Sizing guide

The following sizing guide shows sizing considerations when using Decipher IDP at scale.

Size Decipher Web Clients Rabbit MQ Decipher Server Data
Capture
Client
OCR
Client
Image
Processing
Client
Classification Client Export Client SQL Server
Small (Sizing numbers are per virtual machine (VM). HA and DR sizing are pro rata)

Concurrency of 5–10 users Maximum of 20 batches of 20 pages per hour

Up to 100 Pages per hour Up to 100 Pages per hour Up to 1 Classification training / month Up to 100 pages per hour Up to 100 scanned pages per hour Or up to 150 native PDF pages per hour Up to 100 scanned pages per hour up to 150 native PDF pages per hour Classification of up to 3 document types per batch Export only to Blue Prism Storage for up to 1000 Batches or 20,000 document pages / month

4 CPU (i5)

8 GB RAM

200 GB HDD

2 CPU (i5)

4 GB RAM

200 GB HDD

4 CPU (i7)

4 GB RAM

200 GB HDD

4 CPU (i5)

8 GB RAM

200 GB HDD

Medium (Sizing numbers are per VM. HA and DR sizing are pro rata) Concurrency of 10-50 users Maximum of 100 batches of 20 pages per hour Up to 1000 Pages per hour Up to 1000 Pages per hour Up to 5 Classification training / month Up to 1000 pages per hour Up to 1000 scanned pages per hour Or up to 1500 native PDF pages per hour Up to 1000 scanned pages per hour up to 1500 native PDF pages per hour Classification of up to 3 document types per batch Export to Blue Prism and external DBs / file systems Storage for up to 10k Batches or 1 million document pages / month

8 CPU (i5)

16 GB RAM

200 GB HDD

4 CPU (i5) 8 GB RAM 400 GB HDD

8 CPU (i7)

16 GB RAM

200 GB HDD

8 CPU (i5) 16 GB RAM 400 GB HDD
Large (Sizing numbers are per VM. HA and DR sizing are pro rata) Concurrency of >50 users >100 batches of 20 pages per hour >1000 Pages per hour > 100 Pages per hour > 5 Classification training / month > 1000 pages per hour > 1000 scanned pages per hour Or > 1500 native PDF pages per hour > 1000 scanned pages per hour > 1500 native PDF pages per hour Classification of > 3 document types per batch Export to Blue Prism and external DBs / file systems > 10k Batches or 1 million document pages / month

16 CPU (i5)

32 GB RAM

200 GB HDD

8 CPU (i5) 16 GB RAM 500 GB HDD

16 CPU (i7)

32 GB RAM

200 GB HDD

8 CPU (i5)

16 GB RAM

500 GB HDD

  • Medium and large sizes can be scaled out by adding more VMs with the same VM size as small. The overall count of resources should match the specifications here.
  • All sizing information is for reference only and can vary for each implementation. The language selected for OCR, the format of the image file being processed, the complexity of the document being annualized and the complexity of the Data File Definition can all impact the performance of Decipher IDP.
  • An average of 50 fields including tables are assumed for this sizing.