Thermalright Archon
By: Dan Durland
Introduction
Today I have a new heatsink on the testbed, the Archon from Thermalright. Thermalright, based in Taipei, Taiwan, has been a leader in the heatsink market since 2002. With innovative designs Thermalright has been a favorite among overclockers and enthusiasts for close to 10 years.
Features
- Armor style heatsink fin array, with mirrored copper base, elevates heatsink's overall quality.
- Large heatsink fin array provides dissipating area of up to 150x125mm, making complete use of the airflow, with 140mm or 120mm fan.
- With the max width of 55mm, tall memory heat spreaders are avoided.
- Six sintered heatpipe design, all heatpipes are nickel plated, to slow the oxidation deterioration to the heatpipe, to ensure longer usage and performance of the heatsink for the CPU.
- Supports Multiple-Platforms, and can be used on Socket 1366/1156/775 Platform, and AM2/AM3 Socket. For Intel sockets pressure is adjustable (40-70 lbs.)
- Includes one Thermalright TY-140 PWM fan, Ultra low noise at 21 dBA at max rpm (50cm testing distance)
- Fan-clips are compatible with both 140mm and 120mm fan.
- The Archon package includes a tube of CF III Thermal paste, to provide the best combination for use with Archon.
Specifications
Heatsink
- Dimension: 155mm x 53mm x 170mm (LxWxH)
- Weight: 806g (excluding fan and mounting brackets)
- Heatpipe: 6mm heatpipe x 6 units
- Copper Base: C1100 Pure copper nickel plated with mirror shine
- Dimension: 160mm x 140mm x 26.5mm (LxHxW)
- Weight: 140g
- Fan speed: 900 - 1300RPM (PWM controlled)
- Fan noise: 19 - 21dBA
- Airflow: 56 - 73CFM
- Connector: 4 Pin (PWM Fan connector)
Image Gallery
In our first four pictures we are looking at the packaging arrangement of the Archon. With the box opened the first thing we find is a Thermalright sticker and the installation guides. Directly underneath the paperwork we find the stock fan. The fan is listed as a 160mm x 140mm but in reality it is a 140mm fan with a off-sized enclosure. Removing the fan presents us with a thin white box containing the retention hardware. And the last picture shows the Archon, in a plastic bag, securely protected from possible damage by a form fitting piece of polyurethane foam.
If we open the thin white box we are presented with all of the retention hardware in a sealed plastic bag. With the retention hardware laid out on the left side (top to bottom) we see the AM2/AM3 retention hardware (backplate, X-bracket, 4 spring loaded screws), four (4) anti-vibration strips, and two (2) sets of fan clips. On the right side (top to bottom) we have a angled wrench for tightening the pressure vault mounting plate. Below the angled wrench is the pressure vault mounting plate. Lastly we have the Intel retention hardware (multiple support backplate, skt. 775 backplate cap, anchoring mount, (4) screw pillars, (4) thumbscrews).
In our next set of pictures we see the Archon's four (4) different sides. As advertised, the bottom has a mirrored finish.
In these next three picture we can see the convex base which Thermalright claims improves the contact with the center of the CPU. In the first picture we can see that I've balanced a straight edge across the base lengthwise. The second picture is a close-up of the first picture. We can clearly see the convex surface, only in order to be effective this portion of the base should be flat. In the third picture we can see a close-up of the base with the straight edge oriented width-wise. This surface is flat, however if a convex surface has any hope of improving the performance of the Archon, then this is the surface that should be convex.
This next picture is an example to continue my point on the convex surface. The first thing to look at is how the motherboard has bowed with the heatsink (Thermalright Ultra Extreme) installed. This bow is the whole reason Thermalright claims a convex base improves the contact with the CPU. My first though is that I always screw my motherboard to the motherboard panel of my case. The motherboard panel is flat, when we screw the motherboard to the panel does the motherboard straighten out or does the motherboard panel bend? I'm pretty sure the motherboard straightens out. Some might say the motherboard is still slightly bent at the socket itself, I say the metal socket retention plate, that is bolted to the motherboard (to hold the CPU in place) keeps the CPU itself flat. A convex surface does no good in my opinion, or in the thousands of other users who lap their bases. But lets assume, for a moment, that a convex base will help. Which way should the convex surface be oriented? When considering the Archon we need to go back to the features, provided by Thermalright themselves.
Thermalright said:
Looking at our picture below we can see that this heatsink is oriented horizontally or lengthwise. The Archon, mounted in this position, would fully utilize the convex surface of the base. However, the Archon, being 22.2mm longer (11.1mm on each side) will block the memory slots when it is mounted in this position. When installed to utilize the max. width of 55mm the convex base is fighting the bend of the motherboard. I can only see the convex base of the Archon as a double failure and I hope Thermalright will one day admit the error of their ways.
OK, I'm finished with my complaint so lets move on to our last few sets of pictures. The first two pictures show the Archon with the two different fan configurations I'll be using. The third picture shows the lapped base; the lapping took place after the initial testing was completed. The last two pictures show the two Archon fan configurations installed and running.
Installation
Installing the Archon was pretty straightforward and I encountered no difficulties. Before purchasing this heatsink I strongly recommend checking the Motherboard Compatibility List for issues with your system. Installation does require the removal of the motherboard unless your case has an access hole, located directly behind the CPU socket. A screwdriver is required for installing the Archon with both Intel and AMD systems. Intel users seem to be favored over AMD users as Thermalright's pressure vault system is an Intel only feature. Using the supplied angled wrench, Intel users can adjust the amount of pressure (40 to 70 lbs.) the heatsink excerts upon the CPU. I'm a firm believer in "the tighter the better" theory and I'm disappointed that AMD users are not offered the same benefit. Below you will find a few links to the installation guides and motherboard compatibility list in pdf file format.
