Cooling Editor's Desk Reviews Watercooling

10 CPU water blocks tested on socket LGA1700 and AM5 in a big roundup including a necessary reappraisal

EK-Quantum Magnitude AM4/5?

As EK Waterblocks also offers the Quantum Magnitude for current sockets in addition to the Quantum Velocity 2, I also wanted to include it in our test round. However, finding out that the waterblock even exists was the first challenge. Because if you filter for LGA1700 or AM5 in the EK Shop, you only get Quantum Velocity blocks to choose from. Magnitude blocks can only be found without filtering or by using the search.

As these are quite a bit more expensive than the Velocity 2 offshoots, I naturally didn’t want to buy directly from EK, but first compared prices here in Germany at various web stores. So what do you do? You copy the EAN and throw it into geizhals.de. The block is actually available at a slightly lower price in various online stores, except that the product description says AM4. But the EAN is identical to the EK store and the EK store also has AM4 in the URL. So it must be the same block and only the label will have changed, no? Exactly, no.

After the block arrived, I discovered that the standoffs supplied did not fit the AM5 socket at all. However, the online instructions state that the standoffs should be used with the stock AM5 backplate. But a backplate is included in the scope of delivery. Huh? There is also no description of the package contents in the instructions, so the confusion was complete. Only now did it slowly dawn on me that the difference between the AM4 and AM5 block was probably more than just the sticker on the packaging. If you search for AM5 standoff in the EK store, you will also find a UNC 6/32 standoff kit (EAN 3831109901076) with the correct parts for around 4 euros. This means that Magnitude water blocks for AM4 and AM5 must differentiate between these standoffs, otherwise an AM5 upgrade kit would have to contain more.

Ergo, Magnitude AM4 blocks are not the same products as the AM5 variant and the AM4 variant is not compatible with AM5 without the additional standoff kit. However, the fact that the EAN is the same in each case is highly confusing. Incidentally, this applies to every flavor of the AM4/AM5 Magnitude, whether nickel-acetal, nickel-plexi etc., EK has simply recycled the EAN. According to my understanding of EAN/GTIN numbers, this should not be the case. Feel free to correct me in the forum thread if you are familiar with this.

But now that we have the block and the necessary standoff kit for around 220 euros, what do you get for it? Included is the block itself, an additional backplate made of punched sheet metal, which is probably intended for AM4 and an associated rubber insulation layer, which seems to be a standard part that also has the holes for Intel LGA1xx. There are also 4 sets of screws with M3 thread and H2.5 hexagon socket head, springs and (matching) standoffs. There is also an additional jet plate that can be swapped for the already installed one for better performance for each CPU type, a tube of Thermalgrizzly TIM and a wrench that fits the screws.

Well, relatively suitable. Something seems to have gone wrong in the production or anodization of the screws, because the screw heads are very flat, so that the screwdriver can only transmit very limited force. One screw is not anodized on the inside at all and this one completely tore out after the second assembly, so I only managed to get it loose again by applying a lot of pressure to the head. Too bad, also because 3 screws in black and one in brass looks not great.

Apart from that, mounting on AM5 is easy. The standoffs just go into the threads of the standard AM5 backplate, the block is placed on top and then fixed with springs and the screws. It is interesting to note that EK states in the instructions that the screws should only be tightened until their head is even with the hole in the frame of the water block and no further. The cable for the integrated RGB lighting has a high-quality appearance, is easy to bend and position and has a direct 3-pin connection. The lighting is also extremely bright and powerful, making the Trident Z5 RGB RAM modules look old in comparison – wow!

Conversion to LGA 1700

So what do you have to do to convert the AM4/AM5 Magnitude to LGA1700? There is actually a conversion kit, but you can only find it in the EK Shop if you search for “1700 upgrade” or don’t search at all or filter and simply browse through all waterblock accessory items. If you search for “LGA1700” or “LGA1700 Upgrade” or filter for “Intel” in the accessories section, you won’t find the kit. Incidentally, the kit costs 130 euros. So if you want a Quantum Magnitude water block that is compatible with LGA1700 and AM5, you will have to invest around 350 euros.

