What Are Processing Units (PUs) and How Are They Consumed?

What Are Processing Units?

Processing Units (PUs) are the metric used to measure compute resource consumption on Planet Insights Platform. When making an API requests for tasks such as fetching satellite imagery, performing statistical analyses, or streaming data, PUs are used based on the complexity of the request. The number of PUs consumed depends on factors like the area of interest, spatial resolution, number of spectral bands, and the type of data requested.

Types of Processing Units

  1. Monthly Processing Units: Allocated based on your subscription plan and reset at the beginning of each month. Unused units do not roll over.

  2. Top-up Processing Units: One-time purchases that provide additional PUs for handling large, non-recurring workloads or unexpected usage spikes.

How Can I Monitor My Usage?

  • Dashboard: Track your PU consumption via the Planet Insights Platform dashboard.

    Note: You can track your Usage within Pip for Workspace and run reports  via Workspaces -> Open work space -> Usage. 

  • API Response Headers: Each API response includes an x-processunits header indicating the number of PUs consumed by that request.

How Do I Calculate My PUs?

1. Data Processing (Processing API, OGC API, Statistical API)

A standard request consuming 1 PU typically includes:

  • An output image size of 512 x 512 pixels

  • 3 collection input bands

  • One data sample per pixel
  • An output image format not exceeding 16 bits per pixel

  • Without additional processing such as orthorectification applied

Minimum PU costs per request are:

  • Processing API and OGC API: 0.005 PU

  • Statistical API: 0.01 PU

PUs are deducted only when a request successfully executes. 

2.Data Access & Delivery (Orders, Subscriptions, Data APIs)

  •     Actions                                                                                          PU Cost

          Image Asset Activation                                                                 20 PU per asset
         Clip Tool                                                                                       1 PU per data asset
         Merge Tool                                                                                  10 PU per request
        Coregister Tool                                                                              18 PU per request
        Other Tools (Band Math, Harmonize, Reproject, TOAR, etc.)         2 PU per tool, per data asset
        Egress (cloud delivery)                                                                   200 PU per GB
     

  • Example: 10 fields, 2 assets/day, 292 days, with Clip and Harmonize tools:

          Assets: 2 × 292 × 20 = 11,680 PU
          Clip: 10 × 292 × 1 = 2,920 PU
          Harmonize: 10 × 292 × 2 = 5,840 PU
           Total: 20,531 PU/year

Multiplication Factors Affecting PU Consumption

The actual PU consumption for a request is calculated by applying multiplication factors based on various parameters:

  • Area of Interest: Larger areas increase PU usage proportionally.

  • Number of Input Bands: Using more than 3 bands increases PU consumption.

  • Output Format: Requesting 32-bit float TIFFs or certain streaming formats (e.g., application/octet-stream) increases PU usage due to higher memory and data transfer requirements.

  • Number of Data Samples: Requesting multiple samples per pixel (e.g., for time series analysis) increases PU consumption.

  • Data Fusion: Combining data from multiple collections, especially from different locations, increases PU usage.

For example, requesting a 1024 x 1024 pixel image with 6 bands in 32-bit float TIFF format would consume more PUs than the standard 1 PU, due to the increased area, number of bands, and output format.

For more in-depth explanations of PU consumption, including optional multiplicators, API requests, and example calculations, please visit the Planet Docs site.

Was this article helpful?
0 out of 0 found this helpful

Comments

0 comments

Article is closed for comments.