A couple of questions regarding VP22 and/or VP23 GPU Performance

anthony-chiappette wrote on 9/17/2025, 8:52 PM

Hi everyone. I'm just wondering about VP22 / VP23 performance with GPU acceleration. My Intel Core i7 11700 has built in Intel HD graphics. I currently have this turned off in BIOS as I have my nVidia GTX 1650 Super (4GB) installed. Is there any benefit to turning on the intel graphics to use as the GPU acceleration in Vegas? I'd still use the dedicated GPU for video output. I know the Intel graphics will eat up some system memory, but I have 64GB RAM to work with.

My other question is regarding nVidia GPU's. As stated, I have a GTX 1650 super with 4GB memory as my dedicated graphics, and I use that for GPU acceleration in Vegas. It seems to work fine, I don't find any noticeable bottlenecks. But based on a comment to me regarding the GTX 1650 not having enough memory for Vegas, I bought an RTX 5060 with 8GB GDDR7. However, upon opening it, I see that it only uses an 8X slot, vs. my GTX 1650 which is 16x. And the RTX 1650 is PCIE 5, where as my motherboard only supports PCIE 4. Will I suffer worse performance by installing this card?

For what it's worth, the GTX 1650 has served me well the last several years. I'm not a gamer, so never really paid too much attention to high power graphics. I chose to upgrade to the RTX 5060 as it was the cheapest option for me.

Any replies are appreciated. Thank you.

ASUS Prime Z590-A Motherboard with Intel Core i7 11700 8 Core / 16 Thread 2.50GHZ, 64GB Crucial DDR4 3200 (4 x 16GB), nVidia GeForce RTX5060 8GB GDDR7, SoundBlaster X AE5 soundcard, 3 x 4TB Samsung 860 EVO SATA 3 SSD, 2 x 8TB Samsung 870 QVO SATA 3 SSD, 1 x 2TB Samsung 980 Pro NVME PICE4 SSD, 2 X WD 4 TB NVME PCIE3 SSD, 2 X Viewsonic HD monitors, LG Blu-Ray writer. Windows 11 (latest build), currently using Vegas Pro 22 latest build, and limited VP23 use to gauge performance and ease of use differences. Videos come from 2 x Sony HDR CX-405 Cameras, XAVC-S MP4 @ 50Mbps 1080P 60fps video files. (Previously: 2 x Canon HFR800 cameras, MP4 files at 1920 x 1080 60p 35Mbps).

Comments

RogerS wrote on 9/17/2025, 9:10 PM

For VP 22/23 it now keeps the image data in your GPU VRAM so there's no real need for the iGPU at this point.
It does have enough VRAM for general work, though more is better (HDR, AI Fx, etc. can really benefit from more).

The RTX 5060 will be a major upgrade from the 1650, especially for decoding. While it may not live up to its full performance potential on a PCI 4 motherboard that doesn't mean it's as slow as a GTX 1650. I expect it will be far faster. If you felt like doing before/after benchmarks there are two in my signature you can use.

anthony-chiappette wrote on 9/17/2025, 10:31 PM

Thank you. I had planned on rendering the project I am working on with the 1650 and recording the results, then installing the new card tomorrow and rerunning the render to see the differences. I will give the benchmarks a shot as well.

ASUS Prime Z590-A Motherboard with Intel Core i7 11700 8 Core / 16 Thread 2.50GHZ, 64GB Crucial DDR4 3200 (4 x 16GB), nVidia GeForce RTX5060 8GB GDDR7, SoundBlaster X AE5 soundcard, 3 x 4TB Samsung 860 EVO SATA 3 SSD, 2 x 8TB Samsung 870 QVO SATA 3 SSD, 1 x 2TB Samsung 980 Pro NVME PICE4 SSD, 2 X WD 4 TB NVME PCIE3 SSD, 2 X Viewsonic HD monitors, LG Blu-Ray writer. Windows 11 (latest build), currently using Vegas Pro 22 latest build, and limited VP23 use to gauge performance and ease of use differences. Videos come from 2 x Sony HDR CX-405 Cameras, XAVC-S MP4 @ 50Mbps 1080P 60fps video files. (Previously: 2 x Canon HFR800 cameras, MP4 files at 1920 x 1080 60p 35Mbps).

RogerS wrote on 9/17/2025, 10:37 PM

It would be great to get your results. I just did a similar one going from a RTX 2080 Super to a 5070; I suspect you'll see an even greater difference.

anthony-chiappette wrote on 9/18/2025, 7:54 PM

Hi again. I prepped my PC for the new video card. I downloaded the latest nVidia studio drivers. I downloaded the latest BIOS for my PC. That was an adventure in itself. I ran a render of my last project in VP23 on the GTX 1650 Super 4 GB GDDR5 with these results:

 

I removed the old drivers with Display Driver Utility then swapped out the GTX1650 with the RTX 5060 8 GB GDDR7. I rendered the project again under the new card and got this result:

The average FPS increased from about 77 to 103, an increase of about 26 fps. The render time decreased by 4 minutes for my project of about 21 minutes, 28 seconds. Not a bad increase in performance.

