What kinda of RAM is this!?

[SecretX]

Long story short, I bought memory on AliExpress, I opened Thaiphoon and the software could not read what kind of die this memory has. Is there a way to know for sure what kind of die do this memory have? And yes, I know I am not running D.O.C.P nor adjusted the timings, the thing is the PC parts just arrived yesterday and I didn't had the time nor the patience to actually ajust the timings manually, so that's why I want to use DRAM Calculator for Ryzen.

 

 

 

[.Apex.]

"PC4-3200AA-UB1-11" only returns Micron chips which seems to be an E-Die/Rev.E, not a particularly exceptional die and you could test this by going below 17 TRCD on any frequency, if it fails to achieve those timings then that would confirm that it's an E-Die

[Tan3l6]

32 minutes ago, Syn. said:

"PC4-3200AA-UB1-11" only returns Micron chips which seems to be an E-Die/Rev.E, not a particularly exceptional die and you could test this by going below 17 TRCD on any frequency, if it fails to achieve those timings then that would confirm that it's an E-Die

Is "rev. e" some inferior E-die? Edit:  misread, you just mean e-die

As my E-die sticks do 4000Mhz easily though with high timings. Have not tried to fine tune. But I'm on Intel platform.

[SecretX]

Here the pic of the ram

 

[Tan3l6]

OP - Your Thaiphoon is not the latest. Latest version always adds new info.

[.Apex.]

14 minutes ago, Tan3l6 said:

Is "rev. e" some inferior E-die? Edit:  misread, you just mean e-die

As my E-die sticks do 4000Mhz easily though with high timings. Have not tried to fine tune.

Yeah it is referred to as Micron Rev.E or E-Die, the chip is more suited for high frequency applications, for sub-3600Mhz clocks it is considered to be low tier because of its TRCDRD limitation which on Intel you'd have to adjust TRCDRD and TRCDWR in unison and that makes it even worse, but once you push higher clocks then that won't be much of an issue anymore and it performs well

[Guest]

It's value ram. at best.

[Tan3l6]

16 minutes ago, ShrimpBrime said:

It's value ram. at best.

Maybe for an OC enthusiast there is nothing like Samsung B-die.

I like this test for Ryzen 3000 series - https://lab501.ro/procesoare-chipseturi/amd-ryzen-3000-ddr4-scaling-part-ii-samsung-b-die-vs-micron-e-die-vs-hynix-cjr-english-version

[SecretX]

1 hour ago, Tan3l6 said:

OP - Your Thaiphoon is not the latest. Latest version always adds new info.

 

 

I mean I don't know, I got this version from Wayback Machine since the website does not load.

[Tan3l6]

Just now, SecretX said:

I mean I don't know, I got this version from Wayback Machine since the website does not load.

The website works fine for me. Latest version is 16.3.4.0 by the way.

[SecretX]

Just now, Tan3l6 said:

The website works fine for me. Latest version is 16.3.4.0 by the way.

Would you mind upload it to the cloud then send me the link?

[Tan3l6]

http://www.softnology.biz/files/thphn163.zip

 

That doesn't work? I'd rather not upload it to somewhere else.

 

Also - https://www.softpedia.com/get/Tweak/Memory-Tweak/Thaiphoon-Burner.shtml#download

[SecretX]

5 minutes ago, Tan3l6 said:

Also - https://www.softpedia.com/get/Tweak/Memory-Tweak/Thaiphoon-Burner.shtml#download

This link worked, I think the official website blacklisted Brazil or South America for some reason, 'cause I saw other people reporting same problem (404 timed out) on brazilian forums.

 

Updated info

 

[Guest]

1 hour ago, Tan3l6 said:

Maybe for an OC enthusiast there is nothing like Samsung B-die.

I like this test for Ryzen 3000 series - https://lab501.ro/procesoare-chipseturi/amd-ryzen-3000-ddr4-scaling-part-ii-samsung-b-die-vs-micron-e-die-vs-hynix-cjr-english-version

My B-Die is rated 4267mhz at CL19. 

This is quite a bit different than E-Die at CL19 3200mhz. It's low bin ICs. 

 

I'd be surprised if this kit can run it's XMP rated frequency just by setting DOCP enabled in bios....

 

[SecretX]

13 minutes ago, ShrimpBrime said:

My B-Die is rated 4267mhz at CL19. 

