Welcome to my sixth in a series of CPU block tests I’ll be working on for the new i7-2600K processor, the EK Supreme HF. This is EK’s flagship water-block sporting an acrylic/plexi top, copper bottom, and their new “Easy Mount” system. This particular block is the “Copper Plexi” flavor, but the block also comes in full copper or copper acetal flavors. You can even get the block in blue or green plexi colors if that’s what you want…so plenty of top options to choose from. I was very anxious to get my hands on this block considering the EK Supreme V1 was the thermal winner in my previous Q6600 CPU testing roundup and this newer HF model is the further refined successor to the original.
This sample was sponsored by Eddy from EK waterblocks, thanks!!
Before getting started, how about a picture of the block installed and ready for action!
Packaging & Accessories
I’m no different from anyone else in enjoying a good box opening to see what goodies lie within. I also have had my share of products damaged during shipping over the years, so I consider packaging protection very important. Nothing worse than having to submit an RMA ticket because the UPS man tossed your precious package around more than it could handle.
The block comes in a larger than average box in a sleeve with flashy black/orange/gray graphics. It does give a nice presentation and good first impression.
Upon removing the box sleeve and opening the inner box, you first see the hardware packet, AMD hold down, and user guide.
I read through the user guide and found it to be a good basics guide that should be adequate for your average user. Below you can see a sample of page 1, it comes printed in black and white and a double-sided letter sized paper. I thought the content was good, but not quite as nice as a few color manuals I’ve seen on other blocks. You can also download and print the manual if you want a better copy or loose the original
Digging further into the box, you come to the block nestled and nicely sandwiched between two layers of foam as well as being individually wrapped in a sealed tamper proof plastic bag next to a tube of MX-4 mini tube of compound.
I think this was the first block to incorporate a tamper seal on the block packaging which ensures the block you received has not been tampered with. This is a nice clever touch…I like that. In this world of RMA happy consumers, this is one way to ensure a new product. Very good!!
After all the unwrapping of individual packets, this is the collection of parts you get plus the user manual previously noted:
Short of barbs, you get a very complete package for both AMD and Intel users. The only part I couldn’t make sense of myself were the metal washers in the lower right. These were not explained in the user’s guide, so I went on the assumption that they go below the springs to prevent scratching of the black hold down. They worked well under that assumption
Overall an excellent complete package short of barbs and the packaging protection was also extremely well done.
martinsliquidlab.wordpress.com 
Excellent review once again, thanks martin!
I very much enjoyed reading this review. Keep up the good work 🙂
Martin,
First of all welcome back. I am a grad student working towards my Masters in Mechanical Engineering. I am currently studying water blocks for my thesis. Through the course of my study, we have attempted to use some of your reviews to make some additional observations. However, we have run into a bit of a snag. I was wondering if I you still had some of your original test data? I am also currently looking into building my own test rig for my thesis so any help you can provide would be greatly appreciated.
V/R,
Chris
I have the logged data sheets of these tests. I basically run each mount for about 45 minutes. My crystalfontz generates a time based sheet with a data record (.csv) for all the water & air sensors at one second intervals. I also generate another data sheet using realtemp to log the DTS sensor for all four cores. I then have a template sheet where I copy and paste each of the run sheets into. It uses various look up functions to match up the time from each of the two database sheets into one. I then basically review the logged data over the period of time and trim off the warm up 10 minutes and last minute so I have one constant stream of data points when everything has leveled out. I then average out the whole mess of data points which is several thousand records per mount.
The amount of warm-up loss depends on the radiator and heat load. This is why I’m purposely using a quad radiator for testing. 10 minutes is usually plenty of trimming up front. I’m also purposely using a small 35X reservoir. It bleeds fairly quickly and also not overly large adding to the water volume which would add time to warm up.
The single biggest variable for me has always been the mount and thermal paste contact. I find this depends highly on the block’s bowing of the base and also the processor and IHS shape. My 2600K for example seems to have trouble with some blocks bridging the IHS edges as if there isn’t enough bow or sharp enough bow to ensure strong pressure in the center. This may depend on the backplate support as well..I’m not completely sure there yet either. Many of the top blocks today perform very similar thermally, but it’s that mechanical contact and thin bondline thickness of the TIM material that seems more critical and varies depending on the mount and orientation at times.
Exemplary review, this is the cpu/block combo i have 5k @ 70oC – sweet!!!
necro…. echo….