[Build Log] The Darkness - A Fast Build For Work & Play - Page 3

Yes, I did... I really do like them better (especially the fact that the VRM bank is actively cooled - as opposed to passively on these Razor blocks). However, at the time I wasn't able to find anyone that had them in stock and the cheapest they could be found was right at $155 ea. I paid $176 for BOTH of the XSPC blocks. I'm figuring that I'll be updating the hardware in a year or so anyway and wanted to allocate the 'extra' $150 towards other upgrades.


Thanks and welcome! It should look even better in a month or so after I spend some time on the OC/stability - then it's time for bling... well if you can call sleeving the tubing and adding lights 'bling'... not likely to go crazy.


Thanks... it's definitely not the fanciest around here but it's the nicest one I've ever owned... yet.

Thank you. This is phase one... I'll be working on the aesthetic treatments next and making some adjustments after awhile so stay tuned! For the time being I just want to get to finally use it after a month!

Thanks... can't wait to see what you do with it. Your TJ11 build is a really great one!

Sure, I'll be happy to once I actually get to USE it... I'm not looking for maximum overclock on this system (wanted silence over speed mostly) but I certainly hope to have at least 1GHz on the cores and a small bump on the RAM. I was able to get that much with nominal stability (good for 36-48hrs between crashes bitcoin mining) on the OEM shrouds without any voltage bump and full stability at even higher rates with a voltage bump of 10% - but the heat was scary (~89C) so I didn't run that way for any appreciable time other than a bench run.

The one thing I can say for the blocks is that they flow pretty damn well - that was apparent during my leak-testing period... granted the MCP35X is a powerful pump, but until I put the little anti-cyclone fitting on the intake it was hitting the top of the res with force! From the tests I was able to find on this site and a couple of others it seems that it's in the lower third of comparables in the cooling department (mostly due to passive cooling on the mosfets I believe) - but in the upper third as far as flow rate is concerned.

I will post much more detailed results on my system performance after I get a system on it and run some benches - so I'll let ya know soon!

Some updates... more to follow...

CPU-Z Validation
GPU-Z Validation


As far as the XSPC Razor blocks for the 7970s - I can't find any real flaws. Although it's definitely possible I could hit a higher OC on the cards with a better pair of blocks... it's unlikely that I would be able to hit more than 10-15% higher IMO. Considering the blocks run about 35% less on average than competitive blocks (depending on promotions, etc.) I think they are a great buy. If they had cost more than $90/ea I might have a different opinion, but for what they are and how they look/work... I'm very, very satisfied.

As far as the GPUs themselves... well, although I can't say I'm impressed with the headroom - I wasn't expecting that much and did buy the cheapest reference 7970s on the market (~$395/ea) at the time. Both cards appear to be stable (they're bench-stable and I'm getting ready to mine coins for a 48-hour run to verify) at 1025Mhz/1400Mhz core/mem with a 15% bump in power. The fit between the VRMs and the block is so tight that I can't really fit a thermal probe in there ATM so I can't verify temps but the card appears stable so they can't be out of control.

On the CPU front... well things there are pretty impressive as far as my expectations are concerned. I knew the chip I had wasn't too bad, but considering I'm stable (not 24hrs primed - so just 'usably stable' ATM) at 4.7GHz with only a bump to 1.4V at the core... I'm very confident in this chip being able to hit and pass 5GHz without any significant problems. However, that will require de-lidding and I'm not only hesitant to do it... but I want to actually get to use my system for awhile before I mess with tearing it down and rebuilding it again. I can tell de-lidding would benefit this chip significantly however - considering the difference in temperatures - the difference between Core0 and Core2 is 13C under load!!!

Coupled with the fact that even with the obvious issues with the TIM eveness (under the IHS that is) I'm still barely passing the 80C mark on the hottest core under load (1hr OCCT) - I'm sure after de-lidding I can hit 5GHz at roughly the same temps.

I call this build a success, but I have to do a little more stress-testing before I'm fully confident in my OC stability. Will update further as things progress... next up - appearance and DISPLAY!
Edited by DiGiCiDAL - 11/8/12 at 11:33am