Overclock.net banner

341 - 360 of 362 Posts

·
AMD OVERLOCKER
Joined
·
727 Posts
Got some bad luck, purchased the kit from newegg, arrived in one day, but 1 of the sticks was bad, and would fail to post at anything near XMP. though the other stick was fine.

Didn't have access to much hardware to test it on, but 1 stick was confirmed bad on a Asus ROG Strix X570-E (has the memory in the QVL), as well as an old Gigabyte Z170 Gaming 6 (too old to have such a kit in the QVL).



I will be sending it back to newegg, and I will try to get a replacement as it is b-die, and would love to get a stable pair that I can then hopefully have some luck on tightening timings on.



While the box that it came in had a few dents, the memory sticks did not have any visible damage. Newegg mails the RAM out in a padded envelope instead of a larger box.
The QVL doesn't mean anything.
Many good ram kits are not included there.

Sent from my ONEPLUS A6003 using Tapatalk
 

·
Registered
Joined
·
61 Posts
While that is the case, it does make it more tricky to fully troubleshoot, though in my case, 1 stick was bad enough that it was clear that it was faulty.

In a worst case scenario, if I end up getting errors while it is just at the XMP profiles and increasing voltages to the SOC and other aspects for stability do not help, then it will be hard or impossible for me to accurately narrow things down, thus I will have to hope I don't get 2 bad kits in a row.
 

·
Registered
Joined
·
883 Posts
Anybody experienced regression with the new AGESA ComboAm4PI 1.0.0.6 BIOS that MSI released ?

I am no longer TM5 stable when testing 1900 mhz fclk.

Previous BIOS was rock stable at 1900 mhz.

Cheers
 

·
Registered
Joined
·
911 Posts
Anybody experienced regression with the new AGESA ComboAm4PI 1.0.0.6 BIOS that MSI released ?

I am no longer TM5 stable when testing 1900 mhz fclk.

Previous BIOS was rock stable at 1900 mhz.

Cheers

No regression here and I even changed my RAM from DDR4 4000Mhz 16GB to DDR4 3600Mhz 32GB and still can run 1933Mhz FCLK without errors . Also tested DDR4 3600Mhz @3733Mhz/3800Mhz .

Also forgot Tested my DDR4 4133Mhz without problems also in mem test. DDR4 4133Mhz could run any speed but with 4 sticks of 4133Mhz was limited to DDR4 3600Mhz speed,motherboard limit with four dimms on this board
MSI X470 Gaming Plus with 3800X

EDIT:DOH your talking about Ram but I forgot to add a stupid Cinebench20 test with my new DDR4 3600Mhz Ram

My 3800X is getting Jealous of the New 3800XT but it is still fighting strong with a little BCLK
Score
Single 537
Multi 5333

Rant
Oh the nuts with the CPU voltage is too high ,just FO and learn how the CPU works.Also scores are lower when recording in real time and not using processor affinity to real time to get a higher score and then tell me about how your 3800X got a better score when I know the real truth.
NON OF THE RANT IS DIRECTED AT YOU mongoled,it is just for the nuts out there.



 

·
Registered
Joined
·
883 Posts
No regression here and I even changed my RAM from DDR4 4000Mhz 16GB to DDR4 3600Mhz 32GB and still can run 1933Mhz FCLK without errors . Also tested DDR4 3600Mhz @3733Mhz/3800Mhz .

Also forgot Tested my DDR4 4133Mhz without problems also in mem test. DDR4 4133Mhz could run any speed but with 4 sticks of 4133Mhz was limited to DDR4 3600Mhz speed,motherboard limit with four dimms on this board
MSI X470 Gaming Plus with 3800X

EDIT:DOH your talking about Ram but I forgot to add a stupid Cinebench20 test with my new DDR4 3600Mhz Ram

My 3800X is getting Jealous of the New 3800XT but it is still fighting strong with a little BCLK
Score
Single 537
Multi 5333

Rant
Oh the nuts with the CPU voltage is too high ,just FO and learn how the CPU works.Also scores are lower when recording in real time and not using processor affinity to real time to get a higher score and then tell me about how your 3800X got a better score when I know the real truth.
NON OF THE RANT IS DIRECTED AT YOU mongoled,it is just for the nuts out there.
https://www.youtube.com/watch?v=HP4jJ9IImNo
The rant

:D :D

I understand exactly where you are coming from, lol

So some new information and thanks for relaying your experiences.

