INFO: IFX-14 vs S1283 -- TRUE vs Vendetta 2 - Overclock.net

killa_concept · 09-12-08

The contributors and HSFs tested:
Killa_concept: IFX-14 vs Xigmatek HDT-S1283
Adrienspawn: TRUE vs OCZ Vendetta 2
CudaBoy71: Loaned killa_concept a HS 120x38 Panaflow for testing

Background

A while back, I posted this thread where I attempted to explain a logically backed assumption... the thesis was spawned from my annoyance at people comparing mid-ranged HDT coolers (namely the OCZ Vendetta 2 and Xigmatek HDT-S1283) to higher end models (such as the TRUE and IFX-14)

I based my assumption on these facts:
1) the only changes made w/ the HDT coolers over past coolers is direct contact which equates to faster heat transfer but NOT increased dissipation capacity
2) coolers such as the IFX-14 and TRUE have larger surface area, higher fin density, and larger number of heatpipes
3) Professional reviews comparing the TRUE and Xigmatek do not factor in retention method, thermalpaste application method, high CFM/high pressure fans, and actual heat output (most tests were only done in the 125W range)

I presented the following thesis in the thread:
"Based on these facts, I believe that thermal disipation between HDT coolers and high end coolers such as the TRUE and IFX-14 would be similar at low to mid heat outputs (100-150W)... however, at higher overclocks and higher heat outputs (150W+), the advantages of HDT will become moot and the higher fin density, larger fin area and higher number of heatpipes of the high end TR coolers will shine"

Essentially, I predicted it would go something like this:

Bias?

Obivously, from my prediction and thesis, it might appear that I am biased towards Thermalright products... but that is not the case! I was not defending Thermalright blindly - I defended it based on my above stated logical deductions... One would also hope that a $60 product has some advantages over a $35 one and has EARNED it's name as kill of the cooling hill....Still, people considered me a fan boy and I decided that it was critical to perform an extensive and unbiased test

My Testing Methods:
Evening the Playing Feild

This test was conducted to try to dispell what I believed to be an inaccurate belief - not to create one! Thus, I tried to make this test as fair as possible...

The precautions taken are listed here:

1) The CPU and both coolers were lapped

2) A bolt through kit was used on both coolers... originally, I wanted to use the same kit (tuniq) on both, however it did not fit the IFX-14

3) High CFM/High Pressure fans were used to eliminate potential problems... a FM121 of mine and a high speed 120x38 Panaflow (cortesy of CudaBoy71! Thanks man!)

4) Ambients were monitored throughout the test (the fluctuation did not reach more than 1C)

5) The coolers were remounted several times to ensure reasonable idles (pictures taken after tests)

Xigmatek:

IFX-14:

The Actual Testing (About Time)

During the test, the Q6600 was set to various heat outputs in wattages (you can calculate them using PSU calculator)... The CPU was then stress tested for 5 minutes using Prime95 Small FTTs (long enough to establish the maximum load temperature).... provided below are the wattages and respective settings as well as links to the raw data
75W = 2000MHz w/ 1.2V
IFX-14 -------------- Xigmatek
100W = 2400MHz w/1.29V
IFX-14 -------------- Xigmatek
125W = 2750MHz w/ 1.34V
IFX-14 -------------- Xigmatek
150W = 3000MHz w/ 1.42V
IFX-14 -------------- Xigmatek
175W = 3400MHz w/ 1.44V
IFX-14 -------------- Xigmatek
200W = 3600MHz w/ 1.49V
IFX-14 -------------- Xigmatek
Here are the setups and ambient temps of each heatsink:

Xigmatek:

IFX-14:

Graphing the Outcome and Conclusions

Here's the actual test data in a more user friendly format:

The outcome wasn't quite what I had expected... Based on all the information I had read, the Xigmatek was supposed to be on par with the TRUE at 110Wish... as others know, the IFX-14 is only slightly better than the TRUE so I logically expected the IFX-14 and Xigmatek to perform similarly until 150Wish (which is when I expected the Xigmatek to show signs of hitting it's maximum capacity)

I do admit that the Xigmatek wasn't expertly mounted as is seen in the raw data - there was a 5-8C difference in core temperatures on the Xigmatek as opposed to 3-6C on the IFX-14... HOWEVER, the data was not collected until 6 seperate reseats (while attempting to see decent idles) and the paste did ultimately end up spreading decently as seen in a picture above...

To further omit any claims of bias, I provided the data in various forms:
the average of all 4 cores
the lowest of all 4 cores
and the highest of all 4 cores

Hope my testing helps dispell any misconceptions that others had... Adrienspawn's TRUE vs OCZ Vendetta 2 should show up soon below

Adrienspawn · 09-12-08

TRUE vs VENDETTA 2

Quote:

Info:

- Q6600 G0 revision, VID 1.2875

- Cosmos S case, with grills cut, fans on max, flat on side

- Ambient temp taken via a probe right in the middle of the case.

