Is anyone else bored?

[phide]

It seems like it's been so long since we've seen any fresh, innovative products coming out for us hardcore water cooling enthusiasts. Instead of seeing interesting new block designs, for instance, we're merely seeing re-hashes of old designs re-worked to fit new procs, cards, etc. Same old pin grid, straight channel and (occasionally) wavy channel designs. Rinse and repeat, then charge $60 (or whatever).

I equate some of this to the absence of Cathar (to a point), but nobody "in the game" is even attempting to further maximize the potential of water cooling. The Swiftech Storm has, for a very long time, been the top performing block for 3/8" / 1/2" systems, and it's perpetuated itself into pretty much every 1/2" DIY rig making so many custom loops mirror images of one another. There's little room for further improvement in block design, but room is there, and it seems like someone would be trying to hit the apex, but this doesn't seem to be the case.

It seems that all the industry giants have simply fallen into a position of total stagnancy and non-interest in further advancing the performance and usability of do-it-yourself water cooling.

My own level of interest has been rapidly falling lately, as there's really very little I can do to make my loop any more effective than it is already (with the almost two-year old Storm G5 which I fear is rapidly deteriorating), and nothing that would make my setup any more interesting or appealing.

So, the question at hand: Is anyone else bored?

 

[nonlnear]

The market is mature now. The days of the home tinkerer are past for many aspects of watercooling. The basics of how to build a good block seem to fairly well known (until somebody figures out a better way than impingement). There are a couple solid pumps out there, and they will continue to be improved marginally.

I think there is still room for major improvements in case design. It's a crying shame that so many people are forced to either use up drive bays, or use a radbox. Both are aesthetically repulsive to me. How about some more PSU placement options? Is that too much to ask for? The Lian-Li V series (and similar cases) are a good first step, but there are a lot of other configurations that have potential.

I think there's also an opening for better cooling system hardware/software integration. Whether you like AC or not, their Aquaero USB control is a step in the right direction. Unfortunately, it's at the application level. There is room for a lot tighter integration - hopefully at the BIOS level. That would require a little bit more intelligence to be built into the APIs though. More programmability, and less prefab "threshold temps/RPM monitoring". Especially if you've got GPUs in the loop to worry about monitoring as well.

 

[chris.c]

Not a lot more than can be done now.

Waterblocks, given the pumping power available, are about at the limit. There's no big gains left any more. There are refinements (such as the G7's larger cooling patch to service IHS capped CPU's a little better) and small improvements (G7's lower pressure drop, more even internal flow distribution and increased fine jet density), but overall in terms of thermal transfer efficiency per unit of pumping power, it's pretty tight.

For radiators, I put everything I knew about waterblock flow and lots of added analysis and emperical testing on fan flow and efficiency into the PA design. I specifically targetted making the radiator super efficient for lower powered fans (60cfm and less) 'cos I couldn't see anyone in the industry who was targetting that area. Turns out that the PA radiators still kick butt up to 100cfm fans too, if you can stand the noise.

For pumps, I'd been pushing hard both on-line over at Procooling, and behind the scenes with a number of industry players. While I never spoke directly to Laing, quite a fair bit of feedback and suggestions did flow from me to them through different channels, and now we have the Laing DDC+, which if you read my specs for the "perfect pump" over at Procooling, the DDC+ is almost exactly what I was pushing for. With the after-market 1/2" tops, the 18W DDC+ is exactly what I ever envisioned the ideal PC water-cooling pump to ever be.

Straight from the camel's mouth. Can't have new and innovative products if there is nothing left to innovate.

There's little room for further improvement in block design, but room is there

Yea, room is there, but it comes down to price vs. performance. Cathar's storm G7 uses all of that room, however it will cost ~$300 and the production will be extremely limited.

Am I bored? A better word would be content. I am comfortable knowing I have the best parts in my loop, and I know I won't have to replace them anytime soon because there is nothing left to improve.

 

[el rolio]

The market is mature now. The days of the home tinkerer are past for many aspects of watercooling. The basics of how to build a good block seem to fairly well known (until somebody figures out a better way than impingement). There are a couple solid pumps out there, and they will continue to be improved marginally.

I think there is still room for major improvements in case design. It's a crying shame that so many people are forced to either use up drive bays, or use a radbox. Both are aesthetically repulsive to me. How about some more PSU placement options? Is that too much to ask for? The Lian-Li V series (and similar cases) are a good first step, but there are a lot of other configurations that have potential.

