The Battleship Justy: 2008

Saturday, April 5, 2008

Setting up Hard Drives

I've decided to go with 10 Mtron 7000 Pro's at 64GB each in RAID 0 for my Windows / AutoTrader drive. 6 Mtrons have already arrived, and I've been testing out 5 of them (1 was DOA). Even with only 5 of the 10 drives in, the Battleship Justy scores a 13,584 in PCMark Vantage, which would make it the 4th fastest computer ever tested.

I'm going to have 4 more drives, with a total 3.7 TB capacity installed for storage drives, these will be installed in the next week as their enclosure arrives.

Wednesday, April 2, 2008

Stepping down for stability

I've been trying to get the system as stable as possible today, which means bringing down the Vcore to reasonable levels, and with it, the clock speed. Currently I'm testing 4.4GHz in Vista and it seems to be running stable, even under heavy stress.

Tuesday, April 1, 2008

5GHz in Vista

After upping the CPU voltage to what is probably a near-suicidal 1.75V, I'm posting this from Windows, running at 5GHz. The system doesn't seem to be crashing right away, but SuperPi refuses to run at all, and, trying to take a screenshot, I experience a strange graphics error:

Edit: The strange graphics appear to be a gfx card error, as I'm having the problem even running everything at stock. Upping the voltages a bit more (pushing the 70º C cutoff) got Windows slightly more stable at 5GHz, enough to finish a 1M Superpi run, with a time of 9.454 s.

Monday, March 31, 2008

4.4 GHz, SuperPi = 10.7s (1M)

Just playing around with the BIOS multiplier, we've been able to get to 4.8 GHz in BIOS, and 4.4 GHz booted in Vista. Superpi time to 1M for the latter is 10.7ish. Will post details and pics as I get it up higher: the temperature is still relatively chilly at these clock speeds, so I think this could go considerably higher.

It's getting pretty chilly around here

The phase change cooling is hooked up, and as you can see, it's working beautifully. Running at -53C on stock settings at the time of writing this post (although the more I type the further the temperature falls -- it started at -48C when I began writing this, so it may end up colder than -53C eventually).

Also, to anyone else who tries to run this Vapochill phase change cooling system: do not try to turn it on with your case on it's side. Stand your case up vertically first. We tried with the case on it's side at first and I've never heard a computer make a worse noise, plus the whole case shook (if you did this with normal hdd's spinning it really could just horribly break them too due to the crazy huge vibrations) -- it literally sounded like a lawn mower (perhaps half as loud though), and we thought we had broken everything. We turned the case vertically and things seem okay (no more awful noise/vibrations at all, runs fairly cool actually), but I strongly advise nobody else attempt to try it.

And things start to look up

So, it's looking like we had a bad Asus Striker II Extreme motherboard sent to us from NewEgg, which has set us back about 24 hours in this build. After trying about every other option (different PSU, different CPU, different RAM, different GPU, removing the phase change cooling system, running the whole thing with only the CPU on an antistatic surface to ensure there were no shorts, etc.), we decided it just had to be the motherboard, went to a local computer store this morning (thanks MicroCenter!) and got an E-VGA NVIDIA nForce 970i Ultra SLI motherboard. So far things are looking like they're functioning -- tests are now passing that were previously resulting in completely and total failure. We haven't gotten everything installed yet, but just to give an idea of where we're at, and the mess we've been dealing with, I'm posting some pictures.

Try these on for size:

And we're just now attaching the phase change cooling system...

Can't boot

I've been struggling with a strange problem with the motherboard (Striker II Extreme) for the past 12 hours or so. Taking the motherboard by itself, plugging in just a processor (I've tried 2 different ones), a single stick (or no sticks) of RAM, and connecting the 20+4 pin and 8 pin motherboard power connectors, the system won't turn on at all (using the on button on the board, of course). I've tried 2 different power supplies now, so I'm starting to think the problem is just the board itself. The strangest part is that, connecting the 20+4 power connector and not the 8 pin one, the system appears to power on, but then won't POST at all.

Sunday, March 30, 2008

Few building pictures

Here's what it looks like now, with the phase change cooler tube secured to the processor.

Earlier picture

Saturday, March 29, 2008

Pics of components

Here are the components that I've received so far. I'm starting to attempt to install the necessary heating/cooling components on the motherboard, which could take me a little while.

Thursday, March 27, 2008