Maybe some of you will find this interesting. it's kinda cool to be able to get temps by taking a picture.
system:
Asus PC-DL (w/.08v vdimm mod)
2x1.6G LV xeons @ 2.6 (13x200) (1.6v)
intel (3.06) wind tunnel HSFs
2x 512 crucial ECC pc3200
2x 36GB raptors (non-raid)
pny 5900 ultra @ quadro 3000
antec sonata w/extra 120mm fan
antec 550 eps12v PS
most of these were snapped after half an hour of 4 instances of p95. (100% cpu load) I didn't do anything as far as stressing the gfx card, I suppose if I get some more time I can look into that. I had originally planned on using this to evaluate thermal issues at different vcore voltages, maybe a Cm heatpipe or case upgrade etc, but I just don't have the time.
note that the scales are not always the same (so pay attention when trying to compare) also there's one where I forgot to set it to C instead of F. also keep in mind that this thing only reads surface temps, not core temps.
Here's the back of the case:
Back
cpus & VRM
Cpu1
Here's an interesting one...the max temp on the heatsink (inside the square)is 45 deg C...within a couple degrees of MBM temp readouts
Cpu2
shot of the back of my 5900 ultra - again, I didn't stress this, but the detail is interesting. 48 dec C, no load, while the core readings were much lower- idles at ~40 deg C
Gfx
examining the northbridge heatsink. also of note are some hot little mosfets I discovered while doing this. more on those later. the NB heatsink looks like it's doing an ok job (48 deg c)
NB
the whole mobo. again note that note that the hottest place on the whole board is that pair of un-heatsinked mosfets (just above the gfx card)
MB
closeup of those mosfets. note the high temp. 93 deg C...203 F...you could burn yourself on those things!
Mosfs
I never thought ram got all that hot...but apparently if it doesn't get any airflow it does...62 deg C on the stick that's out of reach of the fan.
Ram
with this brief little study...I need a couple mosfet heatsinks...and maybe a new case. I suppose this could also be a great way to accurately get cpu temps, but that would likely involve ripping the HSF off while running, snap a pic, then shut down asap. Not doing that with my hardware....
~dubb
system:
Asus PC-DL (w/.08v vdimm mod)
2x1.6G LV xeons @ 2.6 (13x200) (1.6v)
intel (3.06) wind tunnel HSFs
2x 512 crucial ECC pc3200
2x 36GB raptors (non-raid)
pny 5900 ultra @ quadro 3000
antec sonata w/extra 120mm fan
antec 550 eps12v PS
most of these were snapped after half an hour of 4 instances of p95. (100% cpu load) I didn't do anything as far as stressing the gfx card, I suppose if I get some more time I can look into that. I had originally planned on using this to evaluate thermal issues at different vcore voltages, maybe a Cm heatpipe or case upgrade etc, but I just don't have the time.
note that the scales are not always the same (so pay attention when trying to compare) also there's one where I forgot to set it to C instead of F. also keep in mind that this thing only reads surface temps, not core temps.
Here's the back of the case:
Back
cpus & VRM
Cpu1
Here's an interesting one...the max temp on the heatsink (inside the square)is 45 deg C...within a couple degrees of MBM temp readouts
Cpu2
shot of the back of my 5900 ultra - again, I didn't stress this, but the detail is interesting. 48 dec C, no load, while the core readings were much lower- idles at ~40 deg C
Gfx
examining the northbridge heatsink. also of note are some hot little mosfets I discovered while doing this. more on those later. the NB heatsink looks like it's doing an ok job (48 deg c)
NB
the whole mobo. again note that note that the hottest place on the whole board is that pair of un-heatsinked mosfets (just above the gfx card)
MB
closeup of those mosfets. note the high temp. 93 deg C...203 F...you could burn yourself on those things!
Mosfs
I never thought ram got all that hot...but apparently if it doesn't get any airflow it does...62 deg C on the stick that's out of reach of the fan.
Ram
with this brief little study...I need a couple mosfet heatsinks...and maybe a new case. I suppose this could also be a great way to accurately get cpu temps, but that would likely involve ripping the HSF off while running, snap a pic, then shut down asap. Not doing that with my hardware....
~dubb
