Patriot Viper Steel 4000-19 Samsung B-Die Overclocking Results

[BerkayZ87]

Test System;

• MSI MPG B550 Gaming Plus
• AMD Ryzen 7 5800X @PBO +200 MHz CO -20 all cores
• Patriot Viper Steel 4000-19 2x8 Samsung B-Die

 

 

 

TestMem5 1usmus_v3 config 10 Cycles no error. Prime95 Large FFTs no WHEA. Prime95 Small FFTs no WHEA. MemTes86 no errors. AIDA 64 51.2ns. This is the tightest results for my bin.

 

 

 

I have bad IMC, can't go up 2000 FCLK; actually I can, but it wants 1.25 SoC and I wouldn't rather it for daily use and it's unnecessary for 67 MHz and 33 FCLK.

 

I took 51.2ns in AIDA64. Enough to me. Good luck for all overclockers. BTW, my CPU is overclocked to PBO2 +200 MHz and -20 CO for all cores.

 

 

[TheDankKoosh]

Very impressive results from this kit, I have the same bin running at 4000 16-16-16-32 1.45v in my 10850k rig, I should've gotten a gigabyte board in hindsight for the t-topology

I will be redoing all of my OC settings when I get the chance to swap my rig into a proper case with a 360mm aio so the z490 gaming edge might surprise me. 

[freeagent]

You are so close to breaking 50ns, you should lean on it just a little more! Good job man.

 

[BerkayZ87]

9 minutes ago, TheDankKoosh said:

Very impressive results from this kit, I have the same bin running at 4000 16-16-16-32 1.45v in my 10850k rig, I should've gotten a gigabyte board in hindsight for the t-topology

I will be redoing all of my OC settings when I get the chance to swap my rig into a proper case with a 360mm aio so the z490 gaming edge might surprise me. 

Giving voltage is the best solution for this chip  If your temps don't go 60 or higher, it is not a problem. I gave the voltage for 1.57V with stock LLC. Also, you can lower the timings, some of these timings are sensitive for voltage. Like tRCDRD, tRFC, tRC and tCL. I can go lower value for tRFC, but it is mindless. 295 is OK for me and for this voltage.

 

8 minutes ago, freeagent said:

You are so close to breaking 50ns, you should lean on it just a little more! Good job man.

 

Bad IMC causes high ns  Unluckiness. But I'm happy with 1967  Thanks. BTW your results are awesome.

[freeagent]

4 minutes ago, BerkayZ87 said:

Thanks.

No problem man!

 

4 minutes ago, BerkayZ87 said:

Thanks. BTW your results are awesome.

Thanks! The FCLK is not stable at that speed, errors start when I pass 2000 1:1. The memory was getting about 1.6v I think.

[RONOTHAN##]

I'm honestly most impressed by the 1967MHz FCLK. Every Ryzen 5000 series CPU I've used won't do above 1900MHz without negative scaling (my 5900X, for example, will do 26 seconds in Y Cruncher 1B at 3800MHz CL16-13-13-21 dual rank, I raise that up to 3866MHz with slightly looser timings, it jumps to over 5 minutes for the same test). You got real lucky with that CPU if it doesn't do that. 

 

 

[freeagent]

My 5900X is unstable once I pass 1900, dual or single rank. It turns into a whea generator. Vsoc, nothing helps. My 5600 is good for 2K, anything after that doesn’t help, and is also a whea generator at that point.

[Guest]

Fairly certain this B-Die at 4000mhz CL14-14-14 should be a pretty doable thing. At least on Intel's platform. 

 

Problem with AMD is that blasted Fclk limitation. While I read even the 5000 chips aren't really stable at 4000mhz, neither was my 2700X which happily does 3933mhz. No post at 4000. And no way to decouple the bastard either.

 

A proper set of B-Die will max out 4000mhz CL12 even CL11. However, my A2 single ranks won't do it. Only a slight bummer. 4200 CL16/17 seemed to produce better 3D scores anyways. 

[freeagent]

My royals will do 3600 flat 12s but my black and whites don’t want any part of C12

[BerkayZ87]

6 hours ago, Guest 5150 said:

Fairly certain this B-Die at 4000mhz CL14-14-14 should be a pretty doable thing. At least on Intel's platform. 

 

Problem with AMD is that blasted Fclk limitation. While I read even the 5000 chips aren't really stable at 4000mhz, neither was my 2700X which happily does 3933mhz. No post at 4000. And no way to decouple the bastard either.

 

A proper set of B-Die will max out 4000mhz CL12 even CL11. However, my A2 single ranks won't do it. Only a slight bummer. 4200 CL16/17 seemed to produce better 3D scores anyways. 

I tried tRCDRD to 14, it posts but unstable. I need more voltage for 14, maybe. But the voltage is already at 1.62V, still unstable That's why no longer I tried it. I dropped voltage to 1.57V, it's fine. When I do 1.56V, it gets unstable again.

[TheDankKoosh]

11 hours ago, BerkayZ87 said:

I tried tRCDRD to 14, it posts but unstable. I need more voltage for 14, maybe. But the voltage is already at 1.62V, still unstable That's why no longer I tried it. I dropped voltage to 1.57V, it's fine. When I do 1.56V, it gets unstable again.

I'd say anything above your voltage range isn't going to be the greatest for a daily OC anyhow, you got some great dimms and that OC as is will still cream 95% of other 5800x latency and bandwidth wise. I might try to copy these settings on my dimms and try to push voltage down if possible

[BerkayZ87]

56 minutes ago, TheDankKoosh said:

I'd say anything above your voltage range isn't going to be the greatest for a daily OC anyhow, you got some great dimms and that OC as is will still cream 95% of other 5800x latency and bandwidth wise. I might try to copy these settings on my dimms and try to push voltage down if possible

I bet that you can't do tRDRDSCL 2 and tWRWRCL 2  I bought this RAM to my two friends too; and they can't pass it. 2 no post, 3 unstable, 4 OK.

[TheDankKoosh]

2 hours ago, BerkayZ87 said:

I bet that you can't do tRDRDSCL 2 and tWRWRCL 2  I bought this RAM to my two friends too; and they can't pass it. 2 no post, 3 unstable, 4 OK.

We shall see, unfortunately asrock timing configurator doesn't have SCL timings included. Might try to do 4400 cl16 if I can push the timings well on 4000