1.5v safe Samsung Bdie?

opethdisciple · 4 May 2020 at 21:23

So I decided to just give it a go. Not run my ram this high in voltage before as been scared.

But I took the dram calc fast timings put them in and upped the voltage to 1.5v from the 1.45v it recommends. (It doesn't work at 1.45v)

I know it's been said before that 1.5v is safe on bdie but I'm just checking again. I don't want to break my mobo or something if the ram was to break at the same time.

I'm running memtest at the moment and it's been running for 17 mins with no issues so I'm getting excited that this might actually work!

opethdisciple · 5 May 2020 at 23:25

Well I tried some CL14 timings which are not as aggressive as the dram calc and everything seems to be fine at 1.45v.

The dram calc suggests timings like: 3600MHz 14-15-14-14-28 at 1.45v which don't work. I haven't tried it at 1.5v however.

I'm a little bit scared to be honest to run it that high and am happy at what I have managed to achieve with them at the moment.

Strangely this Gigabyte motherboard seems to add more voltage than what you specify in the bios. I put in 1.45v and can see it is running the ram at 1.47v.

Any way I've tested it with hci memtest for 400% and have gamed on it for the last two evenings so so far so good.

opethdisciple · 8 May 2020 at 14:47

wookiee87 said:
I've just ordered two kits of Team Group 8Pack RIPPED Edition 16GB (2x8GB) DDR4 PC4-28800C14 3600MHz Dual Channel ram, would like to see if there is a performance difference from my cl18 kit, i suspect there will but, a question if i have to run close to 1.5v to get it stable will airflow be needed or will the ram work, just run warm?

The XMP profile should be all you need to apply.

opethdisciple · 8 May 2020 at 17:38

lnoton said:
I think you'll have to run 1.4-1.5v to get the xmp tbh. Looking forward to seeing the results from you guys. I'm sure they'll be better for many, and only those with crappy IMC will see negligible gains

But that's what I mean, the XMP/DOCP profile sets the voltage and timings not just the timings as that wouldn't be stable.

opethdisciple · 8 May 2020 at 17:40

I have my ram at 1.45v in the bios, but I can see the board is applying 1.47v.

I tried to see how low they would go, but 1.43v wasn't stable, and only 1.44v was left to try so the fact that things where stable at 1.45v I just left it alone.

opethdisciple · 8 May 2020 at 17:43

wookiee87 said:
and temps are ok? around the 55-60 degree mark or lower?

I assume so. I tried looking at temps using hwinfo but it doesn't seem to have a reading for ram. It's been game stable for a couple of gaming sessions of Doom Eternal as well as passes 400% HCI so I assume all is OK.

opethdisciple · 11 May 2020 at 20:14

Shac said:
Samsung "B" die spec sheet.

https://www.samsung.com/semiconduct...11/8G_B_DDR4_Samsung_Spec_Rev2_1_Feb_17-0.pdf

Table of contents number 6. Absolute Maximum DC ratings.

Voltage 1.5V which means they are guaranteed to run and be safe at that.

Operating Temperature Range 0° to 95°C.

Notes 1 and 2 for the above state:
1. Operating Temperature TOPER is the case surface temperature on the center/top side of the DRAM.
2. The Normal Temperature Range specifies the temperatures where all DRAM specifications will be supported. During operation, the DRAM case temperature must be maintained between 0-85°C under all operating conditions

If you wish them to operate in the 85°C to 95°C range then the following changes must be made.
a) Refresh commands must be doubled in frequency, therefore reducing the refresh interval tREFI to 3.9us.
b) If Self-Refresh operation is required in the Extended Temperature Range, then it is mandatory to use the Manual Self-Refresh mode with Extended Temperature Range capability (MR2 A6 = 0b and MR2 A7 = 1b).

So Samsung are saying that 1.5V and below 85°C is safe and within spec.

Excellent news and good find.

opethdisciple · 11 May 2020 at 22:32

Yea, I have mine at 1.45v in the bios but the board is applying 1.47v.

opethdisciple · 13 May 2020 at 11:50

I thought I was stable at 1.45v. It passed 400% HCI and about 7-8 hours of gaming.

However despite being stable in the hecticness of Doom Eternal for hours it was 20mins of BF5 which has shown me it isn't stable.

Currently at 1.46v in the bios (probably closer to 1.48v as the bios applies extra voltage)

It allowed me to get through rounds without a freeze compared to the 10-20mins duration it took to freeze at 1.45v.

Time will tell.

One question is it worth it?

14 15 15 15 35 @ 1.46v @ 3600MHz

vs

16 16 16 16 36 @ 1.35 @ 3600MHz

opethdisciple · 13 May 2020 at 23:51

Well with 1.46v in the bios (1.476-1.488 values actually applied by bios) I was stable for a 3hr session in BFV with no crashing.

For me this is a good sign. I think BFV is a good stability tester. Yesterday at 1.45v in the bios I was crashing every 20 mins or so of gaming.

The internet seems to say running these modules at such a high voltage up to 1.5v is perfectly fine... so I hope this is OK to run like this.

opethdisciple · 15 May 2020 at 12:48

FireTwo said:
very safe tbh 1.5volts

I would feel ok running 24/7 with 1.55 maybe 1.6.

OK, cool, thanks for that. I am now at 1.47v in the bios but actually 1.48-1.5v once the bios overvolts.

I currently don't have GDM or powerdown mode etc disabled and when I try I get memory errors within seconds.

You reckon I have more wiggle room to increase voltage another tick or two?

This will definitely push me over the 1.5v threshold.

opethdisciple · 15 May 2020 at 15:22

Yup, no amount of fiddling with voltages would allow me to disable gear down mode.

1.55v dram and all the other voltages form the dram calc and it still wasn't stable.