I think there's also an opening for better cooling system hardware/software integration. Whether you like AC or not, their Aquaero USB control is a step in the right direction. Unfortunately, it's at the application level. There is room for a lot tighter integration - hopefully at the BIOS level. That would require a little bit more intelligence to be built into the APIs though. More programmability, and less prefab "threshold temps/RPM monitoring". Especially if you've got GPUs in the loop to worry about monitoring as well.

oh wow you hit it on the head. funny enough im running a pa triple, storm block, eddy barbs, oine of hte early modded top DDC+'s and all that jazz, and i have an aquaero too. and i had to HUNT that damn thing down. so yea i wish there was more implementation of the aquaero, more ways for it to work with my pump, ddirectly hook into gpu and cpu stuff like you said etc. at the very least i can have it monitor in line water temps and somewhat adjust fan speeds accordingly, but relaly thats all i can get it to do to go form passive coolign to silent fan cooling.

but with what you said, i think you got it right. the way we need to be thinking now is designing cases better and better for this much larger WC market, start breaking these molds we are in for weird air cooling. silence, wc-directed design is where we should be headed into next year and until i seee thati m not sure hwat case i would get next to migrate my parts into. i love the fact that i relaly wont have to change the main parts of my loop, jsut clean em, but i want better and more elegant ways to incorporate them. WC has lots more room to grow, maybe not block design, but def lots more room.

 

[cyr0n_k0r]

there is nothing left to improve.

that kind of thinking is why people don't innovate.

What if everyone believed when people said you would never need more than 600k of memory. What if they were just like.. "yeah, hes right" and gave up?

Or the people in the patent office back in the 1800's that thought there was nothing left to invent. Should they just close the patent off?

When you starting thinking there is nothing left to invent or improve, there wont be.

 

[phide]

Yea, room is there, but it comes down to price vs. performance. Cathar's storm G7 uses all of that room, however it will cost ~$300 and the production will be extremely limited.

To a certain extent, yes. But it could also be tweaked for specific flow rate ranges, instead of the largely "flow agnostic" G4/G5/G7. There's a slight amount of potential in that particular design, but there may be other design considerations he may have stricken out to achieve performance at a variety of flow rates. What would I want? A design that out-performs the G7, in silver, at flow rates greater than ~4GPM and is somewhat "inferior" at very low flow rates (1-2GPM). According to Cathar, it's entirely feasible. The price factor is irrelevant to me - I just want that envelope to be pushed as far to the limit as possible.

silence, wc-directed design is where we should be headed into next year

I totally agree. Air cooling really has hit the limit, as water cooling has come very close over the past few years, and with heat dissipation increasing almost linearly with time, true mainstream liquid cooling may become the norm. I'm still a firm believer that the ideal water cooling case is still the UFO, and that there are water-oriented cases on the market, but they are just too few and too expensive.

Of course, this isn't something that really removes the "bore" issue. Affordable water-oriented cases may be coming eventually, but they will have little impact on performance, which is what I'm primarily concerned with. Integration? Interesting, and totally desirable, but something I still don't find exciting. I suppose I could be the exception.

 

[nonlnear]

Affordable water-oriented cases may be coming eventually, but they will have little impact on performance, which is what I'm primarily concerned with. Integration? Interesting, and totally desirable, but something I still don't find exciting. I suppose I could be the exception.

Think automatic pump/fan throttling based on programmable parameters. Like CPU and GPU load history as well as temps. i.e. you could set a throttle function based on temps, and a throttle based on loads, and take the max of the two - so that things would spin up preemptively when loads spike, but temps haven't yet risen. (Of course, you'd have to smooth the load curve some, unless all your components are happy with PWM )

That's sort of a pipe dream, but even haveing reliable temperature probes on multiple chips (i.e. GPUs too), with outputs available through a standard interface. That would be something I think every performance-minded watercooler would want.

 

[Arcygenical]

God. Thermal probes that are ACCURATE.

We can clock a chip to 5ghz, but we can't accurately measure it's temperature with onboard components... Gee.

 

[Rocky Mountain High]

We need accurate coolant flow meters with low restriction, similar to those manufactured by Swissflow.

To prevent nozzle clogging, we need low-profile in-line strainers, such as those manufactured by www.specialtymfg.com.

