Thermalright Ultima 90 is out :)

[BonzaiDuck]

Well . . . if the thermal resistance is lower, it's going to cool better . . .

That doesn't mean that the Ultima 90 is the "best," but I suspect, the way TR designs their products, it will match U-120-Ex performance.

Couple years ago when the XP-120 hit the street, there were issues about interference with capacitors around the CPU socket for certain boards, and the side of the XP-120 would make it difficult to fit a heatpipe cooler to the VGA card. They put out the XP-90 and 90C, because it addressed these clearance problems.

Some people are going to be put off by the IFX-14 or the U-120-Extreme because of the size of the cooler. The coolers are tall enough, they may not clear the side access-panel of the case for some narrower midtowers cases. So I can see some enthusiasts wanting a cooler like the Ultima-90 to resolve that problem.

On the matter of the Ninja, it turned in performance benchmarks that were close to the SI-120 per Citarella's OverClockers.com review. It was interesting to see that with a fan speed of just over 800 rpm (using the same fan on both coolers) -- the Ninja outperformed the ThermalRight cooler by a narrow margin, while the ThermalRight trumped the Ninja when the fanspeed was set between 2,200 aod 2,500 rpm.

But imagine how those fanspeeds are going to be useful. People who are absolute noise-phobes are going to run the lower fan-speed regardless of the Scythe cooler's thermal resistance at that speed of something close to 0.2 C/W, while those users who want the performance when it counts will thermally control the fan on either cooler.

I myself had considered getting a Ninja, only because it can pull air through two adjacent sides of the fins in a small midtower case, and in some such cases, would fit near the exhaust fan and PSU intake fan like a glove.

 

[RussianSensation]

Originally posted by: BonzaiDuck
. . . Anyway, combining the information from the Anandtech review of May 5 and the link you show for a forum-review of the Ultima-90, it is obviously better than the Tuniq, which comes in second to the Ultra-120-Extreme. So even if it matches the U-120-Extreme for performance, there may be other reasons (size, shape and other factors) for which it could be a preferred alternative to the U-120-Extreme.

It is important to note that the testing was done on an open test bed though and as Xbitlabs shows in their reviews, there tends to be a difference between a closed case and open test bed due to variation in the flow of air (or trapped hot air in a case). The difference in the testing method used does throw off temperatures between coolers since they can be more efficient in an open test bed for instance (Look at Big Typhoon in a closed case vs. Thermalright 120 and then compare the temperatures in an open test bed) - LINK

The 2nd point to consider is that the Ultima 90 was tested with an Antec fan at 3 settings and at high settings it does outperform the Tuniq but it is mentioned in that review that it sounds much louder. Of course you can swap the fan for a more efficient one. But if we were to compare coolers at fans that spin at similar speeds as was done here (roughly 1400-1600 rpm speed chosen to provide similar noise level), the Tuniq smoked the Big Typhoon (~1400 rpm) by almost 10*C. From the review above you can see inside a closed case the Big Typhoon at the same 1400rpm speed was 5*C inferior to a dual 1200 rpm fans Thermalright Ultra Extreme.

Right now Newegg is selling the Tuniq Tower for $45 with free shipping. If you compare Ultima 90 / Ultra 120/EX and add cost of fans and shipping charges, the Tuniq is still the best value at the moment (of course if you seek ultimate performance it isn't).

The 3rd point to consider is how was the thermal resistance number derived? We dont know the testing methodology manufacturer uses to come up with an absolute value of the thermal resistence due to variation of fans which can be mounted on the heatsinks like the Thermalright. But since that's the number you care about for the Tuniq it is 0.16 - 0.21 °C / W (high - low fan speed).

 

[BonzaiDuck]

Citarella to OverClockers uses a PC test-bed and lab equipment (calorimeter?) to derive two different measures of C/W.

On the matter of the test with dual fans on the TR Ultra-120-Ex, that makes it look as though there's a handicap in favor of the TR, but it doesn't really increase airflow in CFM through the fins by that much. It probably reduces air-leakage from the side of the fins.

On the C/W measure for the Tuniq, I'm surprised that with those values (above 0.15 C/W) that it actually performs well enough to get the rating it received in the Anandtech review. Some of the other coolers -- like the Zalman line -- have C/W values that are lower and closer to ThermalRight's.

ThermalRight may measure thermal resistance of their coolers, but they don't release the information to the public -- it is part of their design, evaluation and production process (as it should be). It's hard to actually find reviewers who attempt to measure t_r__, but if they insinuate the thermal power of the bench-configuration's processor at the loads undertaken, then that plus the idle and load temperatures are sufficient to compare among different coolers.

