Overview
Introduction
The Ascend-DMI tool provides functions, such as bandwidth test, computing power test, and power consumption test, for standard PCIe cards, board cards, and modules of Atlas products. Table 1-1 lists the functions of the Ascend-DMI tool. The system invokes the underlying device control and management interface (DCMI), device system management interface (DSMI), and interfaces related to the Ascend Computing Language (ACL) to implement the tests. The system-level information is queried by invoking the common library provided by the system. You can configure parameters to implement different test functions when using the tool.
Function |
Description |
|---|---|
Bandwidth test |
Measures the bus bandwidth, memory bandwidth, and latency. |
Computing power test |
Measures the computing power of the AI core in a chip and the average power of the chip in full computing power. |
Power consumption test |
Measures the power consumption of the chip or entire card. |
Real-time device status query |
Checks the status of the device in running. |
Fault diagnosis |
Obtains the chip health information, tests the computing power, power consumption, and bandwidth of the chip, and outputs the test results. |
Software and hardware version compatibility test |
Checks the hardware and software compatibility based on the hardware information, architecture, driver version, firmware version, and software version obtained. |
Device topology detection |
Displays the topology between multiple cards in a device. |
Constraints
When the Profiling tool is enabled, the Ascend-DMI tool is unavailable.
For details about the Profiling performance analysis tool, see the following documents:
Inference scenario: See "Environment Preparation" in the CANN V100R020C10 Auxiliary Development Tool Guide (Inference).
Training scenario: See "Overview" in the CANN V100R020C10 Auxiliary Development Tool Guide (Training).
Heterogeneous Computing Architecture
In the heterogeneous computing architecture, the Ascend AI processor interworks with the CPU through the PCIe bus as follows:
- Host: a system running with the CPU. It is an x86 or ARM server connected to the hardware device (such as an Atlas 300I inference card) running with an Ascend AI processor. It uses the acceleration computing capability provided by the Ascend AI processor to complete service processing.
- Device: a hardware device equipped with an Ascend AI Processor. It is connected to the server through the PCIe interface and provides acceleration computing power for the server.
Compatibility Information
Table 1-2 lists the hardware supported by the Ascend-DMI tool.
Hardware |
|---|
Atlas 800 inference server (model 3000) Atlas 800 inference server (model 3010) Atlas 800 training server (model 9000) Atlas 800 training server (model 9010) Atlas 300T training card (model 9000) Atlas 300I inference card (model 3000) Atlas 300I inference card (model 3010) Atlas 500 Pro AI edge server (model 3000) |
