--- title: Intel oneAPI description: >- Gather & visualize metrics from reports generated through Intel OneAPI's vtune profiler breadcrumbs: Docs > Integrations > Intel oneAPI --- # Intel oneAPI Supported OS marketplaceIntegration version1.0.0CDS Intel oneAPI ## Overview{% #overview %} **Intel oneAPI** is a unified programming model designed to simplify development across different architectures. **Intel VTune Profiler**, a component of oneAPI, is a profiling tool that assists developers in optimizing their applications. This integration collects and visualizes metrics from VTune reports, enabling you to monitor and analyze your application's performance directly within Datadog. ### Features{% #features %} - Collects metrics from various VTune analysis types, including: - **Performance Snapshot** - **Hotspot Analysis** - **Memory Access Analysis** - **Threading Analysis** - **System Overview** - **HPC Performance Characterization** - Extracts meaningful metrics from VTune reports for visualization in Datadog. ## Data Collected{% #data-collected %} ### Metrics{% #metrics %} | | | | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Elapsed\_Time**(gauge) | Total elapsed time of the application execution*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.CPU\_Time**(gauge) | Total CPU time consumed*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.CPU\_Time.Effective\_Time**(gauge) | Effective CPU time excluding spin and overhead time*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.CPU\_Time.Spin\_Time**(gauge) | Time spent in spin loops*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.CPU\_Time.Overhead\_Time**(gauge) | Time spent in overhead operations*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Paused\_Time**(gauge) | Time when application was paused*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Total\_Thread\_Count**(gauge) | Total number of threads used*Shown as thread* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Effective\_Logical\_Core\_Utilization**(gauge) | Number of effectively utilized logical cores*Shown as core* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Total\_Logical\_Core**(gauge) | Total number of logical cores available*Shown as core* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Effective\_Physical\_Core\_Utilization**(gauge) | Number of effectively utilized physical cores*Shown as core* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Total\_Physical\_Core**(gauge) | Total number of physical cores available*Shown as core* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Top\_Hotspots.CPU\_Time**(gauge) | CPU time consumed by top hotspots*Shown as second* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Collection\_and\_Platform\_Info.Result\_Size**(gauge) | Size of the result data collected*Shown as byte* | | **cds.intel.one.api.Hotspots\_by\_CPU\_Utilization.Collection\_and\_Platform\_Info.CPU.Frequency**(gauge) | CPU frequency*Shown as hertz* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time**(gauge) | Total elapsed time of the application execution*Shown as second* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time.SP\_GFLOPS**(gauge) | Single Precision Floating Point Operations per Second | | **cds.intel.one.api.HPC\_Performance\_Characterization.Vectorization**(gauge) | Percentage of packed floating point operations*Shown as percent* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time.DP\_GFLOPS**(gauge) | Double Precision Floating Point Operations per Second | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time.x87\_GFLOPS**(gauge) | x87 Floating Point Operations per Second | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time.CPI\_Rate**(gauge) | Cycles per Instruction Retired rate | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time.Average\_CPU\_Frequency**(gauge) | Average CPU frequency during execution*Shown as hertz* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Elapsed\_Time.Total\_Thread\_Count**(gauge) | Total number of threads used*Shown as thread* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Effective\_Logical\_Core\_Utilization**(gauge) | Number of effectively utilized logical cores*Shown as core* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Total\_Logical\_Core**(gauge) | Total number of logical cores available*Shown as core* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Effective\_Physical\_Core\_Utilization**(gauge) | Number of effectively utilized physical cores*Shown