Posted here: http://vapor.skinneelabs.com/i7/Round2/XT/R2i7XT.html
Updated Overall Comparison here: http://vapor.skinneelabs.com/i7/Roun...i7Overall.html
Awaiting my Gigabyte modified base plate before I can start up my system again
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Swiftech Apogee XT review
- Thread starter Main
- Start date
Posted here: http://vapor.skinneelabs.com/i7/Round2/XT/R2i7XT.html
Updated Overall Comparison here: http://vapor.skinneelabs.com/i7/Roun...i7Overall.html
Awaiting my Gigabyte modified base plate before I can start up my system again
Kaldskryke
[H]ard|Gawd
- Joined
- Aug 1, 2004
- Messages
- 1,346
A quick note to anyone who might be confused by that chart: The XT+ and HK+ were modified by the tester using silicone, hence the +. They are not special versions of the block for sale.
Yes - the plus is after the "siliconemod" for both XT and Heatkiller
This highly effective mod is very easy to do - the drawback is a more restrictive block since this mod hinders water flowing above the micropins.
This mod is is highly recommended !
This highly effective mod is very easy to do - the drawback is a more restrictive block since this mod hinders water flowing above the micropins.
This mod is is highly recommended !
Last edited:
Extra-Titanian
2[H]4U
- Joined
- Sep 21, 2009
- Messages
- 2,414
Your updated overall comparison link is busted.
Xilikon
[H]ard|DCer of the Year 2008
- Joined
- Oct 12, 2004
- Messages
- 15,010
Where's the HK 3.0 CU in the chart ???
lollerskater69
[H]ard|Gawd
- Joined
- Jan 3, 2009
- Messages
- 1,231
here wondering where it is as well.
Add yet another to the list.
Gabriel Rouchon
n00b
- Joined
- Mar 4, 2003
- Messages
- 12
Agreed. I'm curious as well...
FYI we conducted (and published) all our test own test data with the HK 3.0 CU version. You may chose to question/ignore our data of course, but it shows 1.5C Delta in favor of the XT. A difference in methodology should be noted between us and Vapor's procedures: we report the best of 5 mounts instead of the average of X mounts. We believe in effect that the best mount reflects what a block can do at its best. Another scientific reason for this approach is to reduce mathematical data massaging versus presenting measured data.
Looks good, but do we now have a filter on top of the CPU? Does this warrant adding in-line filters... or would it be no different than the GTZ?
I'm satisfied with these tests, but now I want data on LAPPED CPUs. I wonder if it'll still edge out the HK 3.0. - just a curiosity. I'm not sure that the XT, being a large block, is compatible with lapped processors... or how easy/difficult it'd be to mod it for compatibility (seems we could lap the perimeter of the block the same amount as the CPU to clear the socket - but maybe there's a more obvious solution... on the socket itself or something).
I'm satisfied with these tests, but now I want data on LAPPED CPUs. I wonder if it'll still edge out the HK 3.0. - just a curiosity. I'm not sure that the XT, being a large block, is compatible with lapped processors... or how easy/difficult it'd be to mod it for compatibility (seems we could lap the perimeter of the block the same amount as the CPU to clear the socket - but maybe there's a more obvious solution... on the socket itself or something).
I've tested the XT every which way till Sunday, and I show it 0.7c behind the HK3 with copper top. The internal pin matrix is chewed up with poor inconsistent milling on the XT, and I can see small pits on the the base that appear to come from casting also.
From what I can tell, the XT is no where near as well made as the GTZ was, but does perform better than the GTZ. That much they got right. Another thing, the XT mounting
system doesn't seem to crank the block down any where near as much as the GTZ mounting system did. You can spin the XT screws down easily to the stops, but you almost always has to use a screw driver to crank the GTZ springs down all the way unless you want to remove skin from the finger tips.
That all being said, I test as "best used", not "as sold" on the other blocks with my own preference of springs, mounting pressure and back plates. So maybe that has something to do with the difference I'm seeing in performance from what others have stated.
The XT is a decent performer with one of the best mounting systems available, but I haven't seen one reason to swap a HK3 out for a XT yet based on any performance or workmanship criteria.
I was much more excited about the GTZ when it first came out and I tested it for Gabe. The XT just doesn't seem to be up the same standard or quality based on the retail version I purchased and tested.
andyc
From what I can tell, the XT is no where near as well made as the GTZ was, but does perform better than the GTZ. That much they got right. Another thing, the XT mounting
system doesn't seem to crank the block down any where near as much as the GTZ mounting system did. You can spin the XT screws down easily to the stops, but you almost always has to use a screw driver to crank the GTZ springs down all the way unless you want to remove skin from the finger tips.
That all being said, I test as "best used", not "as sold" on the other blocks with my own preference of springs, mounting pressure and back plates. So maybe that has something to do with the difference I'm seeing in performance from what others have stated.
The XT is a decent performer with one of the best mounting systems available, but I haven't seen one reason to swap a HK3 out for a XT yet based on any performance or workmanship criteria.
I was much more excited about the GTZ when it first came out and I tested it for Gabe. The XT just doesn't seem to be up the same standard or quality based on the retail version I purchased and tested.
andyc
Last edited:
My block too got some anomalies in the pin matrix, they are in the outer zone - so I don't belive at all that this will affect the blocks performance.
As from I can tell from my 3 days testeing with the block it easily outperforms GTZ and Heatkiller (which I also have both of). During my short tests I am not really sure if I got burnt in the tim enough either - so I might get lower temperatures when I will be finished reconfiguring my system (still awaiting parts so I can start up my mainsystem again
).I neither had any problem with GTZ's mounting, but I always grease threads like that slightly with silicon tim.
Last edited:
My block too got some anomalies in the pin matrix, they are in the outer zone - so I don't belive at all that this will affect the blocks performance.
As from I can tell from my 3 days testeing with the block it easily outperforms GTZ and Heatkiller (which I also have both of). During my short tests I am not really sure if I got burnt in the tim enough either - so I might get lower temperatures when I will be finished reconfiguring my system (still awaiting parts so I can start up my mainsystem again
I neither had any problem with GTZ's mounting, but I always grease threads like that slightly with silicon tim.
I'd be interested in seeing those results and good luck. And yes, letting the TIM set up makes a difference. MX-2 does improve around 0.5-1.0c over several hours depending on the thickness. I like letting my new mounts set up overnight for a least 12 hours before I gather my results.
andyc
See my post here:
http://www.xtremesystems.org/forums/showthread.php?t=236172&page=11
I encountered the same thing. Gabe says there's no performance impact, but I admit I'm a bit disappointed with the quality (or QC on it). In any case, I got a RMA from Swiftech for the base plate, and the new one looks a lot better.
http://www.xtremesystems.org/forums/showthread.php?t=236172&page=11
I encountered the same thing. Gabe says there's no performance impact, but I admit I'm a bit disappointed with the quality (or QC on it). In any case, I got a RMA from Swiftech for the base plate, and the new one looks a lot better.
What in God's name are you spreading all over the inside of that waterblock????
What the hell??? It looks like shit on a stick. I can't imagine that is any good for the system as a whole, but what do I know.
Agreed,
If you want to try something like creating a bow on a block such as the HK-3, or increasing a blocks bow, there's much easier and cleaner ways of doing it than smearing caulk all over the place. That's just asking for trouble. Some people assume to much without the actual experience it takes to make a substantiated claim, and that's when the trouble starts
andyc
What in God's name are you spreading all over the inside of that waterblock????
What the hell??? It looks like shit on a stick. I can't imagine that is any good for the system as a whole, but what do I know.
hehe - yes I was very sceptical myself
So before I did put it in my system I tested it in my testrig with a 133watts peltier attached to the waterblock.
The idea is that this mod will hinder waterleakage over the micropins - and I can assure you that it does really work.
That caulk is soft and stretchable and I am 100% sure it wont come loose.
Agreed,
If you want to try something like creating a bow on a block such as the HK-3, or increasing a blocks bow, there's much easier and cleaner ways of doing it than smearing caulk all over the place. That's just asking for trouble. Some people assume to much without the actual experience it takes to make a substantiated claim, and that's when the trouble starts
andyc
I have been watercooling / phasecooling for some ~10 years - and modded the shit out of "everything". Even made my own waterblocks / and built phasecooler systems - in other words I am not afraid trying new things if it gives me better performance aso.