This is quite a bit different than E-Die at CL19 3200mhz. It's low bin ICs. 

 

I'd be surprised if this kit can run it's XMP rated frequency just by setting DOCP enabled in bios....

I let MemTest86 run for 1hr 30min with this single memory stick running @3200mhz 16-18-18-36 1.3v (yes, just one RAM stick on the moba) and no error occured, it's not that bad...

[Guest]

Just now, SecretX said:

I let MemTest86 run for 1hr 30min with this stick running @16-18-18-36 1.3v and no error occured, it's not that bad...

At what Frequency? 

Stick? Singular only? 

[SecretX]

9 minutes ago, ShrimpBrime said:

At what Frequency? 

Stick? Singular only? 

My bad, I'm gonna edit it. 3200mhz single stick

[Guest]

Just now, SecretX said:

My bad, I'm gonna edit it. 3200mhz single stick

Quote

I am not running D.O.C.P nor adjusted the timings

Sorry, I must have missed where you've been making adjustments.

 

I would aim for a CL14 at max Cas latency for this frequency (3200mhz). You have yet some testing to do.

Test tight at lower frequencies, and At higher, you want to run no looser than the CL19.

 

Then if you purchase a second stick, will you be able to maintain the same settings? 

 

Single stick memory overclocking usually yields the best results of that single stick of memory. This wouldn't be reflective of a dual channel overclock.

I

 

[SecretX]

1 minute ago, ShrimpBrime said:

Then if you purchase a second stick, will you be able to maintain the same settings?

I don't know, probably not but It will take a while for me to be able to buy a second memory stick, and until then I want to enjoy a faster system overall.

 

2 minutes ago, ShrimpBrime said:

Single stick memory overclocking usually yields the best results of that single stick of memory. This wouldn't be reflective of a dual channel overclock.

Yes I know, don't you worry, I am going to readjust RAM timings once I get my second memory stick.

 

BTW what would be a good way to measure how much gain did the system gain with every timing tightening?

[Guest]

Just now, SecretX said:

I don't know, probably not but It will take a while for me to be able to buy a second memory stick, and until then I want to enjoy a faster system overall.

 

Yes I know, don't you worry, I am going to readjust RAM timings once I get my second memory stick.

 

BTW what would be a good way to measure how much gain did the system gain with every timing tightening?

You would use the program called AID64. At the top, in the tools tab, you want to run the cache and memory benchmark. 

[SecretX]

@ShrimpBrime I did it, I run AIDA64 with both "let moba decide" timings vs the timings printed on the memory stick.

 

Auto timings (20-22-22-52) @1.2v

 

Manufacture timings (?) - 16-18-18-36 @1.3v

[Guest]

54 minutes ago, SecretX said:

@ShrimpBrime I did it, I run AIDA64 with both "let moba decide" timings vs the timings printed on the memory stick.

 

Auto timings (20-22-22-52) @1.2v

 

 

Manufacture timings (?) - 16-18-18-36 @1.3v

 

IT's in the general realm or normal I think. Would aim to get sub 70ms latency if possible.

 

 

[imajerec]

if it's micron e-die then is not low tier. it's excelent value. i bought some crucial ballistix 3000mhz cl15 (new model,it's advertised as micron tuned die). running it at 3600mhz cl16 at 1.29V that's crazy low. with a ryzen 3200g. previous kits that i had, some corsair vengeance lpx 3600C18 , and g.skill ripjaws V 3600C19 (i think both kits were samsung c-die) couldn't be ran at xmp preset at the advertised 3600mhz frequency, it simply gave errors in memtest and prime95 large fft after few seconds! the crucial kit i have now is working great, it can even do 3733C16 at 1.35v but i have to raise the soc voltage above 1.2v to stabilise the integrated vega 8 graphics and i'm not too comfortable raising soc voltage that high. 

 

and it could proably be OC even more if the zen+ i'm using had superior IMC i'm pretty sure it could do 3800mhz CL16. so , micron e-die is an excelent OC and great value RAM.

[imajerec]

1 hour ago, SecretX said:

@ShrimpBrime I did it, I run AIDA64 with both "let moba decide" timings vs the timings printed on the memory stick.

 

Auto timings (20-22-22-52) @1.2v

 

Manufacture timings (?) - 16-18-18-36 @1.3v

that read/write should be almost twice higher, you're running single channel and that affects performance considerable