Performance Specifications

For the Atlas training products, this function is reserved and is not supported currently.

Performance Specifications

This section describes only the performance specifications of the Ascend AI Processor of each product, which may differ based on the memory and CPU configurations of the respective product. For details about the product performance specifications, see the official descriptions at the Ascend community.

This section takes a single Ascend AI Processor as an example to describe the performance specifications in basic scenarios, as shown in the table below. The performance at other resolutions can be estimated in the same way.

In the following table, one channel corresponds to one thread, 720p indicates a resolution of 1280 × 720 pixels, 1080p indicates a resolution of 1920 × 1080 pixels, and 4K indicates a resolution of 3840 × 2160 pixels.

Table 1 Ascend AI Processor performance specifications of the Atlas inference products

Scenario

Total Frame Rate

  • Input resolution: 1080p (1920 x 1080)
  • Output resolution: 1080p (1920 x 1080)
  • Input/Output format: YUV420SP
  • n channels (1 ≤ n ≤ 12)

n x 360 FPS

  • Input resolution: 1080p (1920 x 1080)
  • Output resolution: 1080p (1920 x 1080)
  • Input/Output format: YUV420SP
  • n channels (n > 12)

4320 FPS

  • Input resolution: 4K (3840 x 2160)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (1 ≤ n ≤ 12)

n x 90 FPS

  • Input resolution: 4K (3840 x 2160)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n > 12)

1080 FPS

  • Input resolution: 8K (7680 x 4320)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (1 ≤ n ≤ 12)

n x 22.5 FPS

  • Input resolution: 8K (7680 x 4320)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n > 12)

270 FPS

Table 2 Performance specifications of the Ascend AI Processor on the Atlas 200I/500 A2 inference products

Scenario

Total Frame Rate

  • Input resolution: 1080p (1920 x 1080)
  • Output resolution: 1080p (1920 x 1080)
  • Input/Output format: YUV420SP
  • n channels (n < 2)

n x 800 FPS

  • Input resolution: 1080p (1920 x 1080)
  • Output resolution: 1080p (1920 x 1080)
  • Input/Output format: YUV420SP
  • n channels (n ≥ 2)

1600 FPS

  • Input resolution: 4K (3840 x 2160)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n < 2)

n × 200 FPS

  • Input resolution: 4K (3840 x 2160)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n ≥ 2)

400 FPS

  • Input resolution: 8K (7680 x 4320)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n < 2)

n x 100 FPS

  • Input resolution: 8K (7680 x 4320)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n ≥ 2)

200 FPS

Table 3 Ascend AI Processor performance specifications of the Atlas A2 training products/Atlas A2 inference products

Scenario

Total Frame Rate

  • Input resolution: 1080p (1920 x 1080)
  • Output resolution: 1080p (1920 x 1080)
  • Input/Output format: YUV420SP
  • n channels (1 ≤ n ≤ 10)

n x 1440 FPS

  • Input resolution: 1080p (1920 x 1080)
  • Output resolution: 1080p (1920 x 1080)
  • Input/Output format: YUV420SP
  • n channels (n > 10)

14400 FPS

  • Input resolution: 4K (3840 x 2160)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (1 ≤ n ≤ 10)

n x 360 FPS

  • Input resolution: 4K (3840 x 2160)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n > 10)

3600 FPS

  • Input resolution: 8K (7680 x 4320)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (1 ≤ n ≤ 10)

n x 90 FPS

  • Input resolution: 8K (7680 x 4320)
  • Output resolution: 4K (3840 x 2160)
  • Input/Output format: YUV420SP
  • n channels (n > 10)

900 FPS

The VPC processing performance is closely related to the image resolution during processing. The maximum resolution of the input and output images is used as the reference resolution. A larger reference resolution indicates a longer processing time and lower performance.

A VPC batch processing API (whose name includes batch, such as hi_mpi_vpc_batch_crop_resize_paste) allows parallel image processing on multiple VPC hardware units in a digital vision preprocessor (DVPP). This maximizes the total frame rate on a single channel.