I am having to retest again as settings that were "known" to work at different fclk are no longer "valid" with this new BIOS for my X370.

Managed to determine that VDDP voltage now needs to be alot lower than previous agesa version otherwise TM5 errors with codes that I have rarely seen and with no specific "pattern".

Before 3800/1900 required [email protected], [email protected] and [email protected]

Dropping VDDP to 0.805 - 0.850 (still testing) stops almost all of those weird and wonderous error codes in TM5.

Still need to play with vSOC as this probably needs further tweaking along with VDDG CCD/IOD.

And back to your rant,

Currently running/testing [email protected], this gives me 3800/1900, previous fully tested stable BCLK was 106.725 which gave me 3766/1888.

Attempting to balance alot of things with TM5 running all core loads between 4490-4520 mhz while finding the right balance of voltages and LLC to keep everything nice and stable while at the same time maintaining a clean 3800/1900 mclk/fclk combo while at the same time using [email protected]

:p :p

Performance is amazing for this 3600 with the combination described above.

Oh and one small cavet, I have to reduce my PCIe to Gen2 to be able to get to OS with such high BCLK otherwise the VGA goes all wonky

:D :D

I actually get better 3D performance with my hardware using Gen2 rathar than Gen3 even when not using BCLK, so really no loss.

Have yet to experience any anomolies with my nVme/SATA/USB device so far

:thumb:
 

·
Registered
Joined
·
911 Posts
Some good info there and thanks for the tip using Gen2 rathar than Gen3,I am going to test this out.My cheap Motherboard can not handle high BCLK.
 

·
Registered
Joined
·
883 Posts
Some good info there and thanks for the tip using Gen2 rathar than Gen3,I am going to test this out.My cheap Motherboard can not handle high BCLK.
Did you get to do some testing ?
 

·
Registered
Joined
·
4 Posts
I know this is a X370/X470/X570 thread, but I didn't find a dedicated one for the B Series chipsets, so I'm going to ask here: has anyone had any luck pushing Samsung B-Dies to 3800MHz CL12 on a B450 board? I have a F4-3600C16D-16GVK Kit, which was bought last month. According to Thaiphoon each module has an A1 10-layer PCB, but I'm pretty sure it's an A2 layout, so that reading should be wrong. Anyways, I am running a Ryzen 5 3600, which can go as high as 4.6-4.7GHz on air for benchmarking and so far the memory sticks have been running at 3800MHz (1900MHz FCLK) 14-14-14-28-1T-42-260, GDM off. I haven't had any luck booting at CL12 even with GDM enabled, so I'm thinking either the bin of the kit is not that good, or the mainboard is just bad at memory overclocking, it's only a MSI B450 Tomahawk MAX. If anyone has any experience or suggestions, I would be glad to hear them. Cheers.


P.S.: I am about to try the latest BIOS update with the AGESA 1.0.0.6 tonight, gonna report if that makes any difference.
 

·
Registered
Joined
·
26 Posts
I know this is a X370/X470/X570 thread, but I didn't find a dedicated one for the B Series chipsets, so I'm going to ask here: has anyone had any luck pushing Samsung B-Dies to 3800MHz CL12 on a B450 board? I have a F4-3600C16D-16GVK Kit, which was bought last month. According to Thaiphoon each module has an A1 10-layer PCB, but I'm pretty sure it's an A2 layout, so that reading should be wrong. Anyways, I am running a Ryzen 5 3600, which can go as high as 4.6-4.7GHz on air for benchmarking and so far the memory sticks have been running at 3800MHz (1900MHz FCLK) 14-14-14-28-1T-42-260, GDM off. I haven't had any luck booting at CL12 even with GDM enabled, so I'm thinking either the bin of the kit is not that good, or the mainboard is just bad at memory overclocking, it's only a MSI B450 Tomahawk MAX. If anyone has any experience or suggestions, I would be glad to hear them. Cheers.


P.S.: I am about to try the latest BIOS update with the AGESA 1.0.0.6 tonight, gonna report if that makes any difference.
Wow...can you tell me your specific setting on timing and voltage on yours?