We need control devices, such as those manufactured by Crystalfontz, that can receive electronic input, interpret it, display data and make adjustments.

 

[Anavel0]

Everything does seem to have hit a limit here resently. Be it air cooling or water cooling. We just keep seeing larger heatsinks with more heatpipes in air cooling, and most water cooling taking after Swiftech at the moment.

I think the future will be based on several factors. Either an innovation in water cooling/air cooling. Or AMD and Intel will both decide that processors are putting out too much heat and come up with a way to lower proc temp in the first place. If AMD or Intel can pull this off I don't think we will see too much really in terms of cooling advancements, just refinements to what we have now.

 

[MrWizard6600]

@ phide, drill some holes, mount some radiators, build some custom enclosures, mount the radiators in your custom enclosure, extend your loop to include your new wicked sick radiator.

/bordum

@ Anavel0, totally. when Grayhound comes out, we need like an FX 65, while the FX 64 will have awsome performance and run stable under the crap stock hs/f amd provides, the FX 65 would require massive V-mods, and require liquid cooling, as it would produce ground breaking amounts of heat. and while the FX 64s humming along at 2.8GHz, the FX 65 would be crashing along at 5GHz
-- of course along with the release of this products, amd would attach the note "take THIS Core 2!"

that would give a royal kick to waterblock manufacturers to get started and start developing.

 

[phide]

God. Thermal probes that are ACCURATE.

Isn't it funny how insanely difficult and expensive it is get reasonably accurate core monitoring? Everything else seems so easy by comparison.

Or AMD and Intel will both decide that processors are putting out too much heat and come up with a way to lower proc temp in the first place.

Well, there are ways (some that include life-threatening gamma radiation!), but I don't think we'll be there for a while yet. Maybe once we start seeing packages with eight or more logical cores, we're going to see wild nano-pump, lead IHS type dealies. For now, AMD and Intel are trying to keep output low without doing anything too crazy (read: semi-boring process jumps).

We need...

Maybe if we ask Petra real nice-like...

 

[Arcygenical]

The new DD block for the 7950Gx2 is pretty awesome!

 

[sprocket]

I think we're bored because we're in the twilight of the watercooling age altogether.

Laptops increasingly dominate the new sales market and the latest generations of desktop hardware are becoming less power hungry and require less cooling.
With the (temporary) exception of vid cards everything else requires lower wattage (merrily we go green!) so cooling becomes less of an issue/problem.

So, smaller, quieter, cooler...where does watercooling fit in this brave new world except an increasingly narrow niche market?

There may indeed be a "new Cathar" lurking in the bushes- someone who possesses the talent, inquisitiveness and rigorous attention to detail and scientific principle of Stew- but where is the challenge for such a person when the big manufacturers are removing his very reason to exist?
The technology is already in place to deal with a Prescott...how hard do you have to work to cool a 35w nextgen CPU?

I've always known that my watercooling obsession was based on the fact that I'm a hardware junkie...I like to tinker...but the absurdity of my expenditures (both time and money) is driven home every day as I work on Dells, eMachines, etc.
My customer's PCs idle at temps that would give me a coronary if my PC hit them under full load and yet their cheap POS computers are just as silent (easy to do when you only use one fan!) and have been running for years.

I'm enjoying the bounty of H2O gear currently available because I'm pretty sure my next major platform upgrade won't have/need any of it.

*sigh*.

 

[Erasmus354]

Phide : The Storm waterblocks are most certainly NOT flow agnostic. A block like the apogee or MCW6000 are flow agnostic. Look at the flow chart, if it is relatively flat then it is flow agnostic. The Storm charts are far from flat, and improve a lot with increased flow. Just because they still perform well at lower flow does not make them flow agnostic.

 

[Marci]

Think automatic pump/fan throttling based on programmable parameters. Like CPU and GPU load history as well as temps. i.e. you could set a throttle function based on temps, and a throttle based on loads, and take the max of the two - so that things would spin up preemptively when loads spike, but temps haven't yet risen. (Of course, you'd have to smooth the load curve some, unless all your components are happy with PWM )

That's sort of a pipe dream, but even haveing reliable temperature probes on multiple chips (i.e. GPUs too), with outputs available through a standard interface. That would be something I think every performance-minded watercooler would want.

