Exploring the Impact of Professional Workstations for Processing Laser Scan Data

Powerful workstations: The missing piece for a successful laser scanning business

In the short time since laser scanners arrived in the architecture, engineering, and construction (AEC) industries, they have become absolutely essential tools.

These tools capture spatial data for objects and environments at a level of detail and fidelity that was impossible before. By producing rich 3D models of the project site, they have already changed how we plan, design, and execute our most demanding projects.

The problem? If you want to build a successful laser scanning business in the AEC space, you’ll need a powerful computer to work with this 3D data efficiently.

Why is laser scan data challenging to process?

A laser scanner (sometimes called a lidar scanner) uses beamed lasers to measure the distance to objects and surfaces in its environment. 

It records these measurements as XYZ points (with intensity values) in a data set called a point cloud. If the laser scanner includes a camera, you can also record RGB values for colorization purposes.
3D laser scanning piping equipment for chemical plant applications
A laser scanner and its beautiful, data-rich 3D result. Image credit Hive Virtual Plant.

Underpowered workstations

Processing your point clouds with an underpowered workstation can create significant headaches for you and your team.

    • Longer processing times
      A less powerful computer will have more limited CPU resources, so it will take longer to process point cloud data. And don’t expect to work around this limitation by leaving your data sets to process overnight – most point-cloud processing requires manual work from a trained specialist.

    • Scaling issues
      Slower processing becomes an even bigger problem as your point clouds grow in size and number.

      The relationship between the number of scans and the time it takes to process each scan is not linear. Processing times increase unpredictably as you add more point clouds to the process. That means scaling your workflow to handle a bigger project can mean an exponential slowdown.

    • Longer export times
      You might think that an underpowered workstation could at least export your data sets quickly once processing is complete. But you’d be wrong. Exporting a project efficiently in most software requires immense amounts of temporary storage space and computing power.

    • Inability to multitask
      Processing software will use as much processing bandwidth as possible, often to the point of maxing out your CPU. If you try running multiple tasks with an underpowered CPU, you risk crashing the software, losing progress, and having to start all over again. 

      When processing consumes 100% of a computer’s resources, your computer is useless until the job is done.

    • Slow project delivery
      Slow processing and the inability to multitask on your workstation creates a project bottleneck. Every hour you wait for the data to process is an hour the client will have to wait for delivery of their data sets.

    • Limits on business growth and profit
      In the end, this bottleneck limits your productivity, your ability to finish big projects, and even your ability to take on more projects. And that can stop the growth of your business.
Point cloud on a Dell Workstation
Dell workstations, the surest thing for massive point clouds. Image credit Giraffe360.

Testing a more powerful workstation

Like many laser scanning service providers, Nexus 3D Consulting (now Allen 3D) encountered these problems often in their day-to-day work. That’s why they tested a more powerful Dell workstation for their processing workflow, with the hope that they could shorten processing and export times, speed up project delivery, and even help boost the company’s growth and profits. 

To perform this test, they set up a controlled comparison with one of their existing workstations. They captured a laser scan data set in a generic office, typical of a smaller-scale laser-scanning project. Then, they processed this dataset with both computers, while measuring benchmarks like import time, CPU usage, memory usage, and maximum concurrent processes.

Below are the specifications for the computers they used and information about the test data.

Computers

  • Previous Computer
    • Unnamed Gaming Computer
    • Processor: Intel® Core™ i9-9900K CPU @ 3.60GHz
    • RAM: 64 GB
    • GPU: NVIDIA GeForce RTX 2080 Ti
  • New Workstation
    • Dell Precision 3660
    • Processor: 12th Gen Intel® Core™ i9-12900K 3.2GHz
    • RAM: 128 GB
    • GPU: NVIDIA RTX A5500

Data

For testing, the team scanned a small office space with a terrestrial laser scanner, producing:
    • Scans: 135 in a mix of B&W and color 
    • Raw data size: 46.38 GB 
Test Results The Dell Precision workstation, with its powerful Nvidia GPU and Intel Processor, delivered on its promise of greater processing speed. 

But the team was not prepared for how much faster it would be. The Dell workstation was 147.69% faster in completing the initial import processing for the data set. It took only 1 hour and 5 minutes, while the old workstations took 2 hours and 41 minutes.

See below for a more detailed comparison of processing benchmarks.
Import Time (Total)
2 hours, 41 minutes
1 hour, 5 minutes
Import Time (Per Scan)
1 minute, 40 seconds
45 seconds
CPU usage
14-30% (with a few
100% spikes)
50-80% (with 100% spikes every 2-5 minutes)
Threads
5000
7500-8000
Handles
173,717
300,000
Processes
300
340-350
Memory Utilization
40-60%
20-40%
Import Time (Total)
Standard Gaming PC: 2 hours, 41 minutes
Dell Precision Workstation: 1 hour, 5 minutes

Import Time (Per Scan)
Standard Gaming PC: 1 minute, 40 seconds
Dell Precision Workstation: 45 seconds

CPU Usage
Standard Gaming PC: 14-30% (with a few 100% spikes)
Dell Precision Workstation: 50-80% (with 100% spikes every 2-5 minutes)

Threads
Standard Gaming PC: 5000
Dell Precision Workstation:7500-8000

Handles
Standard Gaming PC: 173,717
Dell Precision Workstation: 300,000

Processes
Standard Gaming PC: 300
Dell Precision Workstation: 340-350

Memory Utilization
Standard Gaming PC: 40-60%
Dell Precision Workstation: 20-40%

Long-term business benefits

The team says the faster processing will give them a big advantage in future projects.
    • Faster delivery
      The Dell workstation cut processing time by more than half on smaller data sets. It also has spare power to scale up for larger data sets. That enables the  team to finish projects much faster – and make final delivery to their clients even faster than before.

    • Freed-up resources
      Because the team can process scan data faster, that frees up time and other resources they can use to focus on other project-critical tasks.They can attend to their clients’ needs with greater care, ensuring higher satisfaction.

    • More jobs
      In some cases, those freed-up resources may be used to work on new jobs entirely, helping the company to continue growing its business, handling more projects with the same amount of resources.
Increased productivity and growth
As processing time goes down, productivity & growth go up. Image credit Business Insider.

Powerful computers matter

Don’t underestimate the benefits of using a powerful workstation for processing laser scan data.

This test clearly demonstrates the dramatic impact that a powerful workstation can have on your laser-scan processing times. It also shows how that faster processing can create bigger changes across your business, bringing benefits that can help grow your company while improving your reputation.

Whether working on a fixed workstation or on-the-go with a mobile workstation, Dell Precision systems with the latest Intel and NVIDIA technology offer the performance and reliability professionals need to succeed.Learn more about professional workstations here.

R-CON 2024 Announced

October 8 - 10, 2024
Boise, Idaho

Reserve your ticket, submit a speaker or panelist application, or explore exhibitor and sponsorship options!