┻━┻︵ \(°□°)/ ︵ ┻━┻ NEW YEARS BREAK ┻━┻︵ \(°□°)/ ︵ ┻━┻
Log of Jan 01, 2014: First Work inside the Case, Dual GPU Priming, CPU Block Installation
welcome back and happy new year
Turns out I needed a break for two days and thus didn't do anything on Dec 30/31. But as I'll resume work/study coming monday the 6th, I have to kick it up a notch and get this thing finished asap. Therefore, I've done quite a bit of work today and luckily progressed well. Firstly, I installed the single Corsair AF140 quiet edition fan as rear exhaust. For a second I thought about why I'd ordered the quiet ed. instead of the high performance ed. but then realized that Corsair decided not to release any AF140 fans as high perf editions.
After that, it was time to unpack the RIVBE. What a beauty! I've never in my life have had such an awesome looking motherboard! I'm so glad that I could get such a neutral looking performance beast from the ROG series, as I wanted my build to have a black / blue look from the start of its planning stage. Wouldn't have looked half as nice with a prior rampage board, as they all have a black & red color scheme.
The RIVBE was also my last point of concern that I might not be able to build this thing over the course of the x-mas holiday. That's the way I had planned it for months, but I didn't pull the trigger on all the orders until the first week of December. And since I live in a tiny country that doesn't have a single RL or internetshop to offer watercooling supplies, I had to vastly resort to resellers / manufacturers from the UK, Germany, Slovenia and even overseas from the US and Hong Kong. And so, the last week before the start of the holidays, everything had arrived, apart from the RIVBE, which was 'out of stock' at every single feasible retailer available to me, even though people from the US had received theirs around Nov 20 already. That really pissed me off. And so, when I returned home on friday, I was sure I wouldn't be able to get started, as it still had not yet arrived and the supplier couldn't tell me when it would, after I'd inquired accordingly a week before.
But, in perfect christmas-y spirit, I was surprised with a package the very next day. Apparently, my supplier got a tiny amount of those in (prob around 5 pieces) and as I had pre-ordered and paid almost 3 weeks before, I was one of the few lucky sods to get one.
But enough of that! I installed the 4820k CPU along with the EK supremacy. Didn't realize until that point, that with the 2011 socket, manual backplate installation is obsolete, as this socket features a pre-installed one! Therefor, the block installation was extremely easy. I only had to replace jet plate, put the gelid extreme thermal paste on, and then handscrew the block onto the cpu. That's the first time I've installed a cpu cooler entirely by hand with no tools required for the job; very nice.
Maybe I should mention that I still did a bit of work on Sylvester night: Since my girlfriend thinks of it as a stupid and unnecessary event and - like last year - decided to go to bed early, I had to party on my own and did so by installing the second GPU waterblock and drinking beer
Afterwards, I hooked up both gpu's to the triple parallel bridge and then to the priming loop for a bit of daring leak testing, as I haven't even flushed the blocks / bridge prior to installation. http://www.youtube.com/watch?v=dVSjCvGFMnM
So, below you'll find pics of the 900D, the RIVBE, the crossfire setup and the EK supremacy installed. Enjoy lads
┻━┻︵ \(°□°)/ ︵ ┻━┻
Log of Jan 02, 2014: Motherboard, GPUs, Tube Res and Radiator Installation
Good day, dear sir or madam!
I continued my working frenzy and installed everything inside the case, save for most of the cabling. Starting off with the mobo installation, I then socket the gpu's with the bridge already on them. Lazy? yes, maybe, but I did just unhook it from the testing loop and there was still some water inside of the blocks and I didn't want any on the RIVBE, so I just installed the whole bundle and it worked without any problems, just had to carefully align both cards and voilá! I'm still somewhat worried that I might have screwed up somehow during the waterblock installation, as I've never done that before and - as I stated once already - I haven't even tested the cards with the stock fan still on them, so if they were faulty to start off with, I'm in trouble as I'm quite sure ASUS won't be taking them back: There was one tiny sticker on one of the screws on the backside of the gpu that I had to destroy when taking off the stock cooler / fan / plastic cover. Let's hope I did well
It was then time to install the fittings onto all the blocks and radiators. Was fairly easy with the EK compression fittings but my fingers did hurt a lot after twisting and turning and then untwisting due to noob mistakes and tiredness, then re-twisting... At the end of the day, I was happy once all the fittings were installed and I could rest my hand.
