PC Processors

[rev_night]

So I have a HTPC that I use for light gaming, with the performance target of 1080@60. I recently had a 2200G in there, which was fine, and I got decent FPS at FarCry 5 (below). I chose this game to be the defacto benchmark because of how CPU sensitive it is. I bought a 2600 (non-X), and installed it, and my FPS sucks! Min, Avg, and Max are down significantly. Thinking its a compatibility issue, I upgrade my chipset drivers, bios, graphics drivers, install Ryzen Master, etc.. Still the fps sucks. And yes, I made sure vsync wasn't on, game settings were the same, radeon settings same, etc..

.

In Open Hardware Monitor, I see the 6 cores with Max freq values of 3900mhz, so I know they are hitting their listed top speed. CPU Utilization is 100%, and GPU utilization is close to 100% too. All clocks and temps appear to be fine. In Cinebench R15, R20, and wPrime 1024, my 2600 gets similar scores to other 2600, although a few % points lower.

.

How is my lowly 2200G, and especially my ancient I3, crushing my 2600?

.

1080 Fps: (Min, Avg, Max)

Old I3-3350P: 41, 58, 82

Ryzen 2200G: 54, 67, 94

Ryzen 2600: 36, 46, 72

.

Specs:

Ryzen 2200G (4C/4T)

Ryzen 2600 (6C/12T)

MSI A320M Gaming Pro (Bios was B, now its I. Most recent is K-Beta)

Radeon 480x (Was 20.5.2, now 20.7.2 via DDU/Factory Install)

8gb 2133 Ram

Win 10

[r29k]

I wonder if this is related to your motherboard and the power delivery to the cpu? A320 might be issue!

[rev_night]

How can we quantify this? If power delivery is a problem, that means my cpu/gpu has less watts available, which means less watts used during peak load. So I should be able to run the game, keep Open Hardware Monitor open, look at Max Powers value, and compare this to what they should be.

[r29k]

I'm saying if you look around online there are complaints about that chipset. Not sure what exactly the problem is though. However you do just list game fps, is the game using the same setting across all hardware? Simple way to test what's going on with the CPU would be to benchmark it and see if you're getting the performance you're supposed to get. If you are then you have some other issue.

[rev_night]

These tests were all done in the same 1-2 hour time period. At first, the only change was me swapping the cpus. Then when I saw the bad performance, i started the various driver/chipset/bios upgrades. This had no impact.

I bought this A320/2200G/2133ram combo cheap off Facebook marketplace, so I didn't do the research I should have. Reading around the forums, the A320 chipset is apparently really bad. I have ordered an Asus TUF B450 mobo and I will try that and post up. I also ordered faster and more ram, but I will install that after the mobo - I want to see what only changing the mobo means.

[r29k]

I suspect you will see significant improvement with the B450, Im running an Asus TUF B450M Plus Gaming.

[rev_night]

oh sweet man, I'm looking forward to it.

[rev_night]

As I said before, I ended up replacing the MSI A320M with an Asus TUF B450M Plus Gaming, as well as the Corsair 2133mhz (8gb) with Corsair 3200mhz (16gb) ram. We all know that faster ram is better, especially for Ryzen. And the results? dun dun dunnn

Yeah, the A320M chipset sucks sh!t. Replacing only the mobo (still same 2133 ram and Ryzen 2600), my fps in Far Cry New Dawn jumped from (min, avg, max) 36, 46, 72 to 61, 75, 99. Thats a 20-30 fps jump in mobo alone. Add in the 3200mhz (DOCP/XMP in bios) and a moderate gpu autoOC, and those numbers marginally increase to 62,78,101. Truth be told, I expected more of a jump with the faster and double ram. This machine's performance target  is 1080p@60 with Amd Chill/Enhanched Sync enabled, since the tv is a 60hz. 

Outside of FC, I benched CineR20, CineR15, and Wprime. These largely were the same between A320 and B450. So just something with games the A320M/2600 was horrific. 

Cool beans. thanks all

[fyrel]

Is your Ram 1 stick or 2?

Is the Ram speed 2133 or does it have an XMP profile to run faster?

[rev_night]

It's 2 sticks of 4GB 2133. And this is not a question of how to improve overall, but why the 2600 isn't as good as the 2200g.

[jamesx380fx]

Its odd cause it should be faster in most games on stock settings, and for reference here you can see an example of both processors beeing tested in games and the 2600 being abit faster overall.

the only difference in that setup he has in the video is that he's using a dual channel Hyperx 3000mhz CL15 ram which is significantly faster than the 2133 that your using (CL14 3200 would be ideal for ryzen but is a bit more expensive than the CL15 ones) the overall ram usage for windows also wasnt shown but having played most of the games on that video it should be around 7-8gb (sometimes higher) for about half of the games shown there.

[rev_night]

Looking at that clip for Far Cry 5, which is the same as Far Cry New Dawn, the 2600 is regularly 20-25+ fps more than the 2200g. That's not a little bit.

As for the ram, I have some Corsair Vengeance LPX (2x8GB) 3200MHz C16 already delivered. I'll install it later, after I do the mobo swap.