G.Skill DDR5 6000 MHz CL 36 - Z690 Extreme - Temperatures

[uplink]

Hey there guys!

Thanks to Hopper64 and Shamino, I've managed to run two sticks of G.Skill DDR5 6000 MHz CL36 16GB modules on Z690 Extreme...and now here's my worry.

I know zip about DDR5, I know very little about DDR4 so...

I'm running 2second full run of MemTest on my benchtable and my average temperature of RAM stick is 40°C [+-], most of the time it hovers on 32 - 36°C, but there are ones and zeroes and one more test, that pumps the temps of sticks up to almost 60°C.

I have no errors whatsoever.

My question is, I plan to run two kits, on let's say 5600 MHz, or 5200 MHz, I'll see how they run and if they run together at all. I'd like to keep "vanilla" XMP I. or XMP II. profile with Hopper64's VDD 1.35 and Shamino's VDDQ 1.3575. They keep my RAM steady!

But my temperatures are as follows:



Are 35 - 40°C in most of the tests okay? And 55°C - 60°C in 1s & 0s and hammering in MemTest?

[Silent_Scone]

Hello,

Those temps are normal, especially without active airflow. Thanks to the PIMC the modules pack more heat.

This is well within tolerance, the only "danger" is when pushing the modules to the edge of stability as when stability margins decrease, so does the module's tolerances to temperature and active airflow might be required to sustain overclocking stability (anything above 55-60c can become problematic for stability with Samsung IC).

All par for the course, and simply needs to be kept in mind when pushing the memory IC limits.

[warreng090]

I have the G.Skill 6000 MHz CL40 on an Apex board. It is tuned to C32 and it hits 60C on memtest, 50C while gaming. Front intake fan hitting it as well as a 60mm Noctua fan mounted right above the ram pulling hot air up.

[jackson_lab]

warreng090 wrote:
I have the G.Skill 6000 MHz CL40 on an Apex board. It is tuned to C32 and it hits 60C on memtest, 50C while gaming. Front intake fan hitting it as well as a 60mm Noctua fan mounted right above the ram pulling hot air up.

Z690Apex_Gskill Trident Z5DDR5-6000MHz CL40-40-40-76 1.30V32GB (2x16GB)/I9-12900k
VDD 1.3
VDDQ 1.3
IVR TX VDDQ 1.3
IVR VCCSA 1.1
IMC VDD 1.25

I have same setup but my sticks start erroring out above 52C. Memtest86 easy to pass (no problems there) but OCCT or MemTest5(absolut) for at least an hour- errors start right at 52C and up. Turn fans on max to get sticks below 52C and errors stop. Turn fans off or low , memory temps creep up again and errors start. Can you run OCCT for at least 45 minutes above 55C and still pass? I'm trying to find out if folks with 6000mhz Gskills can keep their ram running error free with just the stock passive heatsinks that come with it. 99% of us will not seek aftermarket cooling solutions for our ram to run at XMP or below. My solution was to manually set my ram to (5600mhz 36-36-36-76) and tweak voltages. Same as above except IMC VDD lowered to 1.15. They now can run all these stressful memory tests topping out at 62C and no errors! 62C seems to be where the stock Gskill heatsinks level off with no external cooling applied. BUT 6000mhz and above is very sensitive to heat. Comments anyone? Thank you.

[Yunior4GT]

jackson_lab wrote:
Z690Apex_Gskill Trident Z5DDR5-6000MHz CL40-40-40-76 1.30V32GB (2x16GB)/I9-12900k
VDD 1.3
VDDQ 1.3
IVR TX VDDQ 1.3
IVR VCCSA 1.1
IMC VDD 1.25

I have same setup but my sticks start erroring out above 52C. Memtest86 easy to pass (no problems there) but OCCT or MemTest5(absolut) for at least an hour- errors start right at 52C and up. Turn fans on max to get sticks below 52C and errors stop. Turn fans off or low , memory temps creep up again and errors start. Can you run OCCT for at least 45 minutes above 55C and still pass? I'm trying to find out if folks with 6000mhz Gskills can keep their ram running error free with just the stock passive heatsinks that come with it. 99% of us will not seek aftermarket cooling solutions for our ram to run at XMP or below. My solution was to manually set my ram to (5600mhz 36-36-36-76) and tweak voltages. Same as above except IMC VDD lowered to 1.15. They now can run all these stressful memory tests topping out at 62C and no errors! 62C seems to be where the stock Gskill heatsinks level off with no external cooling applied. BUT 6000mhz and above is very sensitive to heat. Comments anyone? Thank you.

