Corsair Hydro Series H100i AIO CPU Cooler

Posted: March 13, 2013 in Kits
Tags: , , ,

Welcome to the fourth kit in my value <$150 2x120mm radiator kit series, the Corsair Hydro Series H100i CPU cooler. Corsair has been in the sealed AIO cooler business for a while and although I personally have never had the opportunity to review any of the kits before, I’ve seen the forum following grow over the years so I expect good things. A good following and user base is no accident and usually means good things. I also very much enjoy my Corsair TX650 power supply and regularly run Corsair memory such as my current Vengence memory due to the quality of parts and consistent good result I’ve had with Corsair products. They also have a good reputation for customer service and honoring warranty issues and that means a lot to me as well. When I decided I wanted to do kit testing, the H100i was on my short must do list due to its popularity and amazing low price of just $119. While the H100i doesn’t come designed for expansion due to it’s sealed nature, it is the lowest priced kit with a 240 sized radiator and had the longest warranty so I was very interested in reviewing.


Here is a quick picture of the kit

Frozen CPU Product Information

Product Description

Have a PC case that supports a top-mounted 240mm radiator? Take your CPU cooling to a new level. H100i starts with the advanced design of the Hydro Series H80i, and adds a double-wide radiator for even better performance.

Monitor temperature and control lighting and fan speed on your screen. No additional hardware is necessary using the Corsair Link


Self-contained Cooling System

  • Hydro Series H100i comes pre-filled, and never needs refilling or priming.

Dual Radiator with Custom Fan Design

  • The 240mm top-mounted radiator provides maximum surface area for maximum cooling power. The 120mm fans use wide, low-pitch blades for better static pressure to noise ratio, offering improved efficiency at lower noise levels.

Built-In Corsair Link

  • No additional hardware is necessary – just connect the included Corsair Link cable to a USB header on your motherboard and download the free Corsair Link Dashboard software. You can monitor coolant temperature and adjust cooling performance directly from your desktop.

Tool-free multi platform magnetic mounting bracket kit

  • The modular design makes installation simpler, and it’s compatible with Intel and AMD processors.

Large-diameter, low permeability tubing

  • Minimal coolant evaporation helps ensure long life, and the resilient material offers both high flexibility and excellent leak protection.

Technical Specifications

Technical Specifications
Radiator dimensions: 120mm x 275mm x 27mm
Fan dimensions: 120mm x 120mm x 25mm
Fan speed: 2700 RPM
Fan airflow: 77 CFM
Fan dBA: 37.68 dBA
Fan static pressure: 4mm/H20
Compatibility: Intel™ LGA 1155, 1156, 1366, and 2011
AMD™ sockets FM1, FM2, AM2, and AM3
Hydro Series H100i requires a case with dual 120mm fan mounts with 15mm spacing for a 240mm radiator
Package contents: Corsair Hydro Series H100i High Performance Liquid CPU Cooler
Magnetic Multi-platform mounting kit for all modern CPU sockets
Dual SP120L High Performance Fans
USB Cable for Corsair Link™ Compatibility
Fan and radiator mounting screws
Thermal compound (pre-applied)
Quick Start Guide

So as already noted, the 5 year warranty stands out for me as does the 2700RPM fans.

What it doesn’t specify and what I found out during the review is the lack of Window 8 support in using the Corsair Link software when I did my testing on this unit.

I did just double-check the corsair website and now that I have removed the H100i for other unit testing, it does appear that corsair link is now finally 5 months later, supporting windows 8. If you go to the downloads section, you will now see that the latest software is supposedly supporting Win 8. That has recently been updated.

Also worthy of note for those thinking of modding the kit, the radiator is “aluminum” so any sort of mods will require a corrosion inhibitor because of the mixed copper/aluminum metals.

Pages: 1 2 3 4 5 6 7 8 9 10

  1. Very cool and comprehensive review yet again from the Best! After reading this I feel like I own an H100i, yet it’s not my cup of tea. You know there are some things I like about this product and some things I do not and other that don’t matter. You are a reviewer and gave it a great honest look at, I can only relate from a user POV. Perhaps I don’t explain that right but who cares.

    Just want to comment on some things here. You say the pump is 1.2 Watts, for some reason I am too stupid to find that information from Corsair. The fan speed range is pretty high to comply with the reason to water cool, sound. The radiator fin density may look like the XSPC RS but it looks nothing like an RS240 in build quality.