Last edited:
Main, are you trying to pass this testing off as your own? WTF
You're not the tester....I am (even when I signed up a few years ago, Vapor was taken, meh).
______________
As for why the mod works on the XT, I'm really not sure why it works. When I disassembled the block, the pins went all the way through the silicone (I did a very thin layer--you could still see the o-ring through it), so I don't think much flow goes 'over' the pins. I think it may just increase the velocity of the fluid at the bottom of the pins and creates a slightly different bow--the modification visibly deforms the base on both the XT and the HK3.0. It's also more effective on the HK3.0 and the microchannel's fins don't go through the silicone at all.
Anyway, you should have used a straight edge to wipe away the silicone (can be caulk or RTV or any variety really, just as long as it's 100% silicone) and form a better and thinner surface. If nothing else, a really thick layer will increase restriction a ton.
HK3.0 Cu is in testing right now using the same base from my LT block
You're not the tester....I am (even when I signed up a few years ago, Vapor was taken, meh).
______________
As for why the mod works on the XT, I'm really not sure why it works. When I disassembled the block, the pins went all the way through the silicone (I did a very thin layer--you could still see the o-ring through it), so I don't think much flow goes 'over' the pins. I think it may just increase the velocity of the fluid at the bottom of the pins and creates a slightly different bow--the modification visibly deforms the base on both the XT and the HK3.0. It's also more effective on the HK3.0 and the microchannel's fins don't go through the silicone at all.
Anyway, you should have used a straight edge to wipe away the silicone (can be caulk or RTV or any variety really, just as long as it's 100% silicone) and form a better and thinner surface. If nothing else, a really thick layer will increase restriction a ton.
HK3.0 Cu is in testing right now using the same base from my LT block
Main, are you trying to pass this testing off as your own? WTF
You're not the tester....I am (even when I signed up a few years ago, Vapor was taken, meh).
______________
As for why the mod works on the XT, I'm really not sure why it works. When I disassembled the block, the pins went all the way through the silicone (I did a very thin layer--you could still see the o-ring through it), so I don't think much flow goes 'over' the pins. I think it may just increase the velocity of the fluid at the bottom of the pins and creates a slightly different bow--the modification visibly deforms the base on both the XT and the HK3.0. It's also more effective on the HK3.0 and the microchannel's fins don't go through the silicone at all.
Anyway, you should have used a straight edge to wipe away the silicone (can be caulk or RTV or any variety really, just as long as it's 100% silicone) and form a better and thinner surface. If nothing else, a really thick layer will increase restriction a ton.
HK3.0 Cu is in testing right now using the same base from my LT block
wtf ??? Have I ever said anywhere that I did that test I linked to ????
I linked to Koolance yesterday, does that make it look in your book like I have made the plate heat exchangers ??
On the other hand I do also test waterblocks (amongst a lot of other things) - I have also been heavy involved in the development of waterblocks for a commercial producer. I am also "betatester" for various producers (both software and hardware) - but I don't write reviews anymore since it is too little money in beeing a reviewer.
Don't tell me how to apply the silicone either - what I have done really work perfect. I did measure the space with use of clay before putting on the caulk. And I have reopened the block of course - twice now - to check that the pins aren't bent.
FYI: This "caulk" I have used is very soft and of the sanitary type - meant to be used in moist/wet conditions and are therefore treated so no fungus will start developing. I cant see that you have used a caulk like I have done - if so - you most certainly will end up getting problems.
Likewise putting on too much caulk will cause the baseplate to bend as you say - I became aware of that
Anyway this mod is great (but looks like shit - but frankly who cares at long as it works ?), other mods I did is to put some spacers under the springs on the fasteningbolts to achieve a bit better pressure since my cpu's are extremely lapped. Here too one have to be careful since the tiny 1/8" (or 3mm ? not sure since I don't have a caliper here just now) threads cant take that much load.
I am sorry if you felt like I was taking credit for your test - that was really not the meaning. English is not my native language - and in addittion I speak/write 7 other languages (approx. on the same level as english) so my head is full of words
and misunderstandings because of my way of putting words toghether can and will happen.Main, are you trying to pass this testing off as your own? WTF
You're not the tester....I am (even when I signed up a few years ago, Vapor was taken, meh).