Hi, I have the Trident Z5 32GB 6000 DDR5 on a Maximus Z690 Apex board and have the same high temps. I been going nuts trying to find some clear answer as to if these temps are normal or not. On idle, with XMP on, I average between 40-45c. When I run a stress test, it went as high as 66c before I got nervous and stopped the test. It seemed it had reached peak temp at 66c, but I didn’t want to risk it. I’m using the fractal torrent case which has good airflow together with Kraken z63 cooler. So, is this in deed normal temps? Should I be worried? GSkill doesn’t really state anywhere what normal temps should be for these new RAM.Â*

[wing_man]

jackson_lab wrote:
Z690Apex_Gskill Trident Z5DDR5-6000MHz CL40-40-40-76 1.30V32GB (2x16GB)/I9-12900k
VDD 1.3
VDDQ 1.3
IVR TX VDDQ 1.3
IVR VCCSA 1.1
IMC VDD 1.25

I have same setup but my sticks start erroring out above 52C. Memtest86 easy to pass (no problems there) but OCCT or MemTest5(absolut) for at least an hour- errors start right at 52C and up. Turn fans on max to get sticks below 52C and errors stop. Turn fans off or low , memory temps creep up again and errors start. Can you run OCCT for at least 45 minutes above 55C and still pass? I'm trying to find out if folks with 6000mhz Gskills can keep their ram running error free with just the stock passive heatsinks that come with it. 99% of us will not seek aftermarket cooling solutions for our ram to run at XMP or below. My solution was to manually set my ram to (5600mhz 36-36-36-76) and tweak voltages. Same as above except IMC VDD lowered to 1.15. They now can run all these stressful memory tests topping out at 62C and no errors! 62C seems to be where the stock Gskill heatsinks level off with no external cooling applied. BUT 6000mhz and above is very sensitive to heat. Comments anyone? Thank you.

I found the same thing with my G.Skill DDR5-6000MHz cl36 36 36 76. 5600Mhz was as fast as I could go because of temperature. I just set xmp1 and lowered the frequency. Increasing the voltage just increases the temperature. I could run xmp1 at 6000Mhz but would fail stress test when ram got hot. Need a fan directly on the ram or a larger heat sink on the ram.

[bad_turbulence]

wing_man wrote:
I found the same thing with my G.Skill DDR5-6000MHz cl36 36 36 76. 5600Mhz was as fast as I could go because of temperature. I just set xmp1 and lowered the frequency. Increasing the voltage just increases the temperature. I could run xmp1 at 6000Mhz but would fail stress test when ram got hot. Need a fan directly on the ram or a larger heat sink on the ram.

