Madshrimps OC Team
Around 10 o'clock, Geoffrey woke up so it was time to get the rigs fired up. We started testing the DFI P45 in combination with my E8600, but rather quickly we ran into problems. Apparently, there was no way to run near 600FSB, which was a little disappointing to say the least. Luckily, I had brought my old K7 platform and, to be honest, I was quite excited to finally test this combo under LN2. The DFI Lanparty NF2 Rev.B was one of the best socket 462 motherboards around and in test sessions at home I already hit 270FSB, which is quite good for an nForce2 platform. Geoffrey had brought his soldering gear (and skills) and modified the board to maximize the potential of this motherboard.
Geoffrey's voltmodding the DFI Lanparty Nforce2 Rev.B for an old-skool AXP bench sessionGeoffrey: Having screwed up multiple Abit NF2 boards myself I was really afraid to first power on the board, this time however the Lanparty logo appeared and brought with it a good feeling, for first time I was going to witness an extreme cooled Atlon XP setup, my favorite too. We worked our way past the 3GHz barrier pretty fast and found ourselves in the hwbot top 10 without too many problems. Reboot after reboot we worked our way up to the top of the list with only one goal, be the first to hit a sub30s time in SuperPi 1M calculations.
Massman doing his LN2 magicNext was our goal to hit the highest overclock percentage wise. Our Pentium 4 S478 was sure able to hit over 100% in theory, however I gathered most hardware way too late and I didn't even had the change to quickly snoop around in the BIOS before the overclocking event took place. The fact that it may or may not have clocked like we expected did not matter at the end of the day because the setup never ever really booted successfully. Luckily we had another victim ready for our hwbot scoring rage, the 8800GTS used for the Madshrimps G80 overclocking article is sure ready to take home some well earned medals as well! Another disappointment unfortunately, the mounting gear was missing and we had to go with a crumble backup solution. At first boot we managed a single Aquamark run but it was rather quickly we noticed that the contact between cooler and videocard was just awful
DFI P45 and a 8800GTS 320MB ready for LN2 action
That was just the start though as there was much more to happen the weekend...Pictures of Saturday’s bench sessions
The competition aspect of the event, which is mandatory in overclocking, was well-balanced. On the one hand you had the so-called free competitions in which everyone could enter with any sort of hardware and on the other hand you had the forum competition, in which the different teams attending the event had to compete against each other in a series of tests.
The free competition consisted of a CPU overclocking contest, in which people had to obtain the highest overclock as possible. To even the playing field, the jury had chosen for a relative overclocking contest, which basically comes down to percentile overclocking: gaining as much mhz over the stock frequency. Many people brought along their E21x0 processor, which are known for their high OC potential (although many suffer from an FSB wall), and it can't come to a surprise that the winning setup was an E21x0 setup: 160% overclock!
Next to the CPU overclocking, there was also a GPU/system overclocking contest: Aquamark3. The rules in this competition were quite simple: whoever has the highest Aquamark3 result in its category (air, water or extreme) wins. A little surprising that Aquamark is the benchmark that has been run the most during the event, no?
Apart from the free competition, we also had the forum competition, which is to me the best part of the entire event. Not because of the concept, as many LAN parties already offer this kind of competition, but because of the test procedures the different teams had to go through. Firstly, all teams had to compete in an overclocking / tweaking contest in which everyone had the exact same C2D based system at their disposal. Every team had 45 minutes to achieve the best score in 4 different benchmarks (Superpi 1M and 8M, 3Dmark01 and Aquamark3), but the maximum cpu frequency was limited to 4000,9MHz. Everyone tried to run 500x8, of course, but most of us had to be satisfied with 470x8,5 as the Asus motherboard didn't allow 500FSB. Next to the overclocking part, we also had to throw (yes, throw!!) a 20kg case as far as possible. Gianni-GT, one of the Swiss overclockers, made a huge amount of slow-motion videos of the case-throwing, so have a look and enjoy. The last part of the competition consisted of a keyboard puzzle. We all had to re-assemble a German keyboard within 4 minutes ... which is very little time I can tell you, especially when the Belgian keyboard lay-out is completely different from the German lay-out!G: As addition to this event (or is it maybe the other way around?) there were many gamers represented for the ordinary Lanparty like Unreal Tournament fragging competitions. The combination of the two is cool extra, gamers among overclockers questioning why they're running the same sequences on and on again, I'm sure some gaps have been narrowed here. As do-it-myself modifier and overclocker I can tell you I have seen a thing or two, but you never saw it all at the same time only a square feet away from each other, have a look at some of the coolest overclocking tools we found at this event:
An air conditioning bench box; when the phase change is turned on, the air temperature in the box is around -12°C, helps to prevent condensation
Next to Madshrimps, there was another person trying to set milestones on the AXP platform
Inside of a benchbox
People printed out HWBot results to know what they are aiming for when benching
Monteboy proudly posing for a photo. He's the creator of the Superpi tweaking tool which helps many people perform an ideal Copy-Waza
Mushkin's idea of an overclocker's PC. Didn't overclock that well ... we tried
Watercooling setup and LN2 ready, that's the spirit
Air cooling, the extreme way
LN2 memory cooling ftw?
At the end of the day, when everyone was exhausted from overclocking all day long, Giorgioprimo and Leghorn were not yet tired and fired up their 3-Way SLI system, using 4(!) LN2 containers
A little credit for Jowi69 is at its place here as he obtained the 32M world record in front of a live audience at the event!
Jowi69, he took the SuperPi 32M world record live on AOCM!!
New 32M World record (clickable)