The kit includes the frame for LGA1700 with 78 x 78 mm hole spacing, a matching backplate made of solid steel including a rubber insulating layer, 4 sets of screws with M3 thread and H2.5 hexagon socket head, springs, standoffs and neoprene washers. Also included is another “insert” that guides the water from the inlet connection to the jetplate, the matching O-rings, 2 jetplates, each 0.5 mm and 0.6 mm thick, and a screwdriver. There are no instructions either in paper form or on the website. So you more or less have to figure out the conversion yourself. Alternatively, you can also refer to the instructions for a Magnitude LGA1700 block. Incidentally, the scope of delivery or “What is enclosed” is included here, in contrast to the instructions for the AM5 block.

Actually, only the 4 H2.5 hexagon socket screws on the underside of the base plate need to be loosened. The base plate can then be removed and the remaining inner parts taken out. The RGB strip is also only loose in the block cover and the cable is only threaded through a hole in the frame. Now we simply do the whole thing backwards with the Intel innards. First, the insert with the O-rings is inserted on both sides. Due to the slightly rectangular shape of the block, it can only be inserted rotated by 180 degrees. According to the instructions, EK always guarantees the best possible performance this way. A jetplate is then placed on the insert. Here you can choose between the 0.5C for rather flat CPUs and 0.6C for rather concave CPUs. This modularity is really practical if you want to optimize the block for your own individual CPU, e.g. if it is particularly concave or if you have ground it flat. Unfortunately, the jetplates are not deburred, so that a metal burr of up to 0.8 mm protrudes below the labeling.

Now the RGB strip is threaded through the frame, the frame is slipped over the cover and then the base plate is replaced. It is important that the seal around the cover is in the correct position and that the base plate is correctly aligned. EK has engraved corresponding markings on the base plate and frame for this purpose. Now simply reinsert the screws and tighten them carefully, ideally crosswise, so that the block is tight again. If you were clever, you could also thread the RGB lighting cable through the hole in the frame that is closest to the connection on the mainboard. But I wasn’t.

Before the block is attached to the mainboard, it should of course always be checked for leaks. This also paid off here, as the block was leaking slightly even though all the screws were fully tightened. Disassembling the jetplate again and turning it 180 degrees helped. I can’t say for sure whether this really had anything to do with it or whether it was just a seal that had slipped out of place. But I have the feeling that the protruding burr on the jetplate had something to do with the leak.

The assembly on the LGA1700 then follows the same procedure. First, the backplate is positioned and held in place on the back of the mainboard, which is not exactly easy with the additional rubber insulation, and then the standoffs are screwed in from the front with a neoprene washer. Finally, the block is placed back on top and fixed in place with springs and screws. Here you should also stop turning when the screw heads are level with the holes in the frame. I tried to measure this with my torque wrench, but it was less than its minimum measured value of 0.3 Nm.

But what happens if you tighten the screws further? I would have liked to test it on AM5 as well, but since one of the screws there was defective after 2 assemblies, this comparison is only available for Intel. I then tightened it to 0.6 Nm, as EK recommends in the instructions for its Velocity 2 block. And I can already spoil it: this alone brings 5 – 8 K lower temperatures!

We contacted EK in advance, told them about our findings and observations and asked for feedback. They want to fix the problems with the search and filtering in the store as soon as possible, as well as the missing sections in the Magnitude AM5 instructions. EK justifies the fact that tightening the screws according to the instructions has a significant impact on performance by saying that they want to stay within the Intel specifications for the LGA1700 socket. I quote: “We are doing our best to follow Intel’s mechanical loading specification requirements for the socket, while competitors are hard mounting water blocks. Yes, you can get more performance with more pressure, but you can also get a PC that doesn’t boot. Fundamentally, Magnitude does not have a hard stop, and it can be overtightened and when the head is flush that equates to Intel spec. For the “best user experience” it’s recommended to keep the mounting screws flush, just as stated in the manual.”

This is a valid point and possibly also a shortcoming of other water blocks. Today, we can only evaluate the user-friendliness and measure the cooling performance. The Intel specifications for the cooler contact pressure for LGA1700 and whether a block exceeds them is another matter. However, as the Velocity 2 block also recommends 4 M3 threads with 0.6 Nm, it can be assumed that this value should also be suitable for the Magnitude. However, according to EK, the Magnitude block has also been on the market for some time and the 0.6 Nm torque screwdriver was not yet available at the time of EK’s release. Accordingly, the instructions had to use a different reference point at the time.

