Start cooking your performance analysis. Understand tuning techniques, performance metrics and hardware solutions to collect statistics. Next, drill down to particular tuning or configuration recipes that feature Intel® VTune™ Profiler or its predecessor, Intel® VTune™ Amplifier.
Recipe |
Description |
|---|---|
| Top-down Microarchitecture Analysis Method | Understand how an application utilizes available hardware resources. Enable it to take advantage of CPU microarchitectures. |
| OpenMP* Code Analysis Method | Analyze CPU utilization of your OpenMP* or hybrid OpenMP-MPI application. Identify causes for possible inefficiencies. |
| Software Optimization for Intel® GPUs (NEW) | Estimate overhead when offloading onto an Intel GPU. Analyze the performance of computing tasks offloaded onto the GPU. |
| Core Utilization in DPDK Apps | Examine metrics that characterize core utilization in terms of packet receiving in DPDK-based applications. |
| PCIe Traffic in DPDK Apps | Examine PCIe Bandwidth metrics used in Intel® VTune™ Profiler and investigate PCIe traffic for a packet forwarding DPDK-based workload. |
| DPDK Event Device Profiling | Analyze the efficiency of DPDK Event Device pipeline utilization in your DPDK-based application. Identify issues like inhomogeneous load distribution and worker core underutilization. |
| Effective Utilization of Intel® Data Direct I/O Technology | Use Intel® VTune™ Profiler to examine the utilization efficiency of the Intel® Data Direct I/O technology, a hardware feature of Intel® Xeon® processors. |
| Compile a Portable Optimized Binary with the Latest Instruction Set | Explore the methods of compiling an application binary for distribution, with their pros and cons. Learn to compile a portable binary that uses the latest hardware features while retaining backwards compatibility. |