I find it strange on mine, I have F4-3600C16D-32GTZN and MSI B550 Tomahawk on Ryzen [email protected]@1.275v
I have be able to boot and benchmark with IF:MCLK 1:1 @1900Mhz with CL14-14-14-36-1T GDM on @1.5v, however its not past the memtest as it resulted in error. Than I follow all instructions from Ryzen DRAM Calculactor v1.7.3 in safe and fast setting, and it worsen the error in memtest.

In my testing, upping SoC voltage give negative result on memtest. If I run with all the timing stated in DRAM calculator v1.7.3 and it provide more error in memtest even with DRAM voltage @1.55v. Recently I lowering the timing to CL16-16-16-32-48-1T-GDM on-358, lowering the SoC to only 1.05v, change the setting to AMD overclock in SoC voltage, change the DRAM voltage to only 1.43v, and all the rest of timing to voltage to AUTO, and it PASSED the memtest over 100%. (other setting on RAM are, BGS Disable are Powerdown Disable)

So, it might be in case of B550 or only in my case of RAM and mobo, rising the voltage sometimes not beneficial in RAM overclocking.

I attached my stable setting against safe timing recommended by DRAM Ryzen Calculator v1.7.3 which is not stable.

And since I can boot and bench at 3800Mhz CL14 but not stable in memtest (I have tried 14-15 and 14-16 as well), I am very curious in case any of you with the same ram and mobo can push it to 3800Mhz at CL14 stable?
 

Attachments

·
Registered
Joined
·
883 Posts
Wow...can you tell me your specific setting on timing and voltage on yours?

I find it strange on mine, I have F4-3600C16D-32GTZN and MSI B550 Tomahawk on Ryzen [email protected]@1.275v
I have be able to boot and benchmark with IF:MCLK 1:1 @1900Mhz with CL14-14-14-36-1T GDM on @1.5v, however its not past the memtest as it resulted in error. Than I follow all instructions from Ryzen DRAM Calculactor v1.7.3 in safe and fast setting, and it worsen the error in memtest.

In my testing, upping SoC voltage give negative result on memtest. If I run with all the timing stated in DRAM calculator v1.7.3 and it provide more error in memtest even with DRAM voltage @1.55v. Recently I lowering the timing to CL16-16-16-32-48-1T-GDM on-358, lowering the SoC to only 1.05v, change the setting to AMD overclock in SoC voltage, change the DRAM voltage to only 1.43v, and all the rest of timing to voltage to AUTO, and it PASSED the memtest over 100%. (other setting on RAM are, BGS Disable are Powerdown Disable)

So, it might be in case of B550 or only in my case of RAM and mobo, rising the voltage sometimes not beneficial in RAM overclocking.

I attached my stable setting against safe timing recommended by DRAM Ryzen Calculator v1.7.3 which is not stable.

And since I can boot and bench at 3800Mhz CL14 but not stable in memtest (I have tried 14-15 and 14-16 as well), I am very curious in case any of you with the same ram and mobo can push it to 3800Mhz at CL14 stable?
CL14 needs more voltage than 1.5v and fan on dimms.
 

·
Edgy & on the edge
Joined
·
1,540 Posts
I know this is a X370/X470/X570 thread, but I didn't find a dedicated one for the B Series chipsets, so I'm going to ask here: has anyone had any luck pushing Samsung B-Dies to 3800MHz CL12 on a B450 board? I have a F4-3600C16D-16GVK Kit, which was bought last month. According to Thaiphoon each module has an A1 10-layer PCB, but I'm pretty sure it's an A2 layout, so that reading should be wrong. Anyways, I am running a Ryzen 5 3600, which can go as high as 4.6-4.7GHz on air for benchmarking and so far the memory sticks have been running at 3800MHz (1900MHz FCLK) 14-14-14-28-1T-42-260, GDM off. I haven't had any luck booting at CL12 even with GDM enabled, so I'm thinking either the bin of the kit is not that good, or the mainboard is just bad at memory overclocking, it's only a MSI B450 Tomahawk MAX. If anyone has any experience or suggestions, I would be glad to hear them. Cheers.


P.S.: I am about to try the latest BIOS update with the AGESA 1.0.0.6 tonight, gonna report if that makes any difference.

3800 CL12 ain't happening without at least 1.8V DRAM Voltage which isn't really ideal for daily use. It's almost XOC stuff. Flat 14s at 3800 is already great stuff if it's actually stable.
 