 

[VinDSL]

Originally posted by: BonzaiDuck
On the C/W measure for the Tuniq, I'm surprised that with those values (above 0.15 C/W) that it actually performs well enough to get the rating it received in the Anandtech review. Some of the other coolers -- like the Zalman line -- have C/W values that are lower and closer to ThermalRight's.

I think you're working under a misnomer!

IMHO, fin density is paramount!

All kidding aside, that's the one thing that the Ultima-90 has going for it...

Look at the pics - it has the same fin density as a Tuniq Tower 120 - and that's the reason it works so good!

The 'problem' with high fin density, if you will, is it requires a LOT of air flow - the more the better...

Anyway, all things considered, I'll stick with the Tuniq Tower 120, thank you!

 

[VinDSL]

Originally posted by: BonzaiDuck
Couple years ago when the XP-120 hit the street, there were issues about interference with capacitors around the CPU socket for certain boards, and the side of the XP-120 would make it difficult to fit a heatpipe cooler to the VGA card. They put out the XP-90 and 90C, because it addressed these clearance problems.

Some people are going to be put off by the IFX-14 or the U-120-Extreme because of the size of the cooler. The coolers are tall enough, they may not clear the side access-panel of the case for some narrower midtowers cases. So I can see some enthusiasts wanting a cooler like the Ultima-90 to resolve that problem.

Once again, why not give it all up for a Tuniq?

Clearance has NEVER been a problem with the Tuniq Tower 120!

Look at the clearance on my mobo! That's one of the reasons it's so popular...

Bench build -- MM U2-UFO Opti1203 Tray:

Example #1 (vid card)

Example #2 (caps)

Example #3 (nb & mosfets)

Also note: the 120mm Tuniq (2000 RPM) fan sits low in the chassis blowing air across the NB, MOSFET & video RAM sinks... :thumbsup:

 

[RussianSensation]

VinDSL, have you thought about attaching an additional 120mm fan on the side of the Tuniq facing the back of the videocard? I am wondering if this would provide any tangible difference at all?

I am thinking of setting mine up this way and having it positioned vertically so that the back case exhaust fan kicks the air out. The only question is whether or not to attach a 3rd fan to the back side of the Tuniq (which would mean the heatsink Tuniq fan would be positioned between the back side fan and the back exhaust fan):

<--- Exchaust fan <---- Tuniq fan <--- Rubber band attached optional 120mm back side of tuniq fan

Any thoughts?

 

[BonzaiDuck]

Russian --

That's a viable approach, even for other tower coolers. Nevertheless, it won't draw the coldest air intake through the CPU cooler if that air has already traversed the VGA cooler-fins. I'm not sure what adverse impact it would have on peak CPU temperatures, and it might be negligible. If it means drawing intake from two fans mounted at 90-degree angles, the only impact might be to cool the VGA -- and even drop the CPU temperatures to a minimum for more air-flow. But we all make these compromises, even for drawing air-intake through hard disk cages.

 

[VinDSL]

Originally posted by: RussianSensation
VinDSL, have you thought about attaching an additional 120mm fan on the side of the Tuniq facing the back of the videocard?

Any thoughts?

I would NEVER do that!!!

I hate the kludgey look of having wired or zip-tied fans on the outside of HSFs! To me that defeats the whole purpose of running a "desktop computer" so called...

Sorry, but I'm on the fringe when it comes to this sort of stuff. To me, looks trump performance! If you can beat the system and get both, fine! Otherwise, I might as well be running a Apple Notebook Pro instead of a MM U2-UFO, where the guts are hidden behind a logo, you know?

 

[firewolfsm]

I want to know the price on this.

and what the Ultima 120 will be like.

 

[BonzaiDuck]

Firewolfsm said:

"I want to know the price on this."

__________________________

Not sure which "this" you mean -- we've been discussing at least three different heatsinks. I'm pretty sure, however, that all of them are available at :

Heatsink Factory -- Company located in California

There are several resellers in addition to HF who also carry a wide assortment of heatpipe coolers and watercooling kits among them: FrozenCPU.com, SidewinderComputers.com, SVC.com, CrazyPC.com . . . . and of course NewEgg, EWiz, ZipZoomFly. I've bought heatsinks from Frozen, Sidewinder, SVC, NewEgg and Heatsink Factory. For that, the latter has reasonable prices, and SVC seems reasonable. Frozen tends to have prices that are a tad on the high side, but they carry a lot of specialty stuff, and their tech-man is always responsive by e-mail. Last time I plumbed the web-site at SVC, they had photos of the employees up there in Silly-cone Valley, and the young honeys among them, photographed in SVC tee-shirts, have nice bazoobas.