We also questioned the high price of the Magnitude Block relative to its performance. EK pointed out to us that cooling performance is not the only metric by which a water block should be judged, and here we clearly agree. Factors such as manufacturing quality, flexibility, user-friendliness and, of course, RGB lighting are also important factors that should not be neglected. For example, many parts of the Magnitude blocks are CNC-milled from solid aluminum blocks for a corresponding quality, look and feel. Of course, this is also an important feature of the block. However, the quality of the other components must also match this, and in the case of the screws and the jet plate, at least on my block, this was not always at a level that justified the purchase price, at least in my opinion.

Finally, EK pointed out to us that Magnitude and Velocity 2 blocks have been on the market for some time and that the competition has of course had time to come up with better products since then. This is also a fair point, but we just tested the blocks that are currently on the market. That is also fair. We will of course also be happy to test future water block designs from EK as soon as they are released and of course only if you like my tests.

manual_am5_EK-IM-3831109819258

 

manual_lga1700_magnitude_EK-IM-3831109861813

 

EK Water Blocks Quantum Line EK-Quantum Magnitude D-RGB, AM4, Nickel Acetal (3831109819258)

CSV-Direct.deLagernd im Versandlager - Lieferzeit 2-5 Werktage209,39 €*Stand: 17.05.24 17:26
Proshop.deVerfügbar - 5-7 Werktage Lieferzeit209,89 €*Stand: 17.05.24 14:24
Caseking.deLagernd209,90 €*Stand: 17.05.24 17:29
*Alle Preise inkl. gesetzl. MwSt zzgl. Versandkosten und ggf. Nachnahmegebühren, wenn nicht anders beschriebenmit freundlicher Unterstützung von geizhals.de

EK Water Blocks Quantum Line EK-Quantum Magnitude UNC Standoffs, AM5, Schraubenset 6-32 UNC, 4 Stück, schwarz (3831109901076)

Procooling1-3 Werktage2,90 €*Stand: 17.05.24 14:11
Caseking.deLagernd2,90 €*Stand: 17.05.24 17:29
AlternateNicht lagernd, ab Bestellung versandfertig in 11 Tagen2,99 €*Stand: 17.05.24 17:19
*Alle Preise inkl. gesetzl. MwSt zzgl. Versandkosten und ggf. Nachnahmegebühren, wenn nicht anders beschriebenmit freundlicher Unterstützung von geizhals.de

EK Water Blocks Quantum Line EK-Quantum Magnitude D-RGB, 1700, Nickel Acetal (3831109861813)

Procooling1-3 Werktage249,90 €*Stand: 17.05.24 14:11
Caseking.deLagernd249,90 €*Stand: 17.05.24 17:29
AlternateNicht lagernd, ab Bestellung versandfertig in 11 Tagen249,90 €*Stand: 17.05.24 17:19
*Alle Preise inkl. gesetzl. MwSt zzgl. Versandkosten und ggf. Nachnahmegebühren, wenn nicht anders beschriebenmit freundlicher Unterstützung von geizhals.de

EK Water Blocks EK-Quantum Magnitude Frame 1700 silber, Montagekit (3831109897003)

Caseking.deLagernd119,90 €*Stand: 17.05.24 17:29
Jacob Elektronik direktLieferzeit 3-5 Werktage129,89 €*Stand: 17.05.24 17:22
Procooling1-3 Werktage129,90 €*Stand: 17.05.24 14:11
*Alle Preise inkl. gesetzl. MwSt zzgl. Versandkosten und ggf. Nachnahmegebühren, wenn nicht anders beschriebenmit freundlicher Unterstützung von geizhals.de

 

114 Antworten

Kommentar

Lade neue Kommentare

ipat66

Urgestein

1,367 Kommentare 1,366 Likes

Sauber !!!
Tolle Arbeit Xaver :)

Antwort 6 Likes

s
scotch

Veteran

153 Kommentare 103 Likes

Super Arbeit. Ich nutze auf AM5 den Wasserblock von NTech. Qualitativ und von der Montage ist das für mich das beste Gesamtpaket für den Preis. Von der Kühlleistung und vom Flow gibt es auch nichts zu meckern.

Antwort 3 Likes

FritzHunter01

Moderator

1,159 Kommentare 1,575 Likes

Hey Xaver,

danke für den Einsatz!