Same exact G.Skill here: Shouldn't it be working pretty efficiently at the 4800 speed that the ASUS ROG Strix "E" z690 DDR5 mobo sets it to (with XMP disabled)? I get memory crashes in games. I am running windows 11 and unfortunately, I loaded that ASUS Armory Crate (i have since cleaned all that out).... but I've tried everything and I still get crashes and I am not overclocking my 12900K whatsoever. I called G.Skill and he told me to run memtest86 on each stick separately, which i did and it passes fine separately or even with both in there (all 4 passes are good any way I try), windows 11 memory test passes fine and chkdsk and all other diag utility pass (SFC, etc. ) I just loaded up SIV so that i can monitor the G.Skill temps during gameplay. But sometimes i get a memory fail just loading the game. LIke the application error instruction at XXXXXXXXXXXX referenced memory at Ox00000000005 and things like that (event viewer) CTD's and pop up memory messages like "The memory could not be read" ANd on and on. I've literally tried everything. I'm in Forza 5 and SIV says that each of the two sticks are 35.8C and 33.5C (don't seem to be running hot) after about 20 minutes of running Forza 5. Im sure I'll get a memory error any time since I can't believe I'm at the 20 minute mark. The memory label sticker on the box says "F5-6000U4040E16GX2-TZ5RS" DDR5-6000, 16GBx2, CL40-40-40-76 1.30v". One last thing is that I did find some ASUS services still running in Windows "Services" so I disabled them all (about 5 or 6 of them) and there was this big ASUS folder still in the Windows x86 folder, i went into safe mode and deleted it ..... Can't believe that stuff was still in there even after uninstalling Armory Crate using add/remove programs. Is Amrory Crate needed (at least for my testing while these memory errors are occurring?) Hopefully not since I just took the last stuff out (my guess is that there is still a bunch of Armory Crate stuff in my registry). But anyway, any help is appreciated. Last little nugget is that if i turn XMP to enable, it reads the proper 6000MHz clock on the memory and the proper CL values in the BIOS and I get memory crashes much faster with it enabled. But again, I'm memory crashing even at 4800 with XMP disabled.

[bad_turbulence]

WOW, i changed to windows 10, no errors :confused:

[uplink]

Silent Scone@ROG wrote:
Hello,

Those temps are normal, especially without active airflow. Thanks to the PIMC the modules pack more heat.

This is well within tolerance, the only "danger" is when pushing the modules to the edge of stability as when stability margins decrease, so does the module's tolerances to temperature and active airflow might be required to sustain overclocking stability (anything above 55-60c can become problematic for stability with Samsung IC).

All par for the course, and simply needs to be kept in mind when pushing the memory IC limits.

Thank You! Stone fell off my heart :). I'm just testing new BiOS/UEFi, maybe it won't need the VDD and VDDQ boost, God knows 🙂

warreng090 wrote:
I have the G.Skill 6000 MHz CL40 on an Apex board. It is tuned to C32 and it hits 60C on memtest, 50C while gaming. Front intake fan hitting it as well as a 60mm Noctua fan mounted right above the ram pulling hot air up.

Nice! Great to know! 🙂 And thank You for the information! 🙂

Falkentyne wrote:
Can you link me to the exact camera that you're using?

No camera buddy, it's a phone. Caterpillar S61. Purchased it for 680e or so, because it was much cheaper than similiar full fledged FLiR device to measure temps.

I'd like to upgrade my S61 to S62, but I've no buyer for S61 😄

In general, You can get FLiR module for Your phone, both Android and iPhone, or a FLiR gun, but they cost a fortune :/. Mine has HW 160 x 120px for heat, such gun cost well over 1200e, there's also 320 x 240 but dayum, them cost a fortune...

https://www.flir.eu/products/e6-xt/

[Falkentyne]

uplink wrote:
Hey there guys!

Thanks to Hopper64 and Shamino, I've managed to run two sticks of G.Skill DDR5 6000 MHz CL36 16GB modules on Z690 Extreme...and now here's my worry.

I know zip about DDR5, I know very little about DDR4 so...

I'm running 2second full run of MemTest on my benchtable and my average temperature of RAM stick is 40°C [+-], most of the time it hovers on 32 - 36°C, but there are ones and zeroes and one more test, that pumps the temps of sticks up to almost 60°C.

I have no errors whatsoever.

My question is, I plan to run two kits, on let's say 5600 MHz, or 5200 MHz, I'll see how they run and if they run together at all. I'd like to keep "vanilla" XMP I. or XMP II. profile with Hopper64's VDD 1.35 and Shamino's VDDQ 1.3575. They keep my RAM steady!

But my temperatures are as follows:



Are 35 - 40°C in most of the tests okay? And 55°C - 60°C in 1s & 0s and hammering in MemTest?

Can you link me to the exact camera that you're using?