And SVC's tech-people are also very helpful and responsive.

So . . . . shop until you drop. You might save 40% on one thing, 60% on another -- why that's a saving of 100%!!

[ ]

 

[BonzaiDuck]

FOOTNOTE on the issues of Fin-Density and Thermal Resistance

When they measure the thermal resistance of a cooler, it accounts for the entire product, and therefore fin density as it affects thermal resistance would be encapsulated in an evaluation by thermal resistance alone.

 

[BonzaiDuck]

On the other question, let me say I've watched ThermalRight, used several of their products, and I've spoken to their tech-reps many times over the last few years.

I always want to see performance reviews before I buy anything. However, based on what I've seen so far, and based on my experience with TR's design and production patterns, I'd be willing to make bet that the Ultima 90 is a good alternative to the Ultra-120-Extreme for a couple of reasons:

It isn't a monster-cooler like the Ninja, the IFX-14, the Tuniq or the Ultra-120's.

It is likely to be every bit as good as the ultra-120s.

The drawback -- I think it was mentioned by RussianSensation -- is in its accommodation to 92mm fans, which must spin faster with more noise to get the same airflow. But some fans of that size, like the Panaflos, are fairly quiet excluding the ubiquitous air-turbulence "white-noise" -- the "sshhhh" sound coming through your air-conditioning vents.

I'd go back through this thread and look at the Ultima-90's spec's at reseller sites, but I think it also accommodates a 120mm fan. If that's the case, get one with a beefy motor that can be controlled thermally -- preferably from the motherboard. Some of the 120x38mm Panaflos have a reputation for "quietude." There's a tri-blade Delta 120x38 which can push 140 CFM (at 3,700 rpm and close to 50 dBA), but can be controlled to run at speeds between 1,600 and 2,500 rpm where it is also quiet. These fans have a higher exhaust pressure, which is needed for a cooler with this sort of fin-density -- VinDSL may have mentioned that fact.

 

[RussianSensation]

Originally posted by: BonzaiDuck
FOOTNOTE on the issues of Fin-Density and Thermal Resistance

When they measure the thermal resistance of a cooler, it accounts for the entire product, and therefore fin density as it affects thermal resistance would be encapsulated in an evaluation by thermal resistance alone.

I am saying..

1) Thermal resistance changes based on the fan you use because you are effectively accelerating heat dissipation from the fins

2) How does the manufacturer provide an accurate measurement of thermal resistance given there are dozen + options for fans? So now I am supposed to take their absolute number at face value?

Look at Diagram 7 where it shows how thermal resistance is affected by airflow.

For this reason saying that Thermalright Ultra has resistance A and Tuniq B, it still doesnt tell us with certainty which one would be most effictive but only provides a best guess.

 

[BonzaiDuck]

Yes, RussianSensation, you are absolutely correct.

Whatever measure is taken for a heatpipe cooler, the CFM throughput and therefore the fan determine the overall thermal resistance value.

Citarella at OverClockers.com would use a single fan of the same model and manufacture, applying it to all coolers tested in a comparison review. The test would take a high and low value of fan rpms showing also the measured dBA noise-level.

I can accept this, knowing that fans of a particular amperage rating (@12V) and size (120x25 versus 120x38mm) with an rpm speed fall into a "fan class," and the CFM-throughput will pretty much be in a discrete range for fans in that group.

The measurement of those things can at least give one a pretty good idea of the cooler's performance in conjunction with other factors, like noise-level.

If two coolers are neck-and-neck or the measured tr value or temperature values for the same thermal wattage and clock settings are very, very close, you're going to see that.

You might worry about "fair ratings" since there's a lot of payola afoot in the industry -- maybe not under-the-table cash inducements, but attempts to make a good advertising account look good by avoiding inclusion of better products or "serious competitors" in the published review. That's the problem I have with "Maximum PC Magazine."

But again -- it's like choosing how you will acquire knowledge about anything. If you want to understand the JFK assassination, you don't just read one book about it. One would have a great deal of contempt for someone who thought that level of research was "enough." If you want to understand a field like economics, just picking a very good book like Galbraith's "New Industrial State" doesn't make you an expert, or even "knowledgeable."

So you read as many reviews as you can find from different sources.

 

[swatX]

Ordered Ultima 90 via Jab-tech for $45 + shipping. Any ideas on a good 120mm fan with good cfm without too much noise?

