gabrimor

Thank you YoungBlade. I found this early, but it wasn't enough to me.

Thank you, Shimeji. My bios is updated and the resizable bar option is enabled, but I don't know if this really works. I need a compatible GPU to test. Before buying the arc, I'll try to find more information.

Hello, thanks for reading this topic. I want to buy an Arc 770 graphics card, but it doesn't perform well without resizable bar. I have an Asrock Taichi X370 with Ryzen 2700X. People said resizable bar is not available with this setup. I've searched in my BIOS and resizable bar is available. Do you think resizable bar will work with the Arc 770 with my setup? This is something really new to me, and my research was not enough to understand. Thank you!

Compatible, mate!

The problem stills, but I was wrong about the PCIE speed. I had a bugged GPU-Z version. The new GPU-Z shows x16 3.0 under load. Installed a R9 280X and, yes, about 10% less fps.

I think it's the socket pins. I can't see any bent sockets, but it's one of the last ideas I have.

PCIE @ 16x 1.1, instead of 3.0. Stays 1.1 under load. I think it's the problem!

I think it's not XMP, cause the CPU performance (Physics Test) was very, very good. And XMP alone couldn't cause a 10fps drop. I tried to load optimized default. I ended up changing the DZ77RE-75K, which is one of my favorites mobo, for the Z77 Extreme 4, which now is in use. I'll try to find answers!

Tried another ssd, which has Windows 10 installed. No dice. Tried Titan with a third motherboard, an Asrock Z77 Extreme 4. Titan scored as expected. The low score happens only with the Intel DZ77RE-75K. I can't undestand.

Nothing different in the task manager, mate.

Thanks for still helping me! So, Windows same build, cause I'm using the same SSD. Spectre and Meltdown, I suppose it's in the same condition, cause I'm using the same SSD. XMP and dual channel enabled in both systems. I don't believe it's CPU/RAM related, cause the system with lower graphics performance, have higher CPU performance in 3DMark! About the scores, a simple example: In 3DMark 11 Graphics Test 3, I'm getting about 97fps with the Asrock, while getting 88fps with the Intel. It's a consistent difference between all the graphics test.

Ok, but can you guess what's happening? I put all the components installed in the the Asrock Z77 Pro 4 in the Intel DZ77RE-75K. The 10% difference remains!!!

Looks like something related with the PCIE bandwidth! I swapped all the parts and nothing changed.

Thank you, mate. I'll check it!

Sorry for the poorly organized post, and thanks for your reply. I bought a GTX Titan, last week, and I installed it in a test rig. This test rig are working with a 3570K CPU, an Asrock Z77 Pro 4 Motherboard, 2x2GB G. Skill 1600mhz RAM and a EVGA 500W PSU. I just used 3DMark for testing the graphics card, which is working fine. After testing, I just installed the GTX Titan in another rig, for using it. The other rig are working with a 3770 CPU, an Intel DZ77RE-5K motherboard, 2x4GB Kingston HyperX RAM and a EVGA 500W PSU. In the 3570K test rig, for example, I was reaching about X fps in 3DMark 11 Graphics Test 3. In the 3770 rig, I'm reaching about Y fps, which is 10% less. Yes, this difference is consistent between other tests. I tried to transfer RAM, PSU and the SSD from the test rig for the 3770 rig. The difference still about 10%. So, you'll think it's the CPU or motherboard! I don't think it's the CPU, cause the 3770 is reaching a very high score, compared with the 3570K. So, the problem is the motherboard, I think. But the motherboard is working very well. Tried to change settings in BIOS, but nothing changes.