- TIM is AC MX2.

- Vendetta 2 is bolted down with the TRUE's kit.

- Fan is Ultra Kaze 3000 @ ~2900rpm

VENDETTA 2 RESULTS

Vcore= 1.288V
Ambient= 79.0F

Idle: 32 31 26 27
Load: 43 42 39 38

Vcore= 1.448V
Ambient= 79.6F
Idle: 35 35 31 31
Load: 50 51 48 46

TRUE RESULTS

Vcore= 1.288V
Ambient= 80.6F

Idle: 31 30 26 26
Load: 40 40 36 36

Vcore= 1.448V
Ambient= 80.2F
Idle: 34 34 31 31
Load: 48 48 45 44

Conclusion:

The difference being much smaller between these to coolers than the IFX-14 and HDT-1283, I went with two extremes of wattage, as in between would yield the same difference.

Vcore= 1.288V, 3.2ghz = 129-133W

Vendetta Idle: 32 31 26 27
Vendetta Load: 43 42 39 38

TRUE Idle: 31 30 26 26
TRUE Load: 40 40 36 36

Vcore= 1.448V, 3.6ghz = 181-192W

Vendetta Idle: 35 35 31 31
Vendetta Load: 50 51 48 46

TRUE Idle: 34 34 31 31
TRUE Load: 48 48 45 44

The TRUE was also thinner which allowed me to set up my RAMcooler perpendicularly which I couldn't with the Vendetta 2.

killa_concept · 09-13-08

Bump now that I filled in my portion!

DjQurt · 09-13-08

this outa be good

**Marin** · 09-13-08

Based on the graphs and temps you posted, it proves what I have been saying is right. The Xigmatek coolers do worse under load and higher wattage than Thermalright coolers. The temperature gap between the two coolers becomes wider when the wattage becomes higher, which imo, is not ideal for someone who is OCing.

killa_concept · 09-13-08

Quote:

Originally Posted by Marin

Based on the graphs and temps you posted, it proves what I have been saying is right. The Xigmatek coolers do worse under load and higher wattage than Thermalright coolers. The temperature gap between the two coolers becomes wider when the wattage becomes higher, which imo, is not ideal for someone who is OCing.

Yeah, that was the point I was trying to make originally... I also was going to test the wattage up to 250W to help illustrate the increased slope of the xigmatek.... however, at 200W the Xigmatek hit 75C (2-3C higher than Intel's suggested thermal load) so I halted the test at that

Oh btw, I redid the graph to scale 30C-80C instead of 0-80C to help illustrate the slope

**Marin** · 09-13-08

Quote:

Originally Posted by killa_concept

Yeah, that was the point I was trying to make originally... I also was going to test the wattage up to 250W to help illustrate the increased slope of the xigmatek.... however, at 200W the Xigmatek hit 75C (2-3C higher than Intel's suggested thermal load) so I halted the test at that

Take the risk for OCN

Seriously though, this is a great thread. People constantly ask for proof, well here it is.

mfb412 · 09-13-08

*awaits adrienspawns part*
btw, you couldve included the nh-u12p in this woldnt ya say?

killa_concept · 09-13-08

Quote:

Originally Posted by mfb412

*awaits adrienspawns part*
btw, you couldve included the nh-u12p in this woldnt ya say?

well if I had the cash funds broham, i would include as much as I could.... but sadly, I'm an unemployed 17 y/o finishing off my last year of highschool (hence why I needed to collaborate with Adrienspawn to get everything people wanted)

It just so happened that I had an IFX-14 already and the Xigmatek was $25shipped AMIR when my annoyance reached full capacity

.... Also, Adrienspawn already had a TRUE and happened to have recently purchased the OCZ Vendetta 2... in the world of non-adults, it's just by luck that these sort of reviews happen

logo199 · 09-13-08

bump for your efforts, might comment more thoroughly when I have time to post

09-13-08	#10 (permalink)
logo199 cough Stock cough Join Date: May 2007 Location: Portland, OR Posts: 1,187 Rep: 97 Unique Rep: 87 Trader Rating: 0	bump for your efforts, might comment more thoroughly when I have time to post __________________ avatar by MaKaVeLi They can, unless their processors are designed by ninjas, who only CHOOSE to obey the laws of physics when they want to. Ask a ninja, it's true. -voice Complaining about SLi and 3DMark scores is about as useless here as complaining about the towing capacity of a Ferrari Enzo. -MNiceGuy Bring her flowers, take her out to a fancy dinner, and then pop the question, "Will you buy me a video card?" -RussianGrimmReaper **I realize that buckets of people don't like steam...they view it as communism in the sense that it works well in theory* but not reality... -x7337H***

		Overclock.net - Overclocking.net > Cooling > Air Cooling
INFO: IFX-14 vs S1283 -- TRUE vs Vendetta 2

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)