone Ridge Technology ECHELON ECHELON is a leading reservoir simulation software serving the petroleum industry, including applications in deep-ocean drilling and unconventional reservoirs known for its speed and scalability. [Get ECHELON here.](https://stoneridgetechnology.com/echelon-reservoir-simulation-software/) ## Key Features - Speed: Utilizes GPU technology to perform simulations much faster than traditional CPU-based solutions - Scalability: Handles large, complex models efficiently - Accuracy: Provides precise and reliable simulation results - GA support for MI210 & MI250 ## Supported Versions - ECHELON 2023.3+ ## Related News - [Stone Ridge Expands Reservoir Simulation Options with AMD Instinct™ Accelerators](https://www.hpcwire.com/2024/06/17/stone-ridge-expands-reservoir-simulation-options-with-amd-instinct-accelerators/) - [Eni launches new supercomputer HPC6 that ranks No.5. in the TOP500 list](https://www.eni.com/en-IT/media/press-release/2024/11/eni-launches-supercomputer-hpc6-top500-list.html) ## Technical Documentation - [Installing AMD GPU Drivers](https://www.amd.com/en/support/download/drivers.html) - [Build Recipe](https://github.com/amd/InfinityHub-CI/tree/main/srt-echelon) --- Source: https://instinct.docs.amd.com/latest/simulation/gsplat.html # GSplat [GSplat](https://docs.gsplat.studio/main/) is an open-source, GPU-optimized Python library for differentiable rasterization of 3D Gaussians. GSplat allows you to train and render 3DGS models on AMD Instinct™ MI300X devices, which enables you to create models and scenes from captured 2D images and render these in real time. ## Documentation - For full documentation, installation instructions, and API reference guide, refer to the [ROCm docs site](https://rocm.docs.amd.com/projects/gsplat/en/latest/). - View the ROCm enabled GSplat code on [Github](https://github.com/ROCm/gsplat). - Learn more about Gaussian splatting and see examples of its use in the [GSplat blog](https://rocm.blogs.amd.com/software-tools-optimization/gsplat/README.html). --- Source: https://instinct.docs.amd.com/latest/system-admin/bare-metal.html # Bare metal ![](system-admin/images/System-Administrators-Bare-Metal-AMD-GPU-Driver.jpg) --- Source: https://instinct.docs.amd.com/latest/system-admin/co-tools.html # Containers and Orchestration Tools ![](system-admin/images/gpu-operator.jpg)![](system-admin/images/network-operator.png)![](system-admin/images/device-plugin.jpg)![](system-admin/images/device-metrics-exporter.jpg)![](system-admin/images/container-toolkit.jpg)![](system-admin/images/telemetry.jpg) --- Source: https://instinct.docs.amd.com/latest/system-admin/cluster.html # Cluster Documentation Hub ## Design and Guides ![](system-admin/images/system-acceptance.jpg)![](system-admin/images/telemetry.jpg)![](system-admin/images/mi3xx-reference.png)![](system-admin/images/amd-drivenets-system-reference.png) ## Articles and Overviews ![](system-admin/images/RoCE-comparative-analysis.png) --- Source: https://instinct.docs.amd.com/latest/system-admin/cloud.html # Cloud ![](system-admin/images/instinct-azure.jpg) --- Source: https://instinct.docs.amd.com/latest/system-admin/virtualization.html # Virtualization ![](system-admin/images/instinct_virtualization.jpg)![](system-admin/images/instinct_virtualization_SMI.jpg) --- Source: https://instinct.docs.amd.com/latest/resources/index.html # Tools ![](system-admin/images/System-Administrators-Bare-Metal-AMD-AMD-SMI.jpg)![](system-admin/images/enterpriseAI.png)![](system-admin/images/omnistat.jpg)![](system-admin/images/System-Administrators-Bare-Metal-AMD-Tools.jpg)![](system-admin/images/gpu-operator.jpg) --- Source: https://instinct.docs.amd.com/latest/resources/common-reference.html # Common Reference ![](system-admin/images/System-Administrators-Bare-Metal-AMD-GPU-Partitioning.jpg)![](images/instinct-microarchitecture.jpg)![](images/hipcpp.jpg)![](images/virtualization_image.jpg)![](images/secure-docs.png)