Ken429 Posted December 21, 2015 Share Posted December 21, 2015 I have two "identical" 4790K systems running at 4.5GHz. System 1 uses a custom water cooling system (Swiftech Apogee XL water block and a 2X140 radiator). System 2 uses and old H100 AIO that I had laying around. Both systems use PWM fans controlled by the MB BIOS. Both systems are running in the same location. System 1 (Custom Cooling): Max Vcore under load (IBT 10 passes) 1.272V Max Core Temp 76C. System 2 (H100): Max Vcore under load (IBT 10 passes) 1.264V Max Core Temp 85C. I have seen reviews that say the H100i is much better than the H100. Since I'm getting lazy in my old age and don't want to fool around with another Custom Cooling system - will the H100i get me close to the performance of the Custom Cooling setup? Which version of the H100i is the best choice - I'll probably keep the PWM fan setup rather that fool around with the H100? fan controller. That's based on the H100 controller! Link to comment Share on other sites More sharing options...
snapper69 Posted December 21, 2015 Share Posted December 21, 2015 It would be interesting to see the readings if you swapped the cpus between systems. Not all cpus are created equal, in fact, not all cores in the same cpu are created equal. There's 8 degrees difference between the hottest and coolest cores in my system. Have you considered the H110i GT, bigger rad, better cooling and quieter. I'd be surprised if either of your cases wouldn't accomodate it. Link to comment Share on other sites More sharing options...
Ken429 Posted December 21, 2015 Author Share Posted December 21, 2015 All the cores are hotter under load on System 2. This ain't my first OC. The Custom Cooling system is in a 650D case so 2X140 works. System 2 is in a 600T case so I'm stuck with a 2X120 Radiator (GTX)? Back to original question - is the H100i GTX truly that much better ~5C than the old H100? Link to comment Share on other sites More sharing options...
snapper69 Posted December 23, 2015 Share Posted December 23, 2015 No offense meant, I guess that the fact that you have a custom cooler on rig 1 should have told me that you already know a good bit about the subject. I understand that the H100i was a fair improvement on the H100 (that's why I bought one originally), but how much of an improvement the GTX is, I don't know. I recently changed my case and the new one wouldn't accomodate the H110i GT that I was using, after the problems that I'd had with the H100i (3 of them) being bricked by Link firmware updates, I chose a 240 mm cooler from another manufacterer. That cooler in the new case dropped the temp by 5 degrees, but I don't believe that the cooler is that much better than the H110i GT, so it must be to do with the case and the airflow as well. Link to comment Share on other sites More sharing options...
c-attack Posted December 23, 2015 Share Posted December 23, 2015 It's possible you may get your 5C. It's hard to find test results with similar hardware and I don't use IBT, so it's difficult for me to translate the temperatures. However, there is no doubt the H100i was an improvement on the older H100 (comparison). The H110 I own has always been consistently 4-5C better than an H100i with the same hardware. The link has a w/Priime test for the H110 and 100i GTX. They come in even at 1.30v on a 4790k. So, bootstrapping your way through that should yield a minimum 5-6C. In the other reasons to change category, you can loose effectiveness over time, usually due to flow reduction. The newer GTX apparently had it's fan controller beefed up and stronger hose attachment points. Those were weak points on the prior models. Link to comment Share on other sites More sharing options...
Ken429 Posted December 23, 2015 Author Share Posted December 23, 2015 I went and blew another $100 dollars on a H100i GTX and am very impressed with the look and feel of the cooler verses the original H100 I purchased several years ago. However, after 5 minutes of playing I haven't figured out how to make the Link software do what I want - the Radiator fans remained quiet even on Performance Mode? So I hooked the fans up to the MB controller and ran IBT again. That is the way I had the H100 setup. This time the fans ran full out under load BUT the CPU core temps were still too high compared to the other system with the custom cooling setup. In fact the Max core temp was the same as it was with the old H100 85C! With the fans hooked up to the H100i the Max Core temp reached 88C but the fans were whisper quiet during the whole IBT 10 passes. I have the H100i connected to the Case Fan 3 and it is set to Full in the BIOS but then I really don't know what the Pump is running at since I'm not sure what the H100i is feeding the MB for RPM's using HWINFO64 to monitor. Someone somewhere said to hook the H100i directly to the power supply via a Molex or SATA connector with a three pin fan connector to eliminate the MB variable. I will give it a try. At least at this point, I'm not all that impressed with the H100i vs H100 vs Custom Cooling setup. On the other hand, I may have a 4790K cpu that runs 4.5GHz @ 1,264V but runs hotter that @#$% under load. Link to comment Share on other sites More sharing options...