Next up I installed the Photon 270 D5 pump/res combo. As I had to find out, I was not in the possession of any metal drill bit, only giant ones for poking holes into concrete walls and thus could not drill the four holes required to properly install the tube on its metal bar. So my gf suggested I'd just use zip ties to hold it in place. Thinking of it as a silly idea at first, I quickly had to realize that I didn't have any other feasible options at the moment - I just didn't want to waste any more time going to the hardware store. And so I used zip ties to hold it in place
There's a chance I'll redo that bit at some point in the future, but if its staying steady in place, I'll likely just leave it that way. After all, it's not superheavy, even when fully filled up it shouldn't be putting too much stress on the zip ties. But yes, you may call me a total amateur for doing this.
Afterwards, I installed the fans on both rads and was sorta mad at their relatively short cables, even though they are probably standard size. But inside the 900D, everything has to be slightly longer I'm afraid. I'll be hooking up all 8 fans of each rad to an 8-way pwm splitter, which itself - sadly - only has a very short cable to connect to the mobo header. We'll see how that will play out when the cabling is due.
I then proceeded to install the rads. No easy feat, especially for the top rad, as I - once again - was working throughout the night and couldn't get my sleepy gf to help me hold the thing in place while putting in all the screws. And as if that wasn't enough, I realized that there weren't enough 30mm screws included with either rad. They included 16x 30mm and 16x 35mm screws with each rad; the 35mm ones are only to be used with a shroud in combination with the fan. So, I had to resort to using a silly amount of washers to make up for the 5mm extra that I didn't need for either side of the rad but had to use on one side, as I didn't have any other 30mm screws handy and once again, didn't have the time to get any at that moment. Noob mistake, but doesn't really matter in the end. Still was rather annoying to lay out and count all the available washers of various sizes I had, to calculate how many I could use on each screw, when it was already dawning after a long night's work...
Once I succeeded with the rads installation, I quickly added the drain valve to the extra port of the bottom 480 monsta rad which points towards the front of the case but it works as I had to remove the lowest of the three front intake fans anyway to fit the rad in there. Draining the system will be very easy this way, as the 900D allows for quick and easy removal of the frontal cover and shroud to reveal the valve.
Finally, everything that needed to be inside the case for the 24h leak test was ready to go! At that point, it was 9am already, my gf was up and shaking her head when she saw me still working.
As it was the next day already, I'll split the log up at this point and smother you with more images of today's progress. Once again, have fun
┻━┻︵ \(°□°)/ ︵ ┻━┻
Log of Jan 03, 2014: Tubing, 40h Leak Test
Pulling an all-nighter of hard work consumed what was left of my energy, but I refused to give in to my sleepiness and decided that I wanted to get the tubing done 'real quick' so that I could start the leak test prior to going the bed, so that I wouldn't waste any more precious time to finish this build. And so I did! After another hour or so, the tubing was done and everything was looking fine to me. So I filled the loop by pouring the distilled water into the monsoon bay res. Turned out that the slightly elevated position it was in towards the tube res didn't suffice to fill it up as well. So, I had to unzip it, awkwardly stick it out towards the window side and fill it up there at the same time; dear gf helped at this point.
Started to cycle the pump until the flow indicator starting turning with good speed to indicate that most big bubbles were bled and that my time for bed had finally come. Even though I was super tired, I still managed to take a few pictures for you guys before hitting the haystack; mostly 'napkinshots' to take in, haha. BTW, the green light is once more emitting from my gf's pc which stood behind the build, as I used her PSU to power the pump. Enjoy!
┻━┻︵ \(°□°)/ ︵ ┻━┻
UPDATE of Jan 17, 2014: It's done
As you might be able to tell, midway thru the build log I've stopped writing any more updates. This was mostly due to my frustration that everything was very time consuming and the realization that I'd have to spur on in order to finish this thing before the end of my holiday. So while still taking a lot of pictures in between at pretty much every step, I was unwilling to write / update the build log each day still and instead focused on the work at hand. And so by now, this thing has been standing on my desk for almost two weeks and been in use too, of course. So next up I'll try to cover what I did after the last update up until now:
After filling the loop up and making sure that nothing was leaking (the only leaking thing was the drain valve, that I hadn't properly screwed on, because its G1/4 thread was very long and I was afraid of damaging the rad and thus didn't screw it in all the way - turns out I had to do it anyway though...) I went to bed and didn't get up until the early evening of the 3rd. I proceeded to slightly rock the case to get the remaining bubbles out; precious girly prior kept watch over the first couple of hours while I was sleeping and also cycled the pump for me a couple of times, according to the instructions I gave her
When the 24 hours had passed, everything was bubbles-free and looking good. But still not having fully replenished my own batteries, I left it running for another night until late the next evening. Once 40 hours had passed, I removed the napkins and installed all the cables, which again took longer than I thought it would. And so it was the middle of the night again, when I was finally completely done with it and ready to power it up for the first time ever!