 

The files I always use are from my 2 Canons 1920 x 1080 60p at 35 mbps. Below is the setting I always use to render.

 

 

I think I made a good choice, no?

ASUS Prime Z590-A Motherboard with Intel Core i7 11700 8 Core / 16 Thread 2.50GHZ, 64GB Crucial DDR4 3200 (4 x 16GB), nVidia GeForce RTX5060 8GB GDDR7, SoundBlaster X AE5 soundcard, 3 x 4TB Samsung 860 EVO SATA 3 SSD, 2 x 8TB Samsung 870 QVO SATA 3 SSD, 1 x 2TB Samsung 980 Pro NVME PICE4 SSD, 2 X WD 4 TB NVME PCIE3 SSD, 2 X Viewsonic HD monitors, LG Blu-Ray writer. Windows 11 (latest build), currently using Vegas Pro 22 latest build, and limited VP23 use to gauge performance and ease of use differences. Videos come from 2 x Sony HDR CX-405 Cameras, XAVC-S MP4 @ 50Mbps 1080P 60fps video files. (Previously: 2 x Canon HFR800 cameras, MP4 files at 1920 x 1080 60p 35Mbps).

anthony-chiappette wrote on 9/18/2025, 8:00 PM

I just had a thought ... would it be worth upping the data rate to 35Mbps to render at the same rate as the native files? Would it make a difference?

ASUS Prime Z590-A Motherboard with Intel Core i7 11700 8 Core / 16 Thread 2.50GHZ, 64GB Crucial DDR4 3200 (4 x 16GB), nVidia GeForce RTX5060 8GB GDDR7, SoundBlaster X AE5 soundcard, 3 x 4TB Samsung 860 EVO SATA 3 SSD, 2 x 8TB Samsung 870 QVO SATA 3 SSD, 1 x 2TB Samsung 980 Pro NVME PICE4 SSD, 2 X WD 4 TB NVME PCIE3 SSD, 2 X Viewsonic HD monitors, LG Blu-Ray writer. Windows 11 (latest build), currently using Vegas Pro 22 latest build, and limited VP23 use to gauge performance and ease of use differences. Videos come from 2 x Sony HDR CX-405 Cameras, XAVC-S MP4 @ 50Mbps 1080P 60fps video files. (Previously: 2 x Canon HFR800 cameras, MP4 files at 1920 x 1080 60p 35Mbps).

anthony-chiappette wrote on 9/18/2025, 8:37 PM

OK, so I made a copy of my render preset and changed it to VBR Min 35M Max 35M. The first time I tried rendering, it crashed about halfway through. I rebooted and rendered again and it completed. The render time and average FPS was pretty much the same, a few seconds faster render time and a few avg fps increase. File size increased by about 20%, however. I didn't notice any quality difference during transitions. Certain transitions slightly pixelate when the transition completes, and it looked the same at both settings. I'll just stick with my CBR 28M setting.

 

ASUS Prime Z590-A Motherboard with Intel Core i7 11700 8 Core / 16 Thread 2.50GHZ, 64GB Crucial DDR4 3200 (4 x 16GB), nVidia GeForce RTX5060 8GB GDDR7, SoundBlaster X AE5 soundcard, 3 x 4TB Samsung 860 EVO SATA 3 SSD, 2 x 8TB Samsung 870 QVO SATA 3 SSD, 1 x 2TB Samsung 980 Pro NVME PICE4 SSD, 2 X WD 4 TB NVME PCIE3 SSD, 2 X Viewsonic HD monitors, LG Blu-Ray writer. Windows 11 (latest build), currently using Vegas Pro 22 latest build, and limited VP23 use to gauge performance and ease of use differences. Videos come from 2 x Sony HDR CX-405 Cameras, XAVC-S MP4 @ 50Mbps 1080P 60fps video files. (Previously: 2 x Canon HFR800 cameras, MP4 files at 1920 x 1080 60p 35Mbps).

RogerS wrote on 9/19/2025, 1:01 AM

Yes, I think it's a worthwhile improvement. I bet the timeline performance is smoother as well.

I'd suggest VBR with a higher maximum bitrate than the average to take advantage of the smarter NVIDIA encoding these days. If it pixelates set a higher maximum bitrate and see if that helps. You don't need to match the original files as they may not have 35M worth of actual detail.
I also usually just use the default NVIDIA quality preset.