To those going on about monitoring and control... mCubed TBalancer and BigNG do all you're asking for, and support swissflow monitors etc, and allow response curves to be created based on various variables. Well worth checking out. Also allow interfacing to the common LCD Displays available.

With Cathar's comments... he's talking from an efficiency/performance standpoint. The Storm blocks are as thermodynamically efficient at the most efficient flow rate (for our purposes) as a waterblock can be... soon as C/W is nearing abs 0 (and therefore efficiency nearing 100%), there isn't much tweaking can be done to improve performance, and any tweaking that can be done will be in the fractions of degrees... hence his ceasing work in this field.

 

[nonlnear]

To those going on about monitoring and control... mCubed TBalancer and BigNG do all you're asking for, and support swissflow monitors etc, and allow response curves to be created based on various variables. Well worth checking out. Also allow interfacing to the common LCD Displays available.

Yeah, I guess your'e right. I wish the integration didn't have to be in that form, but I don't see BIOSes becoming any more elaborate any time soon. And on die temp sensors, well we've been waiting for them to improve for years now

I would still prefer to see a TBalancer type device able to interface directly with the BIOS instead of through applications. But given what options are available, they do seem to be the best thing going. Then again, I haven't seriously looked into them for a while now. They might have rolled all those features into a version I didn't notice.

 

[phide]

The Storm waterblocks are most certainly NOT flow agnostic.

I'm well aware of the curves. I should mention that I said, quite specifically, "largely 'flow agnostic'. The quotation marks indicate a Stew quotation.

The block was designed to be as close to "flow agnostic" as possible, thus no devices that would significantly impact performance at low flow rates. My point is that these devices can be implemented and various proportions of components tweaked to create a block that really only performs well at very high flow rates, yielding a less "flow agnostic" design.

 

[R1ckCa1n]

Cathar posed long ago about hitting the "wall" which is why I can't wait to see his pelt water chiller get moving forward. That will be next to get temps more than .5-1c better than current high end blocks. I just not ready for a refrigerator in my office doing something other than cooling my beers.

 

[Cathar]

Hi Guys,

Making water-cooling better cannot be compared to making computers better (more than 640kb or whatever). There are fundamental thermal limitations of the liquids, metals, and gasses that we're dealing with.

It might be fun and fantasy to project to levels of infinite thermal convection rates, and postulate over what's possible, in which case we're only about 60% of the way along the performance curves. i.e. if thermal convection rates were infinite (but still keeping the same thermal paste and metallic boundaries), we could drop the rise above ambient temperatures of a CPU or GPU by around 40% more than what the best hardware today presently achieves. Read that again and appreciate that we're talking about infinite thermal convection rates to achieve that, and still we're only talking about 40% gains.

Of course, the reality is that achieving infinite thermal convection rates really is just fantasy. Read enough research material on the matter, and do the maths (albeit with appreciable margins of error based on measurement inaccuracy) and the top water-blocks are all pushing thermal convection rates that are as high as anyone has ever achieved anywhere outside of very carefully controlled and impractical laboratory conditions. If we take the best of what has been achieved anywhere, and then we're talking about perhaps 15% gains left to achieve, and even then using methods which are wholly impractical for affordable mass production, and/or relatively low maintenance use.

We're not climbing a linear cost/performance curve here. We're climbing an efficiency = 1/x curve. The further you climb, the steeper it gets.

Speaking of which, this is perhaps paralleled by the computing industry now. Yeah, sure. 20-30 years ago we were operating at the lower end of the possibilities of CPU manufacturing techniques and there were huge advances to be made. Nowadays the CPU makers really are pushing up against some pretty hard physical limitations as well. After all, it does take so many silicon atoms to form a hardware gate. If you've ever seen the electron microscope pictures of a CPU gate where you can actually count the physical silicon atoms that make up a CPU gate, it doesn't take a whole lot of brain-power to appreciate that eventually you get down to so few atoms that it's just going to not work any more.

Anyway, we are approaching the limits. Aside from blowing more air through a radiator (whether that be a larger radiator with more fans, or stronger fans), or pushing higher flow rates (and raising radiator capacity to compensate for increase pump heat dump), given some certain fixed pump and fan power, the story really is pretty tight. At this stage it's a bit like expecting the athletes at the next Olympics to exceed previous records by 10% or more. Sure, some years back water-cooling was like the Olympics run with athletes who did not train at all, and the records set were not an accurate measure of what really was possible with the human body. Huge gains were possible in various areas, but fast-forward to today and the blocks, pumps, and radiators are like well honed and trained athletes. They're operating at close to the peak of the game.

