What are the ingredients for the perfect party? The answer from most normal people would usually be plenty of beer and some good friends, but when Barry, or bazx as he's known online, is planning a gathering, he likes to let something a little more interesting than beer steal the show. Barry is one of the lucky few that can get his hands on liquid nitrogen, or LN2 as many people call it, and on 28 December, 2005 he invited seven people round to his place for an extreme overclocking party.
When I arrived, the group was huddled around an Abit AI7 motherboard using a lowly Celeron as a guinea pig for the LN2. Strapped on top of this was a copper flask into which they were pouring LN2 from a saucepan. It looked like something out of a classic sci-fi film, with vapour rising out of the flask and evaporating at an alarming rate.
However, the Celeron eventually stopped booting, which was probably because the motherboard's capacitors near the processor were too cold to function properly (the frosty coating was a bit of a giveaway). The board wasn't dead, but it certainly needed to warm up before it would POST again, which unfortunately meant drying it out as well, as the ice would melt.
Barry's facilities were everything you'd expect from a true enthusiast. Scattered around the room were several phase-change coolers, including a cascade unit from America that he was working on. The most interesting part was the desk he had built for his overclocking experiments. Being an engineer, it was much more than just a desk. The top had a monitor on each side and a Shuttle XPC solely for monitoring temperatures. Then, under that was a fileserver just for storing data, and in the centre of the desk was a cabinet that was a little larger than an ATX case. This was connected to an air conditioning unit and a dehumidifier under the desk, which helped to keep the computer at sub-zero temperatures with little or no moisture. Also, on top of the cabinet was a sealed and airtight hole where the evaporator head from a highly modified VapoChill unit entered, and this doubled up as a hole for the LN2 Flask.
Barry's Athlon 64 FX-57 had been running happily at an astonishing speed of 3.7GHz before the event, and it unfortunately wouldn't go any higher with the LN2. His chip suffered from the 'cold bug', where the chip literally freezes up when it gets below -120˚C. However, much better luck was had with an Athlon 64 FX-55 based on the older ClawHammer core, which took to the cold like an alcoholic takes to a drinks cabinet. Using the LN2, the chip reached 3.85GHz, which is 1.25GHz above its standard clock speed, and was ranked at number seven on the CPU-Z league tables.
We also had time to fiddle with a brand-new 3.8GHz Pentium 4 670, and its virgin POST was cooled with LN2 - a strange existence to be born into. Amazingly, the chip was still happy at an awesome speed of 6.5GHz with an incredible FSB speed of 340MHz (1,360MHz effective). This secured us a ranking of number three in the CPU-Z league tables. We felt the chip could even go a little further, but again, the motherboard's capacitors were too cold at this point.
Overclocking with LN2 is completely different from any other overclocking experience. The coolant evaporates very quickly, and getting a machine that's stable enough to run any significant benchmarks is tricky. The tactic we went for was a bit like 'suicide clocking', where we were happy as long as it was stable enough to get a CPU-Z validation file.
However, you really need to know your motherboard's BIOS inside out to get the machine into Windows quickly enough to run CPU-Z. Lots of confusing and fiddly settings can really hinder you here, so you only really need a stable motherboard that lets you adjust the FSB. It really puts an emphasis on the importance of a motherboard that works straight out of the box. Software such as ClockGen made our life a lot easier too, as it took the boot process out of the equation. Waiting for a few minutes for the PC to boot up can be enough of a load to deplete your supply of LN2.
Normal overclocking is about putting in hours of work to achieve a stable speed, but overclocking with liquid nitrogen is about the thrill of split-second decisions and adrenalin-boosted BIOS fiddling. There's certainly a lot to be said for seeing -200˚C on a temperature probe, and the atmosphere at the event was great. Barry summed up the experience perfectly after overclocking the Pentium 4, saying, 'We were all in the zone then, just for a few minutes. That's what it's all about.'
What are the ingredients for the perfect party? The answer from most normal people would usually be plenty of beer and some good friends, but when Barry, or bazx as he's known online, is planning a gathering, he likes to let something a little more interesting than beer steal the show. Barry is one of the lucky few that can get his hands on liquid nitrogen, or LN2 as many people call it, and on 28 December, 2005 he invited seven people round to his place for an extreme overclocking party.
When I arrived, the group was huddled around an Abit AI7 motherboard using a lowly Celeron as a guinea pig for the LN2. Strapped on top of this was a copper flask into which they were pouring LN2 from a saucepan. It looked like something out of a classic sci-fi film, with vapour rising out of the flask and evaporating at an alarming rate.
However, the Celeron eventually stopped booting, which was probably because the motherboard's capacitors near the processor were too cold to function properly (the frosty coating was a bit of a giveaway). The board wasn't dead, but it certainly needed to warm up before it would POST again, which unfortunately meant drying it out as well, as the ice would melt.
Barry's facilities were everything you'd expect from a true enthusiast. Scattered around the room were several phase-change coolers, including a cascade unit from America that he was working on. The most interesting part was the desk he had built for his overclocking experiments. Being an engineer, it was much more than just a desk. The top had a monitor on each side and a Shuttle XPC solely for monitoring temperatures. Then, under that was a fileserver just for storing data, and in the centre of the desk was a cabinet that was a little larger than an ATX case. This was connected to an air conditioning unit and a dehumidifier under the desk, which helped to keep the computer at sub-zero temperatures with little or no moisture. Also, on top of the cabinet was a sealed and airtight hole where the evaporator head from a highly modified VapoChill unit entered, and this doubled up as a hole for the LN2 Flask.
Barry's Athlon 64 FX-57 had been running happily at an astonishing speed of 3.7GHz before the event, and it unfortunately wouldn't go any higher with the LN2. His chip suffered from the 'cold bug', where the chip literally freezes up when it gets below -120˚C. However, much better luck was had with an Athlon 64 FX-55 based on the older ClawHammer core, which took to the cold like an alcoholic takes to a drinks cabinet. Using the LN2, the chip reached 3.85GHz, which is 1.25GHz above its standard clock speed, and was ranked at number seven on the CPU-Z league tables.
We also had time to fiddle with a brand-new 3.8GHz Pentium 4 670, and its virgin POST was cooled with LN2 - a strange existence to be born into. Amazingly, the chip was still happy at an awesome speed of 6.5GHz with an incredible FSB speed of 340MHz (1,360MHz effective). This secured us a ranking of number three in the CPU-Z league tables. We felt the chip could even go a little further, but again, the motherboard's capacitors were too cold at this point.
Fastest, cheapest 3G mobile broadband dongles from 3, Vodafone, T-Mobile and Orange
from just £10/month