______________
As for why the mod works on the XT, I'm really not sure why it works. When I disassembled the block, the pins went all the way through the silicone (I did a very thin layer--you could still see the o-ring through it), so I don't think much flow goes 'over' the pins. I think it may just increase the velocity of the fluid at the bottom of the pins and creates a slightly different bow--the modification visibly deforms the base on both the XT and the HK3.0. It's also more effective on the HK3.0 and the microchannel's fins don't go through the silicone at all.
Anyway, you should have used a straight edge to wipe away the silicone (can be caulk or RTV or any variety really, just as long as it's 100% silicone) and form a better and thinner surface. If nothing else, a really thick layer will increase restriction a ton.
HK3.0 Cu is in testing right now using the same base from my LT block
LMAO...I actually thought that was you in the OP Vapor. Glad you straightened that out, and how's it going
andyc
hehe - yes I was very sceptical myself
So before I did put it in my system I tested it in my testrig with a 133watts peltier attached to the waterblock.
The idea is that this mod will hinder waterleakage over the micropins - and I can assure you that it does really work.
That caulk is soft and stretchable and I am 100% sure it wont come loose.
I have been watercooling / phasecooling for some ~10 years - and modded the shit out of "everything". Even made my own waterblocks / and built phasecooler systems - in other words I am not afraid trying new things if it gives me better performance aso.
That's cool..............but it still looks like a baboon's ass............
That's cool..............but it still looks like a baboon's ass............
hehe yeah - I know... but let's keep it a secret - noone is gonna see it
Well,
After complaining about the hacked up base I received on my retail XT, Gabe was kind enough to replace it with this.
They got some serious QC issues with that base. Large pic to show milling defects.
Needless to say, I was expecting something a little different:rofl:
I agreed to retest and see if the new base would help the XT I have catch up to the HK-3, but the new one I received today looks as bad as the one I have on the XT now. Funny thing is, both bases are hacked up in the same way, and in the same exact place on the pin matrix.
I guess they can say it doesn't hurt performance, but something's either produced correctly with care, or it's not. I don;t buy it either way to be honest. I'll retest and see if I can get the XT with the new base to catch up to the HK-3 which I show 0.7c better than the XT.
andyc
After complaining about the hacked up base I received on my retail XT, Gabe was kind enough to replace it with this.
They got some serious QC issues with that base. Large pic to show milling defects.
Needless to say, I was expecting something a little different:rofl:
I agreed to retest and see if the new base would help the XT I have catch up to the HK-3, but the new one I received today looks as bad as the one I have on the XT now. Funny thing is, both bases are hacked up in the same way, and in the same exact place on the pin matrix.
I guess they can say it doesn't hurt performance, but something's either produced correctly with care, or it's not. I don;t buy it either way to be honest. I'll retest and see if I can get the XT with the new base to catch up to the HK-3 which I show 0.7c better than the XT.
andyc
Kaldskryke
[H]ard|Gawd
- Joined
- Aug 1, 2004
- Messages
- 1,346
That definitely doesn't look too pretty. I guess I'm a little surprised because the base on my old Apogee GT looks so well made by comparison.
Yep..something's going on and quality has dropped way off from the base on my GTZ. I realize the pins are cut deeper, but still no excuse. With the proper equipment, attention to detail and overall willingness to create something of quality, not that tuff. And it's not like the XT is a cheap block.
andyc
Gabriel Rouchon
n00b
- Joined
- Mar 4, 2003
- Messages
- 12
Sadly, I have to agree with you with respect to cosmetics of the replacement plate that we sent you. This is a plate left from our first production run. Two people are involved in sending you this, and I severely reprimended them.
I am sending you this new plate coming from our latest production run, received this week:
I already said so before, but let's be cristal clear again: there will not be any difference in performance. It's purely cosmetic.
