Accommodate PC with low resources

Topaz Photo AI General
1

With Sharpen AI my comparatively old Windows 10/11 PC could complete any operation. Admittedly it was slower than a more modern PC with higher CPU and GPU would be able to perform. However, it could still work. When more processes were running on my PC then Sharpen AI took longer, but it didn’t crash.

Photo AI seems more insistent on some resources. Perhaps it’s bumping into VRAM exhausted and then aborts. My current graphics card with Nvidia 1650 GPU only has 4 GB dedicated on board memory, which I realize today is low. However, Sharpen AI didn’t crash out even when I was also running other image processing software which consumed / competed for same types of resources.

I’d really like for Photo AI (and successor products) to be able to leverage limited resources and behave more gracefully. Maybe today’s product is especially constrained by using code toolkits which don’t themselves provide any techniques to cope with exhaustion. Maybe the ‘only’ way a program can use a graphics card is via routines which the card manufacturer publishes and application software developers have to conform to that.

I’m old enough that I remember in the 1980s there were .OVL files that ‘extended’ what .EXE programs could use to overcome memory limitations. I remember when virtual memory became a feature to get beyond expensive main system RAM. Of course performance was vulnerable to slow disks. I remember in the scientific community that tough, big problems could be divided automatically into smaller units and they could be spread among additional processors/systems. Of course performance was vulnerable to available communications capacity and amounts of data to be shoved, and also the latency between processing systems.

I’d like for Photo AI (and presumably video products, too) to use the fastest/best components available in the system, and have a means to also benefit from the rest of the system when needed. There should be a configurable limit upon the allowed consumption of resources, such as, perhaps 85%. Leave some room for general system housekeeping. Or checking email or browsing while the image processing has to chug along…

2

I hear what you’re saying Richard. The issue is that software applications, like PhotoAI, are developed to take advantage of new video card (GPU) hardware and software to get the results and features. So robust video GPU cards (and the latest drivers) are even more important. All that and a generous amount of system RAM memory and a relatively fast CPU.

It is irritating that technology many times is the driving force behind the development of software applications and that happens so fast…making what we have become obsolete practically overnight.