    Glad to see Win 8 support, that is a good thing indeed. Their warranty is 5 years, that is outstanding, left a post at Swiftech wondering why they could/would/can’t match this with their 3 year. I fully understand the timetable of new products and technology release, yet this is important for many reasons to the end user.

    Really like the Link software control, it is very cool. But honestly, USB, SATA and 3 pin headers, come on! We want to cool our hardware, not evolve into a new cooling subsystem. That is too much crap to deal with just to cool things off, perhaps Cooler Master gets it right also with no darn software involved either. It is nice and neat, but their must be balance in function and form.

    Bottom line it is a great product for what it is, but it is only what it is. If you like control then Link is the cats’s meow, if you want a cheap no software solution in 120.2 then Cooler Master under $100 is the ticket. Every person is different, their priorities are abundant and unknown. Corsair has most all their bases covered in the market. It is up to the end user to research what they think is important.

    That is why you are so important today Martin. WCing is just getting going, your reviews will enable customers to make the right choice for their needs and wants in the AIO market. You are the only one doing this quality of work like you do, you are the authority in this field

    • Martinm210 says:

      On the 1.2W, that is something that was reported by Frosty Tech on the H100 here:

      Also seen this YT video here on the older H100:

      The H100i is however a different manufacturer, so it could be different. Based on RPM alone, a D5 which is a pretty large motor, only draws about 3-4watts when dialed down to 2000 RPM and the H100i is spinning at 2300 and much smaller. I would guess it is probably in the 1-2W range based on that, but we’ll see when I do some real testing later..:)

      • That is good logic estimating the power of the pump. Maybe Frosty Tech went to CoolIT to find the pump spec or just metered the current themselves at full speed ( I*V=W ), you could always do that also to get a close number. For some reason I always thought the D5 was close to a DDC in power as they are very popular. So the pump in an H220 is better than a D5, WOW. But I also always thought the DDC and 35x pumps were very similar, I need to do some more reading on this it seems.

        Can’t wait until you take one apart to get a look inside that pump. I hear Asetek is suing Cooler Master over the Seidon pump, patent infringement. Saw an exploded drawing of that pump and it is really a whole different animal altogether from a Swiftech pump. They also have a suit against CoolIT, so I gather the pumps are a lot alike.

        Perhaps the change of OEM’s explains the Link thing, but maybe not because the X60 is Asetek with software, only Corsair really knows and they aren’t saying….. 🙂

        • Martinm210 says:

          No, that D5 number was using one of the dialed down tests. A D5 at full speed is a good 18+W pump that can spin upwards of 4600rpm. The H220 is only capable of 3,000rpm in stock form. I was just looking at an RPM comparison at similar speeds. ST says the H220 produces similar performance to the 35x when dialed down to like RPM, bit I haven’t tested that. It is a bit tricky with the integrated base as the restriction will mean I have to compare to a block+pump combo to measure the resulting pressure vs flow remaining. A little different than just measuring pump performance alone.

          I found another YT video of a guy taking apart the H100 pump/block and it looked like it was a DDC-LT pump which is quite a bit more power than what FT reported. It appeared the lower flow rate was primarily due to the restriction of the copper cold plate microchannel density. So anyhow, not sure I trust the net sources much…going to have to test myself..:)

          That’s not at all a bad thing in a CPU only loop design, but it may be the reason for the resulting low flow rate. I suspect both the H220 and H100i may have similarly restrictive bases, but that is something I can measure. I am back at my other home away from testing gear for the weekend, but I did come back to my second H100i used sample.

          I will see if I can’t at least do a little breaking down the h100i block/pump and maybe some current measurements. I think I do have a multimeter down here. Maybe a simple video of the H100i flow rate too. It’s kind of moot in a cpu only loop sealed system, but it may give some insight into modding options or lack there of regarding expansion. I am pretty sure you could add a reservoir and your own tubing and possibly upgrade the radiator, but I’m not too sure the pump would do well trying to push through your typical full cover GPU block. Most FC GPU blocks are pretty low in restriction but they also like having a little higher flow rate.

          I may also have to order a lower flow rate meter. My King Instruments meter is only capable of measuring down to .5GPM and I suspect most of these AIO units will be pushing that lower limit. I probably need a 0-1GPM King to do any sort of good pump curve analysis.

