Jump to content

4000mhz ram oc with dark hero board. Anyone?

Build is 5950x with 4x32 (PC4 32000) Desktop Memory Model F4-4000C15Q2 trident Z royal ram and the dark hero board. Before I update my bios and dedicate a bunch of time to see what I can get to run I’m just wondering if anyone on here has a similar setup and got the ram to at least be stable at 4000mhz on the same board? 

Link to comment
Share on other sites

Link to post
Share on other sites

2 hours ago, StuSavage said:

Build is 5950x with 4x32 (PC4 32000) Desktop Memory Model F4-4000C15Q2 trident Z royal ram and the dark hero board. Before I update my bios and dedicate a bunch of time to see what I can get to run I’m just wondering if anyone on here has a similar setup and got the ram to at least be stable at 4000mhz on the same board? 

Running 4000mhz with 4 sticks of ram is hard or impossible due to instability caused by having 4 ram sticks.I would aim for something lower like 3733 MHZ.I saw(on this forum) somebody try and do the exact same thing and he failed.

amd-ryzen-memory-decoupling-580x326.jpg

Link to comment
Share on other sites

Link to post
Share on other sites

Also running at 4000mhz is WORSE FOR PERFORMANCE for ryzen as at 3600mhz it runs 1:1 with the memory controller thus having the optimal transfer rate throughput.

Link to comment
Share on other sites

Link to post
Share on other sites

Good luck man, I tried with 4x8 and it was pretty tough. Best I could do was 1900 1:1 stable, and occasionally 1933, but not stable. Best I have done is 2100 with 2 sticks 1:1 not stable, and stable at 2K. Try with 2 sticks and see what happens.

AMD R7 5800X3D | Thermalright Phantom Spirit 120 EVO, 1x T30

Asus Crosshair VIII Dark Hero | 32GB G.Skill Trident Z @ 3733C14

Zotac 4070 Ti Trinity OC @ 3060/1495 | WD SN850, SN850X, SN770

Seasonic Vertex GX-1000 | Fractal Torrent Compact RGB, Many CFM's

Link to comment
Share on other sites

Link to post
Share on other sites

9 hours ago, Cheburek said:

Running 4000mhz with 4 sticks of ram is hard or impossible due to instability caused by having 4 ram sticks.I would aim for something lower like 3733 MHZ.I saw(on this forum) somebody try and do the exact same thing and he failed.

amd-ryzen-memory-decoupling-580x326.jpg

1900 fclk and 3800 seems to be the best from what I’m seeing 

Link to comment
Share on other sites

Link to post
Share on other sites

4 hours ago, freeagent said:

Good luck man, I tried with 4x8 and it was pretty tough. Best I could do was 1900 1:1 stable, and occasionally 1933, but not stable. Best I have done is 2100 with 2 sticks 1:1 not stable, and stable at 2K. Try with 2 sticks and see what happens.

This is some pretty expensive ram. I just posted this in hopes someone with the same dark hero board and similar ram would be like “oh yeah after updating bios my rig is loving these timings and fclk @ this ram etc lol

Link to comment
Share on other sites

Link to post
Share on other sites

5 hours ago, jaslion said:

Also running at 4000mhz is WORSE FOR PERFORMANCE for ryzen as at 3600mhz it runs 1:1 with the memory controller thus having the optimal transfer rate throughput.

Where are you reading this? From what I read the sweet spot is 1900 fclk and 3800.  But also there’s a guy who posted his results and was smashing good results on 2000 fclk and 4000. Apparently it’s just not stable and not many have ever got it to boot that way

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, StuSavage said:

Where are you reading this? From what I read the sweet spot is 1900 fclk and 3800.  But also there’s a guy who posted his results and was smashing good results on 2000 fclk and 4000. Apparently it’s just not stable and not many have ever got it to boot that way

I wasn't aware you were also ocing the fclk. If you are doing that make sure to keep it 1:1 ratio for the best performance even a little deviation ruins any gains you get. 2000 is golden sample territory so basically try to push it to 1900 and see what that gives. First get that rocksolid stable and try from there in tiny increments. It is pure luck the moment you start ocing fclk.

Link to comment
Share on other sites

Link to post
Share on other sites

14 minutes ago, jaslion said:

I wasn't aware you were also ocing the fclk. If you are doing that make sure to keep it 1:1 ratio for the best performance even a little deviation ruins any gains you get. 2000 is golden sample territory so basically try to push it to 1900 and see what that gives. First get that rocksolid stable and try from there in tiny increments. It is pure luck the moment you start ocing fclk.

Golden sample territory, nice. One can only hope they have the goods to make that happen. Lol I’m also oc’ing my 5950x on dos oc to give me the best single core and multi core performance. Just hoping I have some top notch bins

Link to comment
Share on other sites

Link to post
Share on other sites

11 minutes ago, StuSavage said:

Golden sample territory, nice. One can only hope they have the goods to make that happen. Lol I’m also oc’ing my 5950x on dos oc to give me the best single core and multi core performance. Just hoping I have some top notch bins

You did test it all before ocing so at stock 5950x and stock 3600mhz memory right? That way you have a baseline for scores as often these kinds of ocs can look right but be wrong and end up in worse performance.

Link to comment
Share on other sites

Link to post
Share on other sites

7 minutes ago, jaslion said:

You did test it all before ocing so at stock 5950x and stock 3600mhz memory right? That way you have a baseline for scores as often these kinds of ocs can look right but be wrong and end up in worse performance.

If testing it using user benchmark was enough:.. yes but if you’re referring to testing it stock anywhere else I have not

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, StuSavage said:

If testing it using user benchmark was enough:.. yes but if you’re referring to testing it stock anywhere else I have not

Userbenchmark is like the worst benchmark you can use. It is NOT at all accurate.

 

Use cinebench. Constant repeatable results and is a known good benchmark.

 

Always test stock. You wouldn't be the first person to be like oh look I got a 500mhz oc and then use the system only to notice that it is worse than someone with a cpu at stock :p. Believe me it is a lot more common than you think.

Link to comment
Share on other sites

Link to post
Share on other sites

1 hour ago, jaslion said:

Userbenchmark is like the worst benchmark you can use. It is NOT at all accurate.

 

Use cinebench. Constant repeatable results and is a known good benchmark.

 

Always test stock. You wouldn't be the first person to be like oh look I got a 500mhz oc and then use the system only to notice that it is worse than someone with a cpu at stock :p. Believe me it is a lot more common than you think.

Lol thanks tremendously 

Link to comment
Share on other sites

Link to post
Share on other sites

3 hours ago, StuSavage said:

This is some pretty expensive ram. I just posted this in hopes someone with the same dark hero board and similar ram would be like “oh yeah after updating bios my rig is loving these timings and fclk @ this ram etc lol

I hear you man. I have some pretty expensive ram too.. cost more than my board did 😄

 

But yeah.. probably wont see anything like that.

AMD R7 5800X3D | Thermalright Phantom Spirit 120 EVO, 1x T30

Asus Crosshair VIII Dark Hero | 32GB G.Skill Trident Z @ 3733C14

Zotac 4070 Ti Trinity OC @ 3060/1495 | WD SN850, SN850X, SN770

Seasonic Vertex GX-1000 | Fractal Torrent Compact RGB, Many CFM's

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×