 

[n7]

Scythe S-Flex
Noctua
Nexus

 

[Noubourne]

just realized how many posts were between the one I was responding to and mine.

I read the XS review of the Ultima 90. It looks good, but I'll wait for the 120mm version.

 

[BonzaiDuck]

Well, you know I'm just guessing, but I think it might be the other way around -- the Ultima 90 is the 92mm version of the U-120-Extreme. Of course, there's the increased fin-density, so maybe they will release a 120mm version.

[Even so, I thought the Ultima had provisions for both fan-sizes. I better double-check . . . ]

Also -- here's a thought about that, too. I'm looking closely now at the stock-cooler fans that come with the retail Q6600 and E6600 boxed-processor. They're pretty beefy at 0.40A and 12V. I'm getting . . . . "some ideas" . . . . again . . . .

 

[tuan209]

Would you have to remove your mobo to install this heat sink?

 

[swatX]

Originally posted by: BonzaiDuck
Well, you know I'm just guessing, but I think it might be the other way around -- the Ultima 90 is the 92mm version of the U-120-Extreme. Of course, there's the increased fin-density, so maybe they will release a 120mm version.

[Even so, I thought the Ultima had provisions for both fan-sizes. I better double-check . . . ]

Also -- here's a thought about that, too. I'm looking closely now at the stock-cooler fans that come with the retail Q6600 and E6600 boxed-processor. They're pretty beefy at 0.40A and 12V. I'm getting . . . . "some ideas" . . . . again . . . .

XS review used 120mm on the ultima 90

 

[BonzaiDuck]

I'll have to give it a closer look, but I'd think so -- initially. That is, you would need to remove the motherboard for the initial installation.

There is a myth afoot for these more recent ThermalRight "tower" coolers that you would need to remove the motherboard each and every time you remove and reinstall the cooler.

Not so. I've said this before in other posts:

If the motherboard has 1/4" or 6mm brass or aluminum standoffs in the ATX pattern, once initially installed, the rear "X"-bracket should remain in place with the threaded extrusions secured in their motherboard holes. If the motherboard has 3/8" or 8mm standoffs, then you can shim the X-bracket with a plastic ruler or popsicle sticks before removing the screws -- to keep the threaded extrusions in place.

WHOA!! I just went to the ThermalRight website to look at this baby!! They never stop improving in their designs!

First, there IS NO X-bracket, as we've come to expect from the Ultra-120- and Extreme, although there is a rectangular (under-the-motherboard) bracket with the same four threaded extrusions. So yes -- must remove motherboard for initial installation of that bracket.

Second, THIS TIME -- they've added another bracket fitting on the motherboard top-side which secures independently to the threaded extrusions. Bear with me a minute on this.

Third, the Ultima-90 is THEN secured to this top-side bracket by two screws, enabling the cooler to be removed without any attention to the motherboard or shimming the underside bracket.

Nice going, ThermalRight!!

 

[gorobei]

Originally posted by: BonzaiDuck


WHOA!! I just went to the ThermalRight website to look at this baby!! They never stop improving in their designs!

Second, THIS TIME -- they've added another bracket fitting on the motherboard top-side which secures independently to the threaded extrusions. Bear with me a minute on this.

Third, the Ultima-90 is THEN secured to this top-side bracket by two screws, enabling the cooler to be removed without any attention to the motherboard or shimming the underside bracket.

Nice going, ThermalRight!!

Actually they are copying Noctua's design. The NH-U12F has a x-bracket for the back and two U-brackets on the top of the MB. Flanges that attach to the HS-base provide threading for two springloaded screws to mount the HS.

http://www.bjorn3d.com/read.php?cID=973&pageID=2627

 

[VinDSL]

Originally posted by: gorobei
Actually they are copying Noctua's design. The NH-U12F has a x-bracket for the back and two U-brackets on the top of the MB. Flanges that attach to the HS-base provide threading for two springloaded screws to mount the HS.

LoL!

I think the Duck has been pwned!

 

[swatX]

Doesnt really matter. The end result is what people are after.

I will be getting my Ultima 90 tomorrow. Unfortunately i cant test until the following week when i get my Q6600 G0.

 

[BonzaiDuck]

They look much alike, but I'd bet some chump-change that the Ultima performs on par with the Ultra-120-Extreme. So -- give credit where credit's due, but I doubt Noctua will be able to sue on patent-infringement.

You can tell from the over-clock level, the processor and the reported temperatures that this Noctua -- good enough -- but not at the top of the pile.