          I’ll report back after having something more done…:)

          • Martinm210 says:

            Update, measured the pump power consumption = 12.10V x .17A = 2.2 Watts. Added photos to the post on page 3 here:

          • What was I thinking, of course the D5 is a beast, did not read your post carefully enough. Yes, as you say it’s going to be tricky if not impossible to just measure pump performance alone on these AIO kits because their all pump/block assemblies. And I bet every company has a bit different designed block, going to have to think about this one.

            DCC-LT? You lost me here, went through some of your pump reviews but not sure what the LT is your talking about………

            The new page where you took it apart is GREAT, the pictures look fine! It’s on page 4, not page 3 by the way. Well you nailed down what kind of power this pump has, 2.2 watts is not that bad, a little better than what FT reported. To my untrained eye it looks a lot like the H220, but they all still look the same to me at this point, until I went to the DC-LT page, now I get it, they sure do look the same except the LT is twice as powerful. That is why your reviews are cool, you got the background to reference stuff.

            All and all it looks pretty good, is Aquatuning connected to CoolIT? The barbed fittings are nice and long, 3 barbs ought to hold pretty darn good. But they did kinda cut the quality with the bronze instead of ceramic bearings. But it’s $40 cheaper than a H220, guess that is why along with the aluminum radiator they can sell cheaper.

          • Martinm210 says:

            Yeah just got done with the internal pics page. I’m really not sure who manufactures what anymore. I have heard things like there are only two major radiator manufacturers in China that make most of the WC rads out there, but I never know. The H220 has a similar looking set of impeller blades but the magnet and shaft is much beefier on the H220. The H220 impeller also appears to be balanced where you can see weights installed. All in all not a bad little pump, but the H220 build quality is still better. I heard rumors that the H220 pump was designed to handle MUCH higher speeds and can produce head pressure higher than the 35x, but it is dialed down to reduce noise in the H220 kit. That makes sense because any time you solid mount a pump, noise will be higher. Anyhow, internal photos are up, but hydraulics testing still on hold until I get back up to my other home..:)

  2. Tingez says:

    That’s funny i thought i left a comment on here .. mmmm interesting

  3. Neo Zuko says:

    Sold my non-expandable aluminum rad weak pump H100i for the expandable copper rad strong pump H220!!!

  4. Tingez says:

    No problem Martin, i will reformulate my thoughts again and get back to you … lol

  5. Sorry, I should have waited until you finished this page, thought you were, my mistake. Just don’t like cluttering up the comments area.

    The way they distribute the water down on top of the micro channels looks just like the way the XSPC Raystorm does it. And those micro channels look just like the ones Cooler Master has in their Seidon block. I kind of like that better than how ST does it with the little spike things.

    • Martinm210 says:

      No problem, that’s what I like about doing the living review format. Gives me a chance to get some help and feedback along the way. No comment is clutter IMHO, just good discussion…:)

    • Yes, Stephen at ST told me they spun the H220 up to 4200rpm and it bettered the 35x in pressure. And they will market them as the “Extreme Edition” down the road sometime. It must be in the blade design. He said that noise was why they tuned it down, something about the whole idea was to make a quite setup, with the QP rad and helix fans at only 1800rpm.

      The pictures look great too, must have a good camera in your phone 🙂

      For no good reason I sometimes think about vibration sitting right on top of my CPU, say if it got out of balance. Not sure what it would hurt, if anything. I agree with you that the H220 is the best AIO pump so far. Don’t think any current product is going to better them right now anyway.

      But as you say for what it is, it serves it’s intended use pretty good, it would not hurt to put a bit more tube on them. Great work, Thanks Martin!

  6. Quick question, just read a H110/H90 review going against the X/60/X40. They said the fan spacing was 20mm on the H110, did you measure the H100i? The fans are 15mm on the X60, both made by Asetek I think.

    • Martinm210 says:

      I didn’t measure it but I know the H100i and H220 were the same or they wouldn’t have bolted in the top of the case. I don’t know too much about 140 options but I did read something on NZXTs site that they hD changed spacing in the top of the Switch 810 case from 20 to 15mm. I think the new standard for 140s is also becoming 15mm if I understand right but there is a little of both out there.

  7. Martinm210 says:

    FYI, Added a flow rate video and expansion piece under page 5:

    Max flow is an extremely low 0.11GPM, I would not recommend trying to expand or modify.