as core* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Total\_Physical\_Core**(gauge) | Total number of physical cores available*Shown as core* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Vectorization.FP\_Arith\_Mem\_Rd\_Instr.Ratio**(gauge) | Ratio of FP arithmetic to memory read instructions | | **cds.intel.one.api.HPC\_Performance\_Characterization.Vectorization.FP\_Arith\_Mem\_Wr\_Instr.Ratio**(gauge) | Ratio of FP arithmetic to memory write instructions | | **cds.intel.one.api.HPC\_Performance\_Characterization.Collection\_and\_Platform\_Info.Result\_Size**(gauge) | Size of the result data collected*Shown as byte* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Collection\_and\_Platform\_Info.CPU.Frequency**(gauge) | CPU frequency*Shown as hertz* | | **cds.intel.one.api.HPC\_Performance\_Characterization.Collection\_and\_Platform\_Info.CPU.Max\_DRAM\_Single\_Package\_Bandwidth**(gauge) | Maximum DRAM single package bandwidth*Shown as byte* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time**(gauge) | Total elapsed time of the application execution*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.CPU\_Time**(gauge) | Total CPU time consumed*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.Total\_Thread\_Count**(gauge) | Total number of threads used*Shown as thread* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.CPU\_Time.User**(gauge) | CPU time in user mode*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.CPU\_Time.Kernel**(gauge) | CPU time in kernel mode*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.CPU\_Time.Unknown**(gauge) | CPU time in unknown mode*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.GPU\_Time**(gauge) | Total GPU time consumed*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.GPU\_Time.Render\_and\_GPGPU**(gauge) | Time spent in render and GPGPU tasks*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.Paused\_Time**(gauge) | Time when the application was paused*Shown as second* | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.Context\_Switch\_Count**(gauge) | Total number of context switches | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.Context\_Switch\_Count.Preemption**(gauge) | Number of preemption context switches | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.Context\_Switch\_Count.I\_O\_Wait**(gauge) | Number of context switches due to I/O wait | | **cds.intel.one.api.System\_Overview.Elapsed\_Time.Context\_Switch\_Count.Synchronization**(gauge) | Number of context switches due to synchronization | | **cds.intel.one.api.System\_Overview.Top\_Hotspots.CPU\_Time**(gauge) | CPU time consumed by top hotspots*Shown as second* | | **cds.intel.one.api.System\_Overview.GPU\_Time\_Percentage\_of\_Elapsed\_time.GPU\_Time\_percentage\_of\_Elapsed\_time**(gauge) | GPU utilization percentage over elapsed time*Shown as percent* | | **cds.intel.one.api.System\_Overview.GPU\_Time\_Percentage\_of\_Elapsed\_time.GPU\_Time**(gauge) | Total GPU time*Shown as second* | | **cds.intel.one.api.System\_Overview.Top\_Hotspots\_when\_GPU\_was\_idle.CPU\_Time**(gauge) | CPU time during GPU idle periods*Shown as second* | | **cds.intel.one.api.System\_Overview.Average\_DRAM\_Bandwidth.Average\_Bandwidth\_Read\_GB\_per\_sec**(gauge) | Average DRAM read bandwidth per package | | **cds.intel.one.api.System\_Overview.Average\_DRAM\_Bandwidth.Average\_Bandwidth\_Write\_GB\_per\_sec**(gauge) | Average DRAM write bandwidth per package | | **cds.intel.one.api.System\_Overview.Average\_DRAM\_Bandwidth.Total\_GB\_per\_sec**(gauge) | Total DRAM bandwidth per package | | **cds.intel.one.api.System\_Overview.Bandwidth\_Utilization.Platform\_Maximum**(gauge) | Platform maximum bandwidth utilization*Shown as second* | | **cds.intel.one.api.System\_Overview.Bandwidth\_Utilization.Observed\_Maximum**(gauge) | Observed maximum bandwidth utilization*Shown as second* | | **cds.intel.one.api.System\_Overview.Bandwidth\_Utilization.Average**(gauge) | Average bandwidth utilization*Shown as second* | | **cds.intel.one.api.System\_Overview.Bandwidth\_Utilization.Elapsed\_Time\_with\_High\_BW\_Utilization**(gauge) | Percentage of elapsed time with high bandwidth utilization*Shown as percent* | | **cds.intel.one.api.System\_Overview.CPU\_Throttling\_Reasons.Percent\_of\_Samples**(gauge) | Percentage of samples where CPU throttling occurred*Shown as percent* | | **cds.intel.one.api.System\_Overview.Energy\_Consumption**(gauge) | Energy consumption by component | | **cds.intel.one.api.System\_Overview.Collection\_and\_Platform\_Info.Result\_Size**(gauge) | Size of the result data collected*Shown as byte* | | **cds.intel.one.api.System\_Overview.Collection\_and\_Platform\_Info.CPU.Frequency**(gauge) | CPU frequency*Shown as gigahertz* | | **cds.intel.one.api.Threading\_Efficiency.Elapsed\_Time**(gauge) | Total elapsed time of the application execution*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization**(gauge) | Effective CPU utilization of the application*Shown as core* | | **cds.intel.one.api.Threading\_Efficiency.Total\_CPU**(gauge) | Total CPU cores available*Shown as core* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization.Wait\_Time\_with\_poor\_CPU\_Utilization**(gauge) | Wait time with poor CPU utilization*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization.Inactive\_Wait\_Time\_with\_poor\_CPU\_Utilization.Inactive\_Sync\_Wait\_Time**(gauge) | Inactive synchronization wait time with poor CPU utilization*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization.Spin\_and\_Overhead\_Time**(gauge) | Time spent in spin and overhead operations*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization.Inactive\_Wait\_Time\_with\_poor\_CPU\_Utilization.Preemption\_Wait\_Time**(gauge) | Preemption wait time during inactive periods with poor CPU utilization*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Elapsed\_Time.Paused\_Time**(gauge) | Time when the application was paused*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization.Total\_Thread\_Count**(gauge) | Total number of threads used*Shown as thread* | | **cds.intel.one.api.Threading\_Efficiency.Effective\_CPU\_Utilization.Total\_Thread\_Count.Thread\_Oversubscription**(gauge) | Time spent in thread oversubscription*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Top\_functions\_by\_Inactive\_Wait\_Time\_with\_Poor\_CPU\_Utilization.Inactive\_Wait\_Time**(gauge) | Inactive wait time in top functions with poor CPU utilization*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Top\_functions\_by\_Inactive\_Wait\_Time\_with\_Poor\_CPU\_Utilization.Inactive\_Sync\_Wait\_Time**(gauge) | Inactive synchronization wait time in top functions with poor CPU utilization*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Top\_functions\_by\_Inactive\_Wait\_Time\_with\_Poor\_CPU\_Utilization.Inactive\_Sync\_Wait\_Count**(gauge) | Inactive synchronization wait count in top functions with poor CPU utilization | | **cds.intel.one.api.Threading\_Efficiency.Top\_functions\_by\_Inactive\_Wait\_Time\_with\_Poor\_CPU\_Utilization.Preemption\_Wait\_Time**(gauge) | Preemption wait time in top functions with poor CPU utilization*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Top\_functions\_by\_Inactive\_Wait\_Time\_with\_Poor\_CPU\_Utilization.Preemption\_Wait\_Count**(gauge) | Preemption wait count in top functions with poor CPU utilization | | **cds.intel.one.api.Threading\_Efficiency.Top\_Waiting\_Objects.Wait\_Count**(gauge) | Wait count for top waiting objects | | **cds.intel.one.api.Threading\_Efficiency.Top\_Waiting\_Objects.From\_Object\_Wait\_Time**(gauge) | Percentage of wait time from object in top waiting objects*Shown as percent* | | **cds.intel.one.api.Threading\_Efficiency.Top\_Waiting\_Objects.Wait\_Time\_with\_poor\_CPU\_Utilization**(gauge) | Wait time with poor CPU utilization for top waiting objects*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Top\_Functions\_with\_Spin\_or\_Overhead\_Time.Spin\_and\_Overhead\_Time**(gauge) | Spin and overhead time in top functions*Shown as second* | | **cds.intel.one.api.Threading\_Efficiency.Collection\_and\_Platform\_Info.Result\_Size**(gauge) | Size of the result data collected*Shown as byte* | | **cds.intel.one.api.Threading\_Efficiency.Collection\_and\_Platform\_Info.CPU.Frequency**(gauge) | CPU frequency*Shown as hertz* | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time**(gauge) | Total elapsed time of the application execution*Shown as second* | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time.Average\_Latency**(gauge) | Average memory access latency in CPU cycles | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time.CPU\_Time**(gauge) | Total CPU time consumed during execution*Shown as second* | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time.LLC\_Miss\_Count**(gauge) | Number of Last Level Cache (LLC) misses | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time.Stores**(gauge) | Number of