·
Registered
Joined
·
193 Posts
I run CL14 and have my mem V set at 1.48V in the BIOS which seems to bring it to 1.51 when in Windows according the HWInfo62.
 

Attachments

·
Registered
Joined
·
883 Posts
I run CL14 and have my mem V set at 1.48V in the BIOS which seems to bring it to 1.51 when in Windows according the HWInfo62.
Did you stress test that with 25 cycles of TM5 and than a couple of hours of Y-Cruncher with all tests enabled ??

As its very easy to complete membench easy and AIDA....
 

·
Registered
Joined
·
193 Posts
Did you stress test that with 25 cycles of TM5 and than a couple of hours of Y-Cruncher with all tests enabled ??

As its very easy to complete membench easy and AIDA....



Yes I did and it was fine. TM5 that is not Y-Cruncher.
 

·
Registered
Joined
·
883 Posts
Yes I did and it was fine. TM5 that is not Y-Cruncher.
OK, TM5 is good, hope you did some large FFT testing also which is what the Y-Cruncher suite does as TM5 is not the be all and end all of stability, many have been TM5 25 cycle stable, but then they find they have instabilities because they didnt do any FFT testing...

:)
 

·
Premium Member
Joined
·
5,021 Posts
Yes I did and it was fine. TM5 that is not Y-Cruncher.
Y-Cruncher is a terrible idea on 3000 series CPUs. Open HWInfo when it's running. Second test and beyond with my BIOS voltage at 1.2625v, 1.248v VR VOUT In HWInfo I'm getting 90c temps while running it with an Optimus Foundation block, 1 360 rad, CPU only on it and cool ambient temps. :eek:

If I run my 24/7 CCX overclock voltages it hits 95C. :(

I never like to go above 80c while stress testing.

Better to use Prime95 1344 FFTs In Place with a 15 minute time interval. Temps stay around 70C and it if passes that overnight you can be sure your CPU is stable with TM5 1usmus_v3 passing 25 rounds done separately. :D
 

·
Registered
Joined
·
883 Posts
Y-Cruncher is a terrible idea on 3000 series CPUs. Open HWInfo when it's running. Second test and beyond with my BIOS voltage at 1.2625v, 1.248v VR VOUT In HWInfo I'm getting 90c temps while running it with an Optimus Foundation block, 1 360 rad, CPU only on it and cool ambient temps. :eek:

If I run my 24/7 CCX overclock voltages it hits 95C. :(

I never like to go above 80c while stress testing.

Better to use Prime95 1344 FFTs In Place with a 15 minute time interval. Temps stay around 70C and it if passes that overnight you can be sure your CPU is stable with TM5 1usmus_v3 passing 25 rounds done separately. :D
So you are cheating your way to stability

:p
 

·
Registered
Joined
·
193 Posts
OK, TM5 is good, hope you did some large FFT testing also which is what the Y-Cruncher suite does as TM5 is not the be all and end all of stability, many have been TM5 25 cycle stable, but then they find they have instabilities because they didnt do any FFT testing...

:)

I am also happy that it never crashes when using it. I do also have a fan over the RAM to keep it cooler as it was warmer than I liked with no fan, I like to keep in in the 30's during normal use and gaming. Running the tests it hit the mid 40's so I am all good.
 

·
Premium Member
Joined
·
5,021 Posts
So you are cheating your way to stability


:p
Open HWInfo when it's running and check your temps Test 2 and past it, it isn't good.

No way you want those temps even while stress testing. :(
 

·
Registered
Joined
·
883 Posts
Open HWInfo when it's running and check your temps Test 2 and past it, it isn't good.

No way you want those temps even while stress testing. :(
It was just in jest

:p

Yeah I am well aware of the temps,

that why I stopped using all core overclock with fixed voltage and reverted to BCLK with EDC bug

;)

And even using PBO to do the clocking I do see such temps but thats the norm for AVX2 workloads, the CPUs are designed for that.

So I abuse it a little with the EDC bug, but im not going to be keeping the CPU for a decade and the stess test is just for several hours so I know its stable with whatever I happen to throw at it.

I prefer peace of mind with regards to stability over worrying about a little heat.

:)
 
341 - 360 of 362 Posts
Top