Grüße

Antwort 4 Likes

ApolloX

Urgestein

1,687 Kommentare 955 Likes

Endlich Mal einer, der sich in der Lage sieht, den Core 1 auf AMD zu testen.
Ansonsten schaut's auch sehr fundiert und reflektiert aus.
Danke Xaver für den Aufwand!

Antwort 5 Likes

F
Falcon

Veteran

115 Kommentare 118 Likes

Top Test!
Und nein, dein Qualitätsempfinden ist in Ordnung. (y)

Wen ich soviel Geld für nen Block ausgebe kann ich auch Erwarten das er passt, ich keine Gewinde Nachschneiden muss, sich der Block nicht bis kurz vor die Selbstzerstörung verbiegt und so weiter.

Antwort 4 Likes

Opa-Chris

Mitglied

85 Kommentare 120 Likes

Top Arbeit und Aufarbeitung. Vielen Dank!

Antwort 1 Like

S
S.nase

Urgestein

1,357 Kommentare 455 Likes

Mir fehlt ja immer noch der Glaube daran, das man nur anhand des Losbrechmoments eines normalen Gewindes, die Kompressionsrate einer Schraubenfeder exakt einstellen kann. Bei meinen Versuchen hat das selbst mit einem feinen Trapezgewinde(Schraube und Mutter aus Edelstahl) nicht wirklich exakt funktioniert. So das am Ende doch wieder die klassische Methode mit dem Nachmessen per Schiebelehre an allen vier Schraubenfedern nötig war.

Kann natürlich auch sein, das es an meinem zu ungenauen Drehmomentschlüssel lag. Daher meine Frage: ob du vorher mal getestet hast, wie genau sich die Anpresskräfte mit der DrehmomentMethode wiederholbar an den Schraubenfedern einstellen lassen. Denn das ist mMn schon sehr wichtig, damit der Kühler gleichmäßig aufliegen kann.

Antwort 1 Like

echolot

Urgestein

977 Kommentare 749 Likes

Prima Test...ich hoffe die Werte für den Core 1 sind jetzt akkurat :D ...wird dann mein nächster Kühler.

Antwort 1 Like

komatös

Veteran

104 Kommentare 78 Likes

Danke schön für die investierte Zeit und das doch ernüchternde Fazit. Für mich zeigt sich einmal mehr, dass der Massenfertigungswahnsinn in Asien, die Qualität der Produkte verschlechtert. Oder wie soll man sonst die Fertigungsmiesere mit zu kleinen Löchern und überstehenden Schraubenköpfen etc. nennen?
Auch wenn das eine oder andere Ergebnis in diesem Test, Soft- und/oder Hardwarebedingt, Tolleranzen hinterläßt, kann man diese als valide ansehen.
Also noch einmal danke schön für die zeitintensive Aufarbeitung und Neubewertung.

Antwort 2 Likes

RedF

Urgestein

4,704 Kommentare 2,576 Likes

Danke für den Test.

Was bei mir allerdings anders aus sah ist die Durchfluss Rate Heatkiller vs Core1.

Bei mir war der Heatkiller 50l/h restriktiver wie der Core1.

Habe auch den highflow next.

Und jetzt weiß ich auch warum ich beim loggen mit HWInfo unterschiedlich viele Datenpunkte habe ^^.

Antwort 1 Like

skullbringer

Veteran

306 Kommentare 329 Likes

Die Länge der Schrauben bzw. Überstand zu Messen setzt ja auch vorraus, dass sich die Schrauben nicht längen bzw. sich alle 4 gleichmäßig längen, oder?

Um es wirklich genau zu machen, müsste man also wirklich die Anpresskraft auf der CPU messen und das notwenige Equipment dafür habe ich stand heute leider nicht.

Also ja, du hast Recht. Wir montieren "blind" nach Vorgaben des Herstellers und vertrauen darauf, dass die Schrauben alle gleich sind bzw. das Drehmoment auch zum gewünschten Anpressdruck führt. So machen es ja auch die allermeisten Endanwender und somit ist es auch im Interesse des Wasserblock-Herstellers, dass die Drehmomente zum korrekten Ergebnis führen.