  8. Geese….I can almost pee faster than that, almost..:)

    Is this at full rpm and through the block? Will be interesting to see what the H220, X20-750 (would have to run it through the Raystorm block) and the 300 can do through it’s block also. We know what they say in their specs, but you could get the exact number. Is there any way you could measure the pressure also?

    Would it be possible to leave the copper plate off and test it? The piece that actually contacts the CPU, that would give you some straight pump data without the block. Then you could calculate the block resistance from those numbers.

    The more I look at this H100i block innards the more I like it for some reason. The way they spray the water precisely over the middle of the micro channels very evenly to give all of them uniform flow out.

    Ya it sure looks like CPU only for sure on this rig, I would bet the rest of the true AIO coolers are in the same boat. It would only make sense I guess to give it just enough to do the job and nothing more. Seems a lot of people only want and need CPU cooling in a convent package, this certainly does a fine job there looking at the temps it gives, with a nice little software suit to boot.

    • Martinm210 says:

      Yes. Back at my other place where I could check RPM. It measured 2334 at max flow which is in line with what my other sealed sample reports. Also tested max head pressure and got 1.14PSI which is similar to what FT reported for the old H100 specs. I think it’s a good design for a sealed system, they have maximized fin density in the base and it certainly performs well thermally, just not something very compatible with adding more to it unless another pump is part of the addition.

      • You got it perfectly in a nut shell there, it’s very good for what it was designed for but nothing else. Yes indeed, I also like that CPU fin design, wish ST had something like that, it looks a lot better than the one on the H220.

        I somehow ran across the ENZOTECH Sapphire CPU block messing around, it’s only $40! They took the ST idea of the little spike things to a whole new level, really beautiful if you asked me. I even saved an image of it :). You can check it out here if you want to, also have a snap of the Arc Midi R2 with a H220 and XSPC 750 installed.

  9. mackan says:

    I’m sure I speak for several other people when I say this: “We want 120mm closed loop reviews!” I think it’d be incredibly interesting to read about since more and more people are moving towards smaller form factors. I know I am since I’m currently awaiting the release of SilverStone SG10.

    • It would be cool to look inside of the NZXT Kraken X40 pump/block. Looked at some new very small cases and many of them support 140mm fans now, the new SilverStone was very cool, forgot which model….

      • mackan says:

        I’d rather see a 1x120mm version of Swiftech H220. Nothing else comes close to it, in my opinion.

        • Do you mean inside of the pump? Or maybe a kit with a 120 instead of the 240…..
          I’d sure like to see them release a 280 kit and go head to head with the X60 and H110.

          • mackan says:

            I mean a 120 version instead of a 240.

          • Martinm210 says:

            I do have an MCR120QP and an MCR320QP in my rad collection Plus an MCres that I could do some side testing on. Still looking to for some GPU expansion testing too, but not really wanting to spend money to cool the 570 since it’s getting a bit older these days.

  10. Pedro says:

    Hi Martin,
    I will need to shorten the tubbing on a H100i and would like to know your opinion on a couple of things:
    – can I reuse the original tubing? I was planning on removing the ends from the radiator, cut them to the size I want and then plug them again onto the radiator barbs. Will I need a clamp to secure them? I’m asking this because the ends on the tubbing seem reinforced.
    – Can I use the original fill port to fill the system? Is the original plug reusable ?
    – Can I replace the original fill port with something else? is it glued ? how is the original fill port secured ?
    – is it feasible to make a hole on the top of the radiator and solder/glue a fill port? I was thinking of making a hole and them use epoxy glue to glue a fill port on the top ( if the original fill port cannot be used again )


    • Martinm210 says:

      yes you can use the old tubing but be prepared to spend some time trying to bleed. I would probably add a reservoir like the swiftech mcres or add a T fitting before the pump. I would also ad some clamps, some work drive or at least some zip ties would work.

      Big thing to know is how low flow rate is. I got about 0.1gpm where you need about .5gpm or so for bubbles to work their way out effectively from flow. This means you will have to manually get the air out and not something I would generally recommend. Most DIY systems are capable of nearly 1.0GPM to 2.0GPM which makes bleeding air very easy. The 0.1GPM of the H100i system is extremely low for user bleeding. I suspect in the factory they fill them with another pump and probably submerge the fill port/bleeding points to make it happen. It could be done, but bleeding will be a challenge for sure. i would highly recommend a good reservoir like the mcres and bleed outside the case where you can rotate parts and work the air out manually.