• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Storm G4 Review!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

SnowRider

Member
Joined
Apr 27, 2004
Location
Central Cali, USA
++ Review Of The Storm G4 ++

Here's the first "real" review I've seen of the Storm G4 blocks. Looks like we have the new King of CPU blocks (not that I'm suprised). This block performs better than the Cascade SS which as far as I know was the best performing block at the time!! Excellent work Cathar :attn::attn: Huge props go to you for all the research, time and money you put into raising the "watercooling bar".
 
Yes congrats. I'm hooking mine up this evening with just my Iwaki MD-20RT and a ProCore. Time for uber flow! :attn: :attn: :attn:
 
Dang that is a sweet block. Excellent review too.

If there is more room for performance improvement in the future, I'm confident Cathar will be the first to do it. :attn:
 
Achilles17 said:
its called the G5 my friend...

.5-1C better than the G4, believe it or not

In the final testing stages of the G5 vs G4 now. Just building up a suite of mounts to achieve a satisfactory level of confidence in my results. I test at a higher heat load than Phaestus does, with a different CPU, and measure temps differently. Historically Phaestus seems to see temps at an average of around 55% of temps that I see. For example, when he sees ~11C above water temps, I would be seeing ~20C above water temps. These are just normal sorts of testbed variations.

Anyway, I'm seeing very close to a full 1.0C difference between the final G5 design and the final G4 on my testbed. For Phaestus's testbed this should roughly translate to a 0.5C difference. Once I complete my full G5 testing, which should be done by the end of this weekend, I'll send it off to Procooling for review as well.

Guderian said:
That makes me sad.

Why?
 
I'm not really surprised too much. Impingement designs need a certain amount of pressure to perform well. The MCW6000 is a much higher flow design than the Storm. Makes me glad I got a Mag 3.

heman.jpg


I HAVE THE POWER!
 
Last edited:
Alacritan said:
I'm not really surprised too much. Impingement designs need a certain amount of pressure to perform well. The MCW6000 is a much higher flow design than the Storm. Makes me glad I got a Mag 3.

he-man.jpg

I HAVE THE POWER!

you have the power?

cathar said over at procooling that he was looking to get better low flow performance IIRC
 
greenman100 said:
I'm surprised at the low flow performance...Cathar, weren't you expecting a little better?
(not cracking on you or your design, just curious)

Crap - wrote a huge reply and lost it. This is a re-type. :(

I personally consider a low-flow setup to be a real-world thing, and not some arbitrary flow rate. I test according to pumping power. In a real-world system, an Eheim 1046 with 3/8" ID tubing would be a "low flow" system by most people's standards. Such a setup would still give around 2.5LPM (0.65gpm) with the Storm/G4, and on this score the G4 delivers exactly as stated, with it exceeding all other blocks that Phaestus has tested (equalled only by the Cascade SS, which is a bit of a special-case block).

I don't believe in designing blocks to work around users who want to artificially choke their systems to silly levels with super-small ID tubing. It's not a setup philosophy that I agree with, and so I don't go out of my way to work to cater for it. Flow rates lower than 2LPM actually requires a fairly determined intent to choke the life out of even the very weak Eheim 1046. If people want to do that, that's fine I guess, but I consider that to be an "artificially low-flow testbed", rather than a realistic scenario.

I personally saw better low-flow performance than what Phaestus did though. The difference is not huge. At a guess what I measure would probably correlate to about 1.0C better at 1LPM, 0.4C better on Phaestus's charts at 2LPM, and about 0.2C better than 4LPM. I do use a larger CPU die though and measure my temperatures differently, so I just put it down to the natural variations between different testbeds due to the different setups. Overall the difference is small and at rather unrealistically low flow-rates so I don't have a problem with his results at all.

The G5 design sub-variant is a much stronger low-flow performer due to the 70% higher structure density. Coupled with the use of silver that'll be the one to keep an eye on if you're interested in seeing something that raises the bar right across the gamut. Phaestus should have it in his hands in about one week from now.
 
greenman100 said:
you have the power?

cathar said over at procooling that he was looking to get better low flow performance IIRC


If my linked image had worked the first time (or I had bothered to check to make sure it worked), it would have made much more sense.


He did get better low flow performance that previous LR blocks, but it's still based upon a design that really excels with high flow.

By the Power of Greyskull!

he-man.jpg
 
I think the next 'evolution' of the procooling graph is to be corrected with availible backpressure data, meaning you won't have to guess how much less flow you'll get going from different blocks like the Cascade to the G4 Storm. That will still require some guessing due to varying flow curves between pumps, but all in all, the graph still does a great job showcasing Cathar's latest triumph. Anyways, the article shows off the block's features very well with the particular angles, but until you have it in hand, its hard to really picture it.
 
Overbrazil said:
Cathar, how many mounts do you make ? and what the average different temps you got ?

Are you referring to my own testing? Or the review at Procooling?

Just recently I retested the G4 to revalidate my older results to compare against the G5. Over 10 mounts I saw a 0.4C variation, excluding clearly erroneous mounts.
 
Back