A few years go, Puget Systems built a workstation optimized for Adobe Lightroom for a customer named Bob. Bob had recently changed a few areas of his workflow and was starting to experience intermittent performance dips.
What part of his workflow changed? Bob began using Topaz Sharpen AI more heavily to repair blurry images, and he started to experience long export times. This delay extended the time to complete projects and took him out of the flow of his work.
When Bob worked with Puget Systems to configure a workstation, he purchased a system optimized around his Lightroom usage, and didn't take the requirements for Topaz Sharpen AI into consideration. On the call, he made it clear that he now relied heavily on this application and wanted to speak with a consultant to purchase a new workstation that could handle the changes to his workflow.
After listening to the customer's concerns, my first instinct was to configure a new workstation that was a slight upgrade from his current machine. This allowed me to compare his current system to a new one.
But something didn't feel right.
As I dug more into the details of each system, and reviewed the hardware requirements for Adobe Lightroom, I noticed the customer was currently running an NVIDIA GTX 2080 graphics card. On the surface, this card is still a good choice from Lightroom, so I did a little more research on the Topaz Sharpen AI add-on.
My research led me to speak with William George from Puget Labs, who has a deep understanding of what applications take advantage of the NVIDIA CUDA cores. After some discussion, William and I came to the conclusion that Topaz Sharpen AI relies on CUDA cores to perform its magic. In short, the more CUDA cores the better!
Our conclusion was confirmed when the customer noted that his CPU and RAM usage wasn't slammed when running Topaz AI, but his GPU usage was nearly at or near 89%.
This shows why I deal with Puget Systems. You took the time to review my system and provide honest feedback.
The customer came to us in search of a new computer. And we could have built him one he'd be happy with, but that isn't what we are about at Puget Systems. The customer didn't need a new computer in order to improve his workflow performance. We pride ourselves on helping the customer allocate his or her budget on those components that impact performance.
In this instance, the customer was better off allocating budget towards a current generation NVIDIA 30-series RTX card. Sometimes a new computer is the best solution, but not in this instance.