AAAAAAAAAAAaaand it worked! The RIVBE cycled thru all the POST steps successfully and stated '34' on the built-in display, meaning that the UEFI was successful in initializing all my hardware but failed to boot as it obviously lacked an OS at this point. Hooray!
Next up I wanted to get it to my room, hook it up, install the OS, benchmark the beauty... but I couldn't lift the damn thing. It was seriously too heavy for me to lift by myself. Maybe it was because I was tired again, but in any case, I had to leave it where it was on the living room table and go to bed.
The next day, before moving it, I decided to install the OS where it stood. At this point, I failed to create the RAID0 on my two SSDs that I wanted. After much googling, I figured out the problem: The RIVBE has an Intel Chip for two SATA 6G and two SATA 3G headers, which supported RAID. Along with the Intel Chip, it also has an ASMedia controller for six further SATA 6G headers, which do NOT support any RAID configurations. And of course, I had hooked up one SSD to an Intel header and one on an ASMedia header. Grrrrrr!
Resolving that, I stumpled upon the next issue: The RIVBE Manual stated you'd need a RAID driver, that had to be copied unto a floppy disk (?!?!!). Not really believing my eyes, I tried to run the windows installation without it and indeed it did fail, as before I got to the partitioning part, it asked me to insert some driver disk. Long story short, after trying various RAID drivers from Intel themselves, some modded ones and then some others... it turned out the problem wasn't the Raid driver, but rather the USB3.0 Driver, as I had plugged the win7 install stick into a 3.0 Port. After plugging it into a 2.0 Port, the message vanished and I could install the damn OS.
Once the OS was running, my significant other helped me heave the thing to my room and so, on Sunday the 5th, the very last day of my vacation, I could finally start using my creation. Initial Benchmarks were looking good, temps were looking fine as well. Started some initial OCing, the CPU from 3.7 to 4.5 Ghz, no problem. It's even undervolted (1.340v) at this overclock! Bumped the GPU's up from 947mhz gpu clock to 1145mhz without raising the voltage and they too peform without any hiccups. Really the only problem I'm facing atm is with the Catalyst Driver 13.11 - Assassin's Creed Black Flag crashes every now and then after mere minutes of gameplay. I didn't know it was the driver's fault until I once got a message bubble from windows stating the the catalyst driver had crashed after the game froze. Just subscribed to the amd catalyst newsletter, really hope the next version will resolve this issue...
So yeah, that's pretty much it. The Build is complete, working fine and in use! Performance is great, any game I threw at it so far (Aion, ACBF, BF4, GW2) ran smoothly at around 60FPS on average on highest graphics settings. The temps are below 5° Delta to ambient temp at idle and hardly go over 10° Delta on the CPU and 15° on the GPUs. Noise is great too, at Idle the fans run at 800-900 rpm and rise up to 1100-1200 under load. However, when I manually crank them up to 2200 rpm (max), the thing sounds like a chopper about to take off
All in all, I'm very happy with it.What's left then?
As you might have noticed, I have not yet uploaded any pictures of the final result. That's because I haven't made any so far. Why? One PWM splitter, to which all the fans of the lower 480 rad are connected, is too short to run along the backside and thus awkwardly hangs in the air, going across everything and looking plain ugly. But fear not, just yesterday I ordered an extension cord, which should arrive sometime next week. After its installation, everything will be looking pretty and I will be making plenty of pics for y'all to enjoy. Along with those, I'll also post some benchmark results and with that final update, conclude this build log. You may look forward to it around Jan 25th!
I won't leave you without a tiny sneak peek though
thanks for reading and following up on my build!
Edited by Valice - 1/17/14 at 5:10am