c-attack Posted December 23, 2015 Share Posted December 23, 2015 The LINK system will run the fans in accordance with water temperature, since the primary purpose is to help remove heat as it flows through the radiator. The test duration may not have been long enough to heat up the water and cause an increase in fan speed. Of course, you don't have to run link and can run from the motherboard as you are doing now. Either way, I am a little surprised you didn't see better results. However, we all know where the limitation is on a Haswell chip. I suspect your custom loop has a larger volume of water in the system, a faster flow rate, or both. It certainly would appear fan speed isn't very relevant. I don't use IBT so I can't really comment on the temps, but a quick 4790k & IBT search will bring back a lot of results worse than yours. You might try another test that runs a different way. I suppose the real question is whether it offers better cooling or lower noise levels for your actual day to day uses compared to the H100. Link to comment Share on other sites More sharing options...
Ken429 Posted December 23, 2015 Author Share Posted December 23, 2015 I needed a readjustment in my thinking! The water temp shown in the Link display changes very slowly and the radiator fans increase speed slightly as it rises. You are correct IBT 10 passes completes in too short of a time for the H100i GTX to detect a change in the water temp. I can run AIDA64 Stress Test forever and the CPU Temps stay below 75C. I guess that's why my Swiftech custom cooling setup cost $400+ versus $100 for the H100i GTX. The Swiftech stuff is setup to react to the CPU temp changes and it has more liquid, pump volume and water block cooling than the H100i GTX. But...I thought the H100i GTX would perform better than the old H100. So much for Internet reviews! As an aside, does anybody know where to get a short UBS cable for the H100i that will connect to a standard USB port. The ASRock Z97 Extreme6 MB I'm using has a separate USB port on the MB. If I could use it for the H100i it would free up both of the USB 2.0 onboard connectors for my four 600T case ports. Link to comment Share on other sites More sharing options...
c-attack Posted December 24, 2015 Share Posted December 24, 2015 You can try this. They come in various lengths and some with right angle mini-connectors, if that helps. Amazon has dozens, but you may still be able to find some at your local Big Box. Link to comment Share on other sites More sharing options...
Ken429 Posted December 24, 2015 Author Share Posted December 24, 2015 Thanks for the link. I looked earlier and only found ones with a straight connector. I'll give it a try. Link to comment Share on other sites More sharing options...
Datakiller24 Posted December 31, 2015 Share Posted December 31, 2015 Hi, Just one question to these H100i GTX testing. If You are not running the CL SW (not setting things up) than my question is if You run the pump of the cooler in the "silent mode or performance mode"? :) When I installed this cooler first time ... and checked how it is setup default than CL gave me these base setup: - Fans: balanced mode - Pump: silent mode I have as first shot set the pump to performance mode and created some custom fan profiles (lower rpm, i like silent operation when idle) and set up bigger steps to raising temperatures of the cooler (example 32°C - 30% fan speed, 33°C - 75% ...) so it never allows for the water to heat up to much. The pump speed change lowered my temps after changing around 5°C in idle and 8°C under full load). My cooler is powered also through an akasa Molex to 4Pin PWM splitter to gave the cooler direct 12V from my PSU. Runs perfect. Example: http://www.akasa.com.tw/update.php?tpl=product/product.detail.tpl&no=181&type=Cables&type_sub=Fan%20Cable%20Adapters&model=AK-CBFA03-45 BR, Peter Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.