For the record, and contrarily to what you are asserting, if there is a Company on the market that cares for quality and attention to detail, and has proven so over and over again in the past 10 years, it is Swiftech. On the other hand, people and tools are not perfect, and constant vigilance is needed to keep us on track. This new pin matrix has been a technological challenge, and comparing it to the GTZ only shows a lack of understanding of this manufacturing process on your part. I can't blame you for that, after all, we certainly don't go about advertising how we do things. Suffice to say that it is a difficult process, and that practice makes perfect.
Now talking about performance, I'd like to know the date at which you conducted your tests with your HK. What I am getting at is this: when you conduct comparative testing, do you validate the performance recorded earlier by running a baseline test? Typically it is always preferable to conduct your comparative testing between blocks back to back.
Sadly, I have to agree with you with respect to cosmetics of the replacement plate that we sent you. This is a plate left from our first production run. Two people are involved in sending you this, and I severely reprimended them.
I am sending you this new plate coming from our latest production run, received this week:
I already said so before, but let's be cristal clear again: there will not be any difference in performance. It's purely cosmetic.
For the record, and contrarily to what you are asserting, if there is a Company on the market that cares for quality and attention to detail, and has proven so over and over again in the past 10 years, it is Swiftech. On the other hand, people and tools are not perfect, and constant vigilance is needed to keep us on track. This new pin matrix has been a technological challenge, and comparing it to the GTZ only shows a lack of understanding of this manufacturing process on your part. I can't blame you for that, after all, we certainly don't go about advertising how we do things. Suffice to say that it is a difficult process, and that practice makes perfect.
Now talking about performance, I'd like to know the date at which you conducted your tests with your HK. What I am getting at is this: when you conduct comparative testing, do you validate the performance recorded earlier by running a baseline test? Typically it is always preferable to conduct your comparative testing between blocks back to back.
Thanks and look forward to the new base. I just saw Michelle's email for tracking. Multi mounts, sets and data pulls for well over 5 mounts on both blocks done during the same time periods. I was so thrown by the discrepancy I saw with my results from what others have shown, that I actually did extra mounts. I'm well aware of the sound methodology it takes to preform comparative testing I deem critical (ie, "best used" as opposed to "as sold" for mounting purposes using my own preference of springs, mounting pressure, back plates, etc for the HK-3 CU), and I applied that methodology to both the HK-3 and the XT when I tested.
I look forward to getting a clean base and retesting,
andyc
DogsofJune
Supreme [H]ardness
- Joined
- Nov 7, 2008
- Messages
- 4,638
Wow.....
Any 100% silicone caulk will do, available from any home improvement store
Just got my XT today, but theres something wrong here. Can anyone tell what that is?
PS: I am totally pissed off!
The base is hacked, oh wait there is a missing screw LOL. So much comparison between this POS and the HK 3.0 yet the HK 3.0 beats this crappy cast block in workmanships by several miles. Sadly i don't see the price difference.
On the flip side, you gonna be hard pressed getting tech support from HK in the US while Swifty is here and goes out of the way to keep you satisfied. I assume you already contacted Swifty? And how easy would be it get in touch with someone from HK?
Swiftech makes crappy junk, i've had 2 GTZ backplates failed on me and didn't bother to get a 3rd one. I had to use another backplate from some 3rd party heatsink that happanened to have m3 screws. Yes great support but doesn't help since they keep sending you junk like what happened to Andy C.
Do i care about the inexistent support of HK 3.0? sure i do but hey the block is quality machined from a forged block of copper instead of some cheap cast blank from china then trying to market it like the big boys.
I will have no problem paying $50 for the XT but $80 is a joke at best. Then here comes Gabe from swiftech saying "oh as long as it passes our QC is good to go since it shouldn't affect temps" and we see shitty bases and missing screws. What a joke of QC and who is Gabe to tell me if i should be satisfied with my purchase or not.
You guys are right about the workmanship. I can't believe it was missing a screw! I am so fn pissed off! WS going to be my weekend project) guess in ordering angled compression fittings for my heatkiller that's sitting in my draw! Thought my XT would have better temps but am afraid of leaks especially when ghetes a screw missing!
Well, if it's crappy junk, why do you keep buying it?
No shit! i should of known better specially after realizing gabe has been cooking the numbers to favor his own blocks
now i call that a good marketing ploy. I will not buy asian made junk at high end german machined prices that is a no no