I won't depress everyone and say that nothing is going to improve from here on, but to expect anything big is not going to happen, because the physical limitations tells us that this is not possible. There will be some small gains, but given a fixed pump and fan power, and radiator size limitations, we're pretty close to the wall.

Sorry to bore you with the long post.

 

[el rolio]

thanks for the post boss, and again i gotta say what im more interested in, is for all the other things that can matter to be dealt with. case design with watercooling in mind. case design with silence in mind. case design with both in mind. more ways of display, measurement and control. we have the main parts of the machine done (engine, intake and exhaust), now lets get the stuff that makes it a more enjoyable experience and that can make it more powerful in some respects (better wiring, better plugs, better mounts, better clutch, suspension, alu space frame, aerodynamics, etc). its a loosely drawn analogy but if you ever tuned a car beyond upgrading a filter you should know what im talking about here.

im not bored cuz i got the good core parts (storm, PA triple, DDC+) now i want to bring together a cohesive, powerful, and most importantly: TUNED package, that gives me performance and much closer to the silence. and clean.

 

[Cathar]

Yep, "packaging" is where it's at now.

 

[n00btard]

so we're close to being done with sheer performance... what about improving upon other aspects? Near-silent high-performance pumps? Radiators that only require quiet, low-RPM fans for awesome cooling performance? FORM FACTOR (lower-profile blocks, thinner rads, and such)?

or do we move to a different cooling medium from now such as liquid metal, and start seeing newer technologies in its infancy like electromagnetic pumps silently pushing said cooling medium through tubing that doesn't necessarily have to be 1/2" ID?

or do we stick with what we have, and rice it up like what Thermaltake has been - unsuccessfully - doing for years?

 

[el rolio]

Radiators that only require quiet, low-RPM fans for awesome cooling performance? FORM FACTOR (lower-profile blocks, thinner rads, and such)?

i direct you to the Thermochill PA 120.3 radiator. nothing comes close. i see no performance difference if my fans are at 5v or 9v. and nothing that changes te temps or OC'ability at full 12v. and these are yate loons im talkin about so there isnt even AIRFLOW whoosh sounds. yea we got the performance.

 

[Cathar]

so we're close to being done with sheer performance... what about improving upon other aspects?

Such as?

Near-silent high-performance pumps?

Laing DDC+, with optional 1/2" replacement top if that's your thing.

Radiators that only require quiet, low-RPM fans for awesome cooling performance?

Thermochill PA120.x radiators.

FORM FACTOR (lower-profile blocks, thinner rads, and such)?

There are a number of 1U compatible blocks on the market. There are some decent thinner rads, the old Black Ice Pro, or the Swiftech QP rads.

or do we move to a different cooling medium from now such as liquid metal, and start seeing newer technologies in its infancy like electromagnetic pumps silently pushing said cooling medium through tubing that doesn't necessarily have to be 1/2" ID?

Can go to as low as a well designed 8mm ID loop before the tubing and fittings start to noticably affect performance. Between 8mm and 1/2" the differences are minor, but people like to not choke flow more than necessary.

Liquid metal solutions, for all that they purport to promise, is hindered by severely low flow rates that at the end of the day mean that they perform no better than a good water-cooling system. For all the increased thermal convectional rate that the liquid metal solutions offer over water, the sheer low thermal capacity per second of the fluid transfer is such that it balances out. You are then still left with the problem of what to do with the heat. The BIG problem is how to transfer the heat into the air, and for that there's no substitute for a big radiator that dwarfs the size differences between a compact water-cooling system versus a compact liquid metal solution. If the liquid metal has to carry the heat to some decent radiator, that's more liquid metal fluid that has to be used, and so that drives the cost up 'cos liquid metal is not cheap. If you cut through the marketing hype and analyse the liquid metal solutions, they occupy a size/price/performance niche somewhere between a good air/fan-only heatsink and a mediocre water-cooling solution.

or do we stick with what we have, and rice it up like what Thermaltake has been - unsuccessfully - doing for years?

i.e. back to packaging concerns. The teutonic systems focused on packaging first and performance second. The anglo systems have traditionally been the other way around. In the last few years there's been some good cross-pollination middle ground going on, where systems are being built as a bit of a best possible compromise of compact, near silence, and high performance. Of course the DIY crowd don't like compromises, but the truth is that there's no longer a big difference between a DIY setup of the best components as opposed to a top-end compact kit product.