Angezogen wird zudem über Kreuz, schrittweise und in gleichmäßigen zügen, um "stick and slip" zu vermeiden. An einer einzelnen Schraube mit festem Anschlag habe ich auch mal die Wiederholgenauigkeit meines Amazon-China-Drehmoment-Schlüssels getestet und bis auf 0,02 Nm schien das reproduzierbar zu sein. Aber mit 4 Schrauben kann sich der Block ja auch verkannten, evtl. sind nicht alle Schrauben gleich gefertigt oder die Wärmeleitpaste ist auf einer Seite der CPU dicker aufgetragen, sodass eine Schraube mal mehr Kraft transportieren muss. Alles ziemlich viel Blindflug :D

Ein Test der Schrauben, Anpressdrücke, Wiederholgenauigkeit wäre aber definitiv ein Thema für einen extra Artikel. Danke für den Input! :)

Antwort 2 Likes

Klicke zum Ausklappem
skullbringer

Veteran

306 Kommentare 329 Likes

auch Heatkiller IV Pro oder ein leicht anderes Modell?

Antwort Gefällt mir

N
NilsHG

Mitglied

86 Kommentare 52 Likes

Großartiger Test, vielen Dank dafür!
Ich habe meine Kühler von AquaComputer und Sockel 1155 auf AM4 umgerüstet. Sollte es kein Umrüstkit für den nächsten Sockel, den ich verwenden werde, geben, kann man sich hier sehr gut orientieren. Die größte Verbesserung meiner Wakü in den vergangenen Jahren war die Erweiterung des Kreislaufs auf einen externen Mora 420. Damit steigen die Wassertemperaturen trotz neuer GPU nicht mehr über 34°C. ;-)

Antwort Gefällt mir

RedF

Urgestein

4,704 Kommentare 2,576 Likes

Gerade die Rechnung rausgesucht.
HEATKILLER® IV PRO (AMD, AM4 ready) ACRYL
Bestellt am 5.10.2020

Habe ihm in der ganzen Zeit allerdings nicht gereinigt, vielleicht liegt es daran.

Habe eine laing DDC 1T Plus und eine Grafikkarte vor der CPU im Kreislauf.

Antwort Gefällt mir

HerrRossi

Urgestein

6,789 Kommentare 2,246 Likes

Vielen Dank für den Test, super Arbeit und kommt mir gerade recht.

Der Core 1 ist ja wirklich gut und gefällt mir auch vom Design her gut. Aber 20 EUR für 4 Schrauben und 2 Plastikteile (aka Performance Kit) ist echt frech, auch bei dem recht geringen Unterschied.

Antwort 3 Likes

Schnuffel

Veteran

185 Kommentare 153 Likes

Damals mit 2 oder 4 Kernen konnte man recht einfach unter Last vorsichtig die Verschraubungen lösen bzw fester drehen und dabei die Temperaturen beobachtet. Fingerspitzengefühl vorausgesetzt. Scheint inzwischen kompliziert.
Federn vertraue ich grundsätzlich nicht. Wer schon einmal einen Zylinderkopf eingemessen hat (Ventilfedern) der kennt die Serienstreuung - im kleinen wie im großen halt.
Guter Artikel übrigens @skullbringer 👍

Antwort 3 Likes

Megaone

Urgestein

1,757 Kommentare 1,652 Likes

Was soll man sagen? Du hast einen Top-Job gemacht. Danke für die Arbeit.

Antwort 5 Likes

echolot

Urgestein

977 Kommentare 749 Likes

Ja, sie nehmen es von den Lebenden. Aber bedenke wie lange Du daran hast. I.d.R. mehrfach einsetzbar und immer noch besser als jede Lukü.

Antwort 1 Like

N
NilsHG

Mitglied

86 Kommentare 52 Likes

Sofern es weiterhin Mountingkits für neue Sockel geben wird, ist das absolut richtig. An der Wakü Hardware ist i.d.R. kein Vergang. Ggf. werden Schläuche, Dichtungen, Kühlmittel gewechselt, aber Kühlblöcke/Radiatoren halten ewig. Beim Neukauf orientiere ich mich dann gerne an solchen Tests. Bei dem was eine Wakü zusätzlich zur PC Hardware kostet, spielen 20 € (für mich) auch keine Rolle mehr.

Antwort 3 Likes

Danke für die Spende



Du fandest, der Beitrag war interessant und möchtest uns unterstützen? Klasse!

Hier erfährst Du, wie: Hier spenden.

Hier kannst Du per PayPal spenden.

About the author

Xaver Amberger (skullbringer)

Werbung

Werbung