store instructions executed | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time.Loads**(gauge) | Number of load instructions executed | | **cds.intel.one.api.Memory\_Usage.Bandwidth\_Utilization.Platform\_Maximum**(gauge) | Platform maximum bandwidth utilization | | **cds.intel.one.api.Memory\_Usage.Bandwidth\_Utilization.Observed\_Maximum**(gauge) | Observed maximum bandwidth utilization | | **cds.intel.one.api.Memory\_Usage.Bandwidth\_Utilization.Average**(gauge) | Average bandwidth utilization | | **cds.intel.one.api.Memory\_Usage.Bandwidth\_Utilization.Elapsed\_Time\_with\_High\_BW\_Utilization**(gauge) | Percentage of elapsed time with high bandwidth utilization*Shown as percent* | | **cds.intel.one.api.Memory\_Usage.Elapsed\_Time.Total\_Thread\_Count**(gauge) | Total number of threads used*Shown as thread* | | **cds.intel.one.api.Memory\_Usage.Paused\_Time**(gauge) | Time when the application was paused*Shown as second* | | **cds.intel.one.api.Memory\_Usage.Collection\_and\_Platform\_Info.Result\_Size**(gauge) | Size of the result data collected*Shown as byte* | | **cds.intel.one.api.Memory\_Usage.Collection\_and\_Platform\_Info.CPU.Frequency**(gauge) | CPU frequency*Shown as hertz* | | **cds.intel.one.api.Memory\_Usage.Collection\_and\_Platform\_Info.CPU.Max\_DRAM\_Single\_Package\_Bandwidth**(gauge) | Maximum DRAM single package bandwidth*Shown as byte* | | **cds.intel.one.api.Performance\_Snapshot.Elapsed\_Time**(gauge) | Total elapsed time of the application execution*Shown as second* | | **cds.intel.one.api.Performance\_Snapshot.Elapsed\_Time.IPC**(gauge) | Instructions per Cycle (IPC) during the elapsed time | | **cds.intel.one.api.Performance\_Snapshot.Elapsed\_Time.SP\_GFLOPS**(gauge) | Single Precision GFLOPS during the elapsed time | | **cds.intel.one.api.Performance\_Snapshot.Elapsed\_Time.DP\_GFLOPS**(gauge) | Double Precision GFLOPS during the elapsed time | | **cds.intel.one.api.Performance\_Snapshot.Vectorization**(gauge) | Percentage of packed floating point operations*Shown as percent* | | **cds.intel.one.api.Performance\_Snapshot.Elapsed\_Time.x87\_GFLOPS**(gauge) | x87 Floating Point Operations per second | | **cds.intel.one.api.Performance\_Snapshot.Elapsed\_Time.Average\_CPU\_Frequency**(gauge) | Average CPU frequency during the elapsed time*Shown as hertz* | | **cds.intel.one.api.Performance\_Snapshot.Logical\_Core\_Utilization**(gauge) | Number of logical cores utilized*Shown as core* | | **cds.intel.one.api.Performance\_Snapshot.Total\_Logical\_Core**(gauge) | Total number of logical cores*Shown as core* | | **cds.intel.one.api.Performance\_Snapshot.Physical\_Core\_Utilization**(gauge) | Number of physical cores utilized*Shown as core* | | **cds.intel.one.api.Performance\_Snapshot.Total\_Physical\_Core**(gauge) | Total number of physical cores*Shown as core* | | **cds.intel.one.api.Performance\_Snapshot.Vectorization.FP\_Arith\_Mem\_Rd\_Instr.Ratio**(gauge) | Ratio of floating point arithmetic instructions to memory read instructions | | **cds.intel.one.api.Performance\_Snapshot.Vectorization.FP\_Arith\_Mem\_Wr\_Instr.Ratio**(gauge) | Ratio of floating point arithmetic instructions to memory write instructions | | **cds.intel.one.api.Performance\_Snapshot.Collection\_and\_Platform\_Info.Result\_Size**(gauge) | Size of the result data collected*Shown as byte* | | **cds.intel.one.api.Performance\_Snapshot.Collection\_and\_Platform\_Info.CPU.Frequency**(gauge) | CPU frequency*Shown as hertz* | | **cds.intel.one.api.Performance\_Snapshot.Collection\_and\_Platform\_Info.CPU.Max\_DRAM\_Single\_Package\_Bandwidth**(gauge) | Maximum DRAM single package bandwidth*Shown as byte* | ### Service Checks{% #service-checks %} **crest\_data\_systems\_intel\_one\_api.status** Returns `CRITICAL` if authentication with datadog has failed. Returns `OK` otherwise. *Statuses: ok, critical* ## Support{% #support %} For support or feature requests, contact Crest Data through the following channels: - Support Email: [datadog.integrations@crestdata.ai](mailto:datadog.integrations@crestdata.ai) - Sales Email: [datadog-sales@crestdata.ai](mailto:datadog-sales@crestdata.ai) - Website: [crestdata.ai](https://www.crestdata.ai) - FAQ: [Crest Data Datadog Marketplace Integrations FAQ](https://docs.crestdata.ai/datadog-integrations-readme/Crest_Data_Datadog_Integrations_FAQ.pdf) This application is made available through the Marketplace and is supported by a Datadog Technology Partner. Click Here to purchase this application.