D9 Ram - Does it last longer if run naked?

[CryptiK]

As the title says, if you have run D9 modules naked, did they last longer/suffer less degradation than modules on which you retained the heatspreader?

I'm thinking that keeping the heatspreader on may be causing the IC's to run hotter than they would if they were naked. In the case that the heatspreader makes good contact and gets hot (CellShock, some Ballistix) the heatspreader, being heated by all the IC's, would be holding each IC at the heatspreader temp. In the case that the heatspreader makes less good contact, the spreader may be trapping heat and thereby also causing the IC's to run hotter.

Post your experience/ideas on this issue.

If anyone has directly measured (with IR temp gun for example) the temp of naked D9 IC's under load, please post the results.

[Ket]

I tried getting the spreaders off my Cellshocks, but 2\4 screws on each spreader are riveted in or something.. bastards wouldn't budge.

[CryptiK]

Are you using the correct start shaped tool? (I forget the correct name of them) Push down hard when you're trying to undo them.

[Marvin_The_Martian]

Doesn't this depend on airflow? If you don't have active cooling the hs will cause a bigger surface area to dissipate heat, if you have active cooling bad contact could create hot spots on certain ic's and kill your stick through those.

[RPGWiZaRD]

If you use active cooling on them maybe, but I don't know if it's worth voiding the warranty to take them off. I've been using naked Team Xtreem sticks at 2.2v with 2x80mm fans sitting directly ontop of the sticks that I bought of another user that also ran them 24/7 at 2.2v and occasionally higher for benching and these are manufactured in beginning of 2005 so it's had a fairly long life to be D9GMH microns. It wasn't until recently one of the sticks started having problems to boot but if I managed to get it to boot it would work just fine. My Cellshock D9GKX chips I recently ordered went bad in ~2 weeks time though at 2.25~2.3v usage. I know it's a silly comparision, must have been bad samples from start but still... :p:

[gillbot]

I keep a fan on my firestix, they are a few years old and i've abused them but they still work fine at ddr550+.

[CryptiK]

Yeah it does depend on cooling, but I have my CellShocks actively cooled and they still are pretty warm. I'd have to have an annoyingly loud pair of fans over them to keep them cool. My TG's don't really even get more than warm even using 2.3v with no active cooling (just did it once for a test) and that concerned me. I subsequently removed the heatspreader and the IC's were only making part contact and the imprints in the thermal tape were uneven and feint. I think in this case they would be better off naked.

My issue with the CellShocks is that all 16 IC's are heating up the one heatspreader, rather than if they were naked each IC is just dissipating its own heat and could have direct airlfow. I liken it to the difference in quad core/dual core cpu temps - with 4 cores under the same sized heatsink, it gets much hotter. This may not be true with the ram though, and that's why I want someone with an IR temp gun to measure the naked IC temp.

If you use active cooling on them maybe, but I don't know if it's worth voiding the warranty to take them off. I've been using naked Team Xtreem sticks at 2.2v with 2x80mm fans sitting directly ontop of the sticks that I bought of another user that also ran them 24/7 at 2.2v and occasionally higher for benching and these are manufactured in beginning of 2005 so it's had a fairly long life to be D9GMH microns. It wasn't until recently one of the sticks started having problems to boot but if I managed to get it to boot it would work just fine. My Cellshock D9GKX chips I recently ordered went bad in ~2 weeks time though at 2.25~2.3v usage. I know it's a silly comparision, must have been bad samples from start but still... :p:

Right so 3 years or so running naked with 2.2v and actively cooled. And yeah your CellShocks must have been defective to begin with no way they'd die that fast.

[Marvin_The_Martian]

I don't think the total amount of heat from the ic's combined will cause the hs to become hotter then a single ic would on it's own IF the hs has proper airflow. Even if you get someone to measure the temps with an ir gun it will not give a definitive answer to your question I'm afraid, though it's always interesting :up:

[CryptiK]

Yeah, that's what I'm trying to figure out here, I know my ram chips on my video card get pretty hot and they are naked, so D9 may be similar, and a good heatspreader may be helping. I'd just like some numbers if it's possible.

[Ket]

Are you using the correct start shaped tool? (I forget the correct name of them) Push down hard when you're trying to undo them.

Yep and yep. Damn funky screwheads :shakes:

[bartx]

I have a pair Ballistix 2x1GB PC8500 double sided - I have them for maybe a half of year. I'm running with heatspreaders and with good airflow everything is ok (if I don't provide enough airflow, they are running hot even on 2,0V). So heatspreaders have a good contact. I heard that degradation more depends on motherboard/PSU than high voltages. I have DFI DK P35 T2 and Fortron based PSU. I used them in a few almost all day sessions - running on 2,45-2,65V and 1140-1200MHz CAS4 and everything is ok till now (but always with cooling). For 24/7 they are running 1030-1060MHz 5,4,4,12 1,925V (it's enough for me and I don't want give them more voltage). I think that removing heatspreader won't give anything, only risk of broke chips from PCB.

[Bei Fei]

I have a 120mm fan on my sticks and they are run 24/7 with 2.1v @ ~DDR1000 4-4-4-12 no problems. I have another set that stopped working all together and now they work fine. I don't know what the problem was but they work fine now. I run them at 2.1v ~DDR1150 no problems.

[mad_skills]

Last longest with properly placed heatspreader and active cooling.:up: