AquaXtreme™ MP-1 Universal GPU Water Block

jaydee · 09-04-2005, 08:50 PM

Looks like another decent GPU block.

http://www.cooltechnica.com/Assets/A...MP1-GPU_WB.htm

Once I round up the cash I think I will pick one up along with a AquaXtreme 50Z-DC12 and a video card that will work with the block and use the MP-05 Sp I have for the CPU block. I think the blocks will be restrictive but I believe the performance is going to be really good.

BRiT · 09-05-2005, 01:37 AM

Looks awefully nice...

Roscal · 09-05-2005, 07:28 AM

Really nice, Bruce is hot with all these new products

BRiT · 09-05-2005, 11:55 AM

Can someone post up the dimensions of the MP1 GPU block? I'm trying to determine if it's usable in a DFI SLI setup.

Thanks...

dacooltech · 09-05-2005, 08:31 PM

thx guys

Quote:

Originally Posted by BRiT

Can someone post up the dimensions of the MP1 GPU block? I'm trying to determine if it's usable in a DFI SLI setup.

Thanks...

Here you go:
AquaXtreme MP-1 GPU

BRiT · 09-05-2005, 09:29 PM

Quote:

Originally Posted by dacooltech

thx guys

Here you go:
AquaXtreme MP-1 GPU

Sweetness...

RVWinkle · 09-16-2005, 07:34 PM

I like the look of this block but should we be worried about increased restriction compared to other GPU blocks?

jaydee · 09-16-2005, 08:26 PM

Quote:

Originally Posted by RVWinkle

I like the look of this block but should we be worried about increased restriction compared to other GPU blocks?

Bruce told me this block is not as restrictive as it looks. He tuned it for better flow rates. Not entirly sure how that compares to other GPU blocks. Even with lower flow rate though the performance increase should over come that loss. Maybe pH will get around to testing it sooner or later.............

RVWinkle · 09-17-2005, 12:15 PM

Thanks for the reaffirmation Jaydee. I suppose that to my untrained eye this block has the potential for "best in class" performance. Allthough I'm not sure what that means since my impression is that with GPU blocks the goal has resided in minimal restriction. My unsderstanding is that restriction is always an issue to take into account when introducing multiple blocks in a small loop.

A while back I bought one of the asetek GPU blocks with no pins or channels (aside from manufacturing defects). I purchased it because of the price and because of the unscientific assumption that it would be the least restrictive solution. Aside from the fact that I fully expect it to eventually crack, I'm more than satisfied with its' performance and minimal impact on cpu temps.

The fact that the product page recomends 1/2" or 7/16 tubing seems odd to me and I've never seen a GPU block advertised this way. This lends me to beleive there's potential for restriction issues. While I like the look of this part, I would likely install it in a 3/8" loop.

I'm in the midst of rebuilding my loop and there's a good deal of equally priced GPU solutions out there. At the same time there's little data available. Ultimately my question boils down to how would the MP-1 affect performance compared to the MAZE4, Swiftech, etc. I suspect the answer is 'no significant difference' within a limited testing environment.

My planned setup: 3/8" loop with a MCW6000, DDC, MCRES1000, and a PA160 cooling an OC'ed Venice A64 and 7800GT. I would welcome anyone's comments if they think one GPU block would be better than another with this system.

jaydee · 09-17-2005, 12:53 PM

I hope to have one of these blocks soon with a AquaXtreme 50Z pump. I plan to build a system with those and the MP-05 SP block and see how they all work together in a real computer as opposed to just die sim. I think the next few years there will be a larger demand for high performance GPU blocks as CPU's are not making much progress. That means more restrictive blocks. Might be something we have to get used to and work with it.

Just specualation at this point. pH said he was going to start GPU tests so hopefully that will give us more insight into the problem.

RVWinkle · 09-18-2005, 04:30 PM

Well I look forward to formal and informal testing of this block. I think even speculation at this point by people with more experience in block design would have merit too. In the meantime, I intend to try this block out with some of those DD 3/8" high flow fittings as they seem to have the proper thread.

I have some questions that perhaps Bruce can answer if he's not too busy.

-Why the strong suggestion to use 1/2" and 7/16" ID tubing? It seems unusual.

-Someone also recently suggested that blocks similar to this often have clearance issues with popular ramsinks, namely the OCZ copper BGA. Can we have confirmation that this isn't an issue?

maxSaleen · 09-18-2005, 08:04 PM

CPU blocks have sort of hit a wall (see Cathar's post about the G7). Making them perform better isn't really economically feasible. I belive, however, that GPU blocks can get a lot better. Think about it. The die size is about the same as a CPU (actually a lot larger most of the time, but that should make them easier to cool), and the thermal output is lower, yet we consistantly see higher temps (compare a 6800 ultra to a 4000 Sand Diego).

High performance/restriction (the two seem to go hand in hand after a certain point) GPU blocks will lead to an increased need for more powerful pumps. This is fine and dandy in the world of the ubergeek, but powerful pumps are usually too expensive for most of the market. Yet I don't see much improvement in low flow performance unless we can figure out a way to pump the heat off of the die to a larger surface area (think TEC). The guy who started liquid metal cooling is working on some sort of TEC like device to be used BETWEEN the die and the heatsink. He was a chief thermal engineer for IBM, and specialized in "micro refrigeration". Besides blabbing about it in CPU magazine last month, he's kept everything on the down low. No details, no mention of it on Liquid Metal Cooling's website. I'd be interested to see how well it works.

Time to stop rambling......

dacooltech · 09-19-2005, 05:28 AM

Quote:

Originally Posted by RVWinkle

Well I look forward to formal and informal testing of this block. I think even speculation at this point by people with more experience in block design would have merit too. In the meantime, I intend to try this block out with some of those DD 3/8" high flow fittings as they seem to have the proper thread.

I have some questions that perhaps Bruce can answer if he's not too busy.

-Why the strong suggestion to use 1/2" and 7/16" ID tubing? It seems unusual.

-Someone also recently suggested that blocks similar to this often have clearance issues with popular ramsinks, namely the OCZ copper BGA. Can we have confirmation that this isn't an issue?

RVWinkle Jaydee and MaxSaleen are absolutely right... You can not treat high end GPUs releasing 100W + heat any different than high end CPUs anymore... Just like the CPU block designs, the GPU block designs must evolve accordingly...

As for the suggestion to use 7/16"ID and 1/2"ID tubing with high flow 1/2" barbs, some folks e-mailed us asking if it's OK to use 7/16"ID tubing instead of 1/2"ID with 1/2"OD barbs, and the answer is absolutely OK to use 7/16"ID tubing with 1/2" barbs. And that's the reason why I put Cooltechnica recommends 7/16"ID and 1/2"ID tubing with 1/2" barbed fittings...

I'm getting some high flow 3/8" barbs with large inside diameter made as well...

FYI I personally use 3/8"ID tubing over 1/2"barbed fittings, and use 1/2"ID tubing only for Tee F&B, or between the outlet of a reservoir and the pump. You can simply heat up the ends of 3/8"ID tubing and stretch it over larger 1/2"OD barbs...

Blackeagle · 10-05-2005, 10:27 AM

Bruce I'd also like to know how tall RAM sinks used with your block can be if 1/16" walled tube is used.

I suspect the underside of the lid shows a raised area between the inlet/outlet to force the water to flow through the pins/channels, can you confirm this, or better, post a pic of the inside of the blocks top?

From first look I like it. Larger ID barbs than most to help with pressure drop & good structure to improve cooling.

Nice work, thanks.

09-04-2005, 08:50 PM	#1
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	AquaXtreme™ MP-1 Universal GPU Water Block Looks like another decent GPU block. http://www.cooltechnica.com/Assets/A...MP1-GPU_WB.htm Once I round up the cash I think I will pick one up along with a AquaXtreme 50Z-DC12 and a video card that will work with the block and use the MP-05 Sp I have for the CPU block. I think the blocks will be restrictive but I believe the performance is going to be really good.

09-05-2005, 01:37 AM	#2
BRiT Cooling Neophyte Join Date: May 2005 Location: Cleveland, OH Posts: 39	Looks awefully nice... __________________ --\|BRiT\| / WTF?! "That which does not kill you, makes you wish it did."

09-05-2005, 11:55 AM	#4
BRiT Cooling Neophyte Join Date: May 2005 Location: Cleveland, OH Posts: 39	Can someone post up the dimensions of the MP1 GPU block? I'm trying to determine if it's usable in a DFI SLI setup. Thanks... __________________ --\|BRiT\| / WTF?! "That which does not kill you, makes you wish it did."

09-17-2005, 12:15 PM	#9
RVWinkle Cooling Neophyte Join Date: Oct 2004 Location: Richmond Posts: 13	Thanks for the reaffirmation Jaydee. I suppose that to my untrained eye this block has the potential for "best in class" performance. Allthough I'm not sure what that means since my impression is that with GPU blocks the goal has resided in minimal restriction. My unsderstanding is that restriction is always an issue to take into account when introducing multiple blocks in a small loop. A while back I bought one of the asetek GPU blocks with no pins or channels (aside from manufacturing defects). I purchased it because of the price and because of the unscientific assumption that it would be the least restrictive solution. Aside from the fact that I fully expect it to eventually crack, I'm more than satisfied with its' performance and minimal impact on cpu temps. The fact that the product page recomends 1/2" or 7/16 tubing seems odd to me and I've never seen a GPU block advertised this way. This lends me to beleive there's potential for restriction issues. While I like the look of this part, I would likely install it in a 3/8" loop. I'm in the midst of rebuilding my loop and there's a good deal of equally priced GPU solutions out there. At the same time there's little data available. Ultimately my question boils down to how would the MP-1 affect performance compared to the MAZE4, Swiftech, etc. I suspect the answer is 'no significant difference' within a limited testing environment. My planned setup: 3/8" loop with a MCW6000, DDC, MCRES1000, and a PA160 cooling an OC'ed Venice A64 and 7800GT. I would welcome anyone's comments if they think one GPU block would be better than another with this system. Last edited by RVWinkle; 09-17-2005 at 12:37 PM.

10-05-2005, 10:27 AM	#14
Blackeagle Thermophile Join Date: Oct 2002 Location: U.S.A = Michigan Posts: 1,243	Bruce I'd also like to know how tall RAM sinks used with your block can be if 1/16" walled tube is used. I suspect the underside of the lid shows a raised area between the inlet/outlet to force the water to flow through the pins/channels, can you confirm this, or better, post a pic of the inside of the blocks top? From first look I like it. Larger ID barbs than most to help with pressure drop & good structure to improve cooling. Nice work, thanks. Last edited by Blackeagle; 10-05-2005 at 11:17 AM.

09-05-2005, 07:28 AM	#3
Roscal Cooling Savant Join Date: Apr 2002 Location: North of France Posts: 198	Really nice, Bruce is hot with all these new products

09-16-2005, 07:34 PM	#7
RVWinkle Cooling Neophyte Join Date: Oct 2004 Location: Richmond Posts: 13	I like the look of this block but should we be worried about increased restriction compared to other GPU blocks?

09-17-2005, 12:53 PM	#10
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	I hope to have one of these blocks soon with a AquaXtreme 50Z pump. I plan to build a system with those and the MP-05 SP block and see how they all work together in a real computer as opposed to just die sim. I think the next few years there will be a larger demand for high performance GPU blocks as CPU's are not making much progress. That means more restrictive blocks. Might be something we have to get used to and work with it. Just specualation at this point. pH said he was going to start GPU tests so hopefully that will give us more insight into the problem.

09-18-2005, 04:30 PM	#11
RVWinkle Cooling Neophyte Join Date: Oct 2004 Location: Richmond Posts: 13	Well I look forward to formal and informal testing of this block. I think even speculation at this point by people with more experience in block design would have merit too. In the meantime, I intend to try this block out with some of those DD 3/8" high flow fittings as they seem to have the proper thread. I have some questions that perhaps Bruce can answer if he's not too busy. -Why the strong suggestion to use 1/2" and 7/16" ID tubing? It seems unusual. -Someone also recently suggested that blocks similar to this often have clearance issues with popular ramsinks, namely the OCZ copper BGA. Can we have confirmation that this isn't an issue?

09-18-2005, 08:04 PM	#12
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	CPU blocks have sort of hit a wall (see Cathar's post about the G7). Making them perform better isn't really economically feasible. I belive, however, that GPU blocks can get a lot better. Think about it. The die size is about the same as a CPU (actually a lot larger most of the time, but that should make them easier to cool), and the thermal output is lower, yet we consistantly see higher temps (compare a 6800 ultra to a 4000 Sand Diego). High performance/restriction (the two seem to go hand in hand after a certain point) GPU blocks will lead to an increased need for more powerful pumps. This is fine and dandy in the world of the ubergeek, but powerful pumps are usually too expensive for most of the market. Yet I don't see much improvement in low flow performance unless we can figure out a way to pump the heat off of the die to a larger surface area (think TEC). The guy who started liquid metal cooling is working on some sort of TEC like device to be used BETWEEN the die and the heatsink. He was a chief thermal engineer for IBM, and specialized in "micro refrigeration". Besides blabbing about it in CPU magazine last month, he's kept everything on the down low. No details, no mention of it on Liquid Metal Cooling's website. I'd be interested to see how well it works. Time to stop rambling......

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)