Overview of Problem Statement

A global production/post production house (the Customer) needed a more economical way of uploading data to the Cloud, including large volumes. For its full resolution video content, it sought much faster speeds and greater cost savings than its current network optimization software solution for uploading content to the Cloud.

Executive Summary of Solution and Top Line Value to the Customer

The Customer, Pixspan Data and Amazon Web Services (AWS) conducted a proof of concept (POC) to compare the speed and cost savings of the Customer uploading large media production files from its data center in Los Angeles to its AWS S3 bucket in Northern California with Pixspan Data’s PixMover software versus the Customer’s current pay as you go network optimization software. Both methods sent production data over its internet and AWS Direct Connect connections to the Cloud.

The results of this POC show that cloud uploads using both methods offer the same speeds, while Pixspan Data’s PixMover is significantly less expensive. The key reason for the difference in costs is the difference in business models – Aspera and Signiant cloud network optimization solutions are SaaS based and charge per gigabyte (up to $0.30 to $0.95 per gigabyte, up to approximately tens of thousands of dollars per month) vs. Pixspan Data which charges a fixed monthly subscription for $1,000 per month for all you can transfer.

In addition, the POC showed even faster (approximately 3 times faster) transfers with Pixspan Data’s built-in Bit Exact Round Trip technology for transferring full resolution content, with an additional 65% savings on Cloud storage, transfer, and egress charges.

Customer Objective/Detailed Problem Statement

The Customer has global media production and post production operations in Los Angeles and in other cities in the Americas, Europe and Asia. Given the increasing size and quality demands of 4K-8K content, the growing demand and flexibility required for video production/post production work around the world, and the tighter timeframes being driven by new streaming services, the Customer wanted to assess a new, less expensive and faster way to deliver media content to the cloud — regardless of how often it transfers, how large its transfers are, or what file format the content is in.

Like most enterprises, the Customer faced many of the common limitations of uploading data to the cloud — namely that it is expensive and slow, requiring sufficient bandwidth to the Cloud. Plus, increased capacity for new projects typically requires incremental capital expenditures, involving months of lead time. Most enterprises today require more variable, and on-demand bandwidth and cloud storage to enable the businesses to flex more effectively.

The Customer faced some unique constraints given that media production files can be very large, involving multiple terabytes per day per movie or TV episode. The largest files are full-resolution video, requiring the very largest file transfers.


The Customer is one of the largest worldwide leaders in the media and entertainment sector, with multiple billions in revenues. The Customer provides creative services including VFX, production, postproduction, and animation in more than a half dozen principal locations worldwide, serving all major motion picture studios as well as newer digital entrants.

Pixspan Data software provides the fastest way to move data to, from, and within the cloud. Pixspan Data also offers the option of Bit Exact Round Trip, the fastest mathematically lossless compression with 50% to 80% savings on uncompressed image and camera raw files. Pixspan Data is an AWS Select Partner.

AWS enables businesses and developers to use web services to build scalable, sophisticated applications. Amazon S3 provides a highly durable storage infrastructure designed for mission-critical and primary data storage.

Proof of Concept Design and Results

The test was designed to measure uploads from the Customer’s data center location in Los Angeles with 10 gigabit connections to AWS S3 storage via a) AWS Direct Connect and b) internet. Two sets of digital image files were used: 1) 5 Terabytes of full resolution image files, and 2) 5 TB of 4K digital intermediates compressed using Pixspan Data’s Bit Exact Round Trip technology (.pxz file format).

In this joint test, the Customer provided its on-premise storage, Internet, Direct Connect, and S3 resources; the Customer operated its Pixspan Data’s PixMover software with its intuitive graphical user interface; and an AWS Solution Architect provided technical support for AWS Direct Connect.

Here is a summary diagram of the POC:


  • Upload times and speed were calculated by measuring the time to accumulate data in Customer’s target S3 bucket. The speed was measured against the known I/O speed of its on-premise storage, which was approximately 550 MB/sec.
  • With test uploads, Pixspan Data’s PixMover moved the 5 Terabytes of full resolution image files to the S3 bucket at approximately 550 MB/sec, both for the internet connection and the AWS Direct Connect. This indicates that the uploads were accomplished at the speed of the “reads” from storage. Thus, PixMover was able to upload the data at approximately the speed of the bandwidth, up to the limit of the storage I/O.
  • The 5 TB files of 4K digital intermediates compressed using Pixspan Data’s Bit Exact Round Trip technology (.pxz file format) were also uploaded over the same AWS Direct Connect and internet bandwidth circuits. Pixspan Data’s PixMover uploaded the compressed data from the same onpremise storage to the Customer’s AWS S3 bucket. The resulting effective upload speed was approximately 1,500 MB/sec. This is roughly 3 times faster than the uncompressed files and about triple the maximum throughput of the Customer’s storage I/O “read” speeds.
  • The costs were measured comparing PixMover versus SaaS cloud upload services offered by network optimization companies Aspera and Signiant. Each uploads content to the cloud at roughly the same speed, but with different business models and vastly different pricing. PixMover is subscription-based, costing about $1,000 per month, whereas the SaaS methods cost from $0.30 per gigabyte to $0.95 per gigabyte, with annual minimums of $10,440 to $12,000 per year (volume discounts may apply). As is shown in the chart below, at volumes of 10 GB per day, the capacity of a 10 megabit circuit, the cost are similar, but PixMover is significantly less expensive at higher volumes. The higher volumes shown are: 10GB/day, the maximum of a 10 megabit circuit; 100GB/day, the maximum of a 100 megabit circuit; 3,000 GB/day, the limit of a 300 megabit circuit; and 500 GB/day, which requires a 500 gigabit circuit.

  • Greater cost savings were measured with PixMover transfers of compressed content with Pixspan Data’s Bit Exact Round Trip technology. In addition to uploads three times as fast, this compressed content saves approximately 65% on cloud storage, cross-region transfer, and egress charges. The data can be decompressed easily to its bit-exact original with PixMover software at the same on distant end.


This POC successfully demonstrated that Pixspan Data’s PixMover is able to upload files to S3 up to the speed of the I/O of originating storage and the approximate practical speed of internet and AWS Direct Connect circuits, at dramatically lower costs:

  • The Customer’s media workflows can be enhanced by uploading files of any type to the cloud. This workflow enhancement allows access to scalable cloud storage, on demand and without capital expenditures. Once in the cloud, Pixspan Data’s CloudMover application for moving content at ultra-high speeds is also available, on-demand.
  • Pixspan Data’s PixMover moves data to the cloud with cost savings significantly greater than competing network optimization methods. The greater the volume of data, the greater the savings.
  • PixMover is able to upload files to S3 with its Bit Exact Round Trip option, which increases the effective data transfer rate for full resolution media content, in this POC by approximately 3 times. This increases the savings versus alternative methods, while saving additionally on cloud storage, transfer, and egress charges.