Thermaltake is a very bad example to use. Alphacool is perhaps a better example, albeit somewhat costly for anything other than a basic setup. There's a marketing niche in there for something that marries the Alphacool bling-performance approach with the Swiftech functional-performance approach, although the marketing gap between the two is shrinking. Koolance's whole-case solutions are perhaps better, but then just fall into the same old trap of trying to be trendy but and up castrating the DIY'ers creativity.

What would be nice would be to see some water-cooling friendly cases. Something like the Lian-Li V2000 cases, but with the lower sections specifically designed for use with radiators and pumps, and then let people individualise the base-case.

 

[el rolio]

What would be nice would be to see some water-cooling friendly cases. Something like the Lian-Li V2000 cases, but with the lower sections specifically designed for use with radiators and pumps, and then let people individualise the base-case.

as the other guy and myself have been saying, this is where im most definately NOT bored. right now we have cases with 120mm fan slots right? no more 80mm fan slots. now its better grills stamped and all that. what aabout optional pump brackets that coem as a case accessory. what about premade fan/vent slots that you can put a double or triple radiator on? top of case? bottom of case? with a BUILT in grill. how about a case thats built that requires no hole-saw drilling, deburring, sanding, hole punching or masking-taping? and how about that case coming from lian li, silverstone, antec, etc?
since the parts arent changing much, these thigns could easily be done with the 3 or so mounting holes and setups since its only the difference between HWL / thermochill / german... etc. ok im done for tonight. ima go watch a movie!

 

[n00btard]

@ Cathar's comments...

now we all know what happens when an engineer analyzes a post made by a guy who based his knowledge on a year-old CPU article and minimal research...

I think we've pretty much reached the apex of the watercooling mountain here.

 

[Petra]

what about premade fan/vent slots that you can put a double or triple radiator on? top of case? bottom of case? with a BUILT in grill. how about a case thats built that requires no hole-saw drilling, deburring, sanding, hole punching or masking-taping? and how about that case coming from lian li, silverstone, antec, etc?
since the parts arent changing much, these thigns could easily be done with the 3 or so mounting holes and setups since its only the difference between HWL / thermochill / german...

Awww.... but that takes all the fun out of it

 

[phide]

Laing DDC+, with optional 1/2" replacement top if that's your thing.

I think that was the "big new thing" for this year. I'm somewhat surprised I haven't bought a couple - what the hell has happened to me?

In the last few years there's been some good cross-pollination middle ground going on

Hah! That certainly paints an interesting image

how about a case thats built that requires no hole-saw drilling, deburring, sanding, hole punching or masking-taping?

I think the main issue is that case manufacturers aren't concerned about accomodating Very Large Radiators. The manufacuters you've mentioned are aware and probably interested in this "water cooling" concept, but aren't tuned in to what enthusiasts are looking for, and aren't aware of the potential market for such case designs. The market's not massive, and certainly never will be, but by swapping a few parts and panels for alternate, pre-configured panels (a real easy trick), they can appease the hardcore and move enough units to make it worth their while.

Perhaps if those interested in such cases would simply shoot off a few e-mails, it would clue these guys in a bit. An interesting thought, and probably worth a shot (if you can break any potentially tricky language barriers, that is).

 

[S14 Racer]

Hi Guys,

Sure, some years back water-cooling was like the Olympics run with athletes who did not train at all, and the records set were not an accurate measure of what really was possible with the human body. Huge gains were possible in various areas, but fast-forward to today and the blocks, pumps, and radiators are like well honed and trained athletes. They're operating at close to the peak of the game.

I only come on the [H], for nostalgia purposes anymore, and to try to keep a slight eye (that eye having 20/50 vision looking from 50' away haha) on cutting edge.

But I saw this and remembered the feelings from back then. Long gone are the days of the cross drilled water block, the pin fin water block, oil coolers with turbulators in them (although unknown for many running them at the time), and the obnoxious but cutting edge bong cooler

I would spend days on end trying anything I could think up, and even when I messed up, still saw gains... Those were the fun days. I can see why people are getting bored with water cooling today.