Testing
Test System
- Intel i7 920 D0
- EVGA E757 SLI LE
- 12 GB OCZ DDR3 1600 (6x2GB)
- EVGA GTX 280
- Corsair 750w PS
- 60GB Agility 2 SSD (OS)
- Seagate 250 GB HDD (Data, Page File)
- Lite-On Blu-ray
- Cooler Master Stacker 810
- Thermalright Archon
With a price in the $80 range the Archon is in direct competition with the very best heatsinks available on the market today. So that's exactly what I'll test it against:
- Thermalright Archon - The Thermalright Archon will be tested exactly as it comes out of the box, with its convex base and the included 140mm fan. However, the Archon will also be tested with the convex base lapped to a flat mirrored finish with a pair of Yate Loon 140mm High-Speed 140 cfm fans installed.
- Thermalright Ultra Extreme Copper - The TRUE Copper, as it name implies, is made out of copper. It is a single tower design heatsink capable of utilizing one or two 120mm fans. Copper has the second highest thermal conductivity after silver which allows it to absorb a great deal of heat away from the CPU. This ability to absorb heat also means it doesn't dissipate heat as well as other metals like aluminum. Because of this I'll be testing the TRUE Copper in a single and dual fan configuration using Silverstone FM121 120mm fans. The TRUE Copper, like the Archon, comes from the factory with a convex base. The base of the TRUE Copper has therefore been lapped to a flat, mirrored finished. The TRUE Copper, priced in the $100 range with no fans included, is the most expensive heatsink that will be tested today.
- Noctua NH-D14 - The NH-D14 is an extremely large aluminum, dual tower design, heatsink that ships with two fans. The NF-P12 is a 120mm fan located on the outer edge of the heatsink blowing air at and through the heatsink. The NF-P14 is a 140mm fan located between the dual towers pulling air through one tower and pushing it through the other. Being composed mainly from aluminum, which is only 59% as thermally conductive as copper, it does not absorb heat quite as well as copper by it does do a superior job at dissipating the heat it does absorb. Priced in the $80 range the NH-D14 is a main competitor with the Archon.
- Corsair Hydro H70 - The H70 is a All-in-One, self-contained water cooling solution that utilizes two 120mm fans. The fans are capable of operating at 2000 rpm or 1600 rpm through the use of a 7v resistor wire adapter. Priced in the $100 range, the H70 comes eqipped with everything it need to operate making it the second most expensive cooling solution in our tests.
Test Configurations
I will be testing the Archon against its competitors at two different CPU speeds, stock and overclocked. I'll also be using two different fan speed categories, Low Speed Fan and High Speed Fans.
Stock CPU Speed
- The stock CPU speed tests will be at 2.66 GHz with all the bios setting at default with the following changes.
- Turbo Mode Disabled
- CxE Function Disabled
- Intel SpeedStep Disabled
- with VDroop
- DIMM Voltage @ 1.65v
- Memory Latencies @ 7-7-7-24-54-1T
Low Speed Fan
High Speed Fans
- Archon (factory base & fan)
- Archon (lapped base & stock fan)
- TRUE Copper (1 fan)
- Corsair H70 (1600 rpm fan speeds)
High Speed Fans
- Archon (lapped base with two Yate Loon fans)
- TRUE Copper (2 fans)
- Noctua NH-D14 (2 fans)
- Corsair H70 (2000 rpm fan speeds)
Overclocked CPU Speed
- The overclocked CPU speed tests will be at 4.0 GHz with the following bios setting changes.
- Turbo Mode Disabled
- CxE Function Disabled
- Intel SpeedStep Disabled
- without VDroop
- CPU VCore @ 1.3125v
- CPU PLL @ 1.725v
- DIMM Voltage @ 1.65v
- QPI PLL @ 1.3v
- IOH VCore @1.375v
- IOH/ICH I/O Voltage @ 1.625v
- ICH VCore @ 1.3v
- CPU Host Frequency @ 200 MHz
- Memory Frequency @ 1067 MHz (2:8)
- Memory Latencies @ 7-7-7-24-54-1T
Low Speed Fan - High Speed Fans
At 4.0 GHz the Archon, with the factory base & fan, could not complete the stress test and as such that configuration was eliminated from the results. When the Corsair H70 was tested, at 4.0 GHz with the fans at 1600 rpm's, the temperatures were abnormally high so the testing has halted. Because these two configurations were eliminated from the testing all of the results were able to fit into one graph. The following heatsink and fan configurations were tested at 4.0 GHz.
At 4.0 GHz the Archon, with the factory base & fan, could not complete the stress test and as such that configuration was eliminated from the results. When the Corsair H70 was tested, at 4.0 GHz with the fans at 1600 rpm's, the temperatures were abnormally high so the testing has halted. Because these two configurations were eliminated from the testing all of the results were able to fit into one graph. The following heatsink and fan configurations were tested at 4.0 GHz.
- Archon (lapped base and stock fan)
- Archon (lapped base with two Yate Loon fans)
- TRUE Copper (1 fan)
- TRUE Copper (2 fans)
- Noctua NH-D14 (2 fans)
- Corsair H70 (2000 rpm fan speeds)
Testing Software and Procedure
- LinX 0.6.4 64-bit (ALL memory selected)
- Real Temp 3.60 (Logging enabled)
LinX will be run for a minimum of 50 minutes followed by a 30 minute idle period. Idle and Load temperatures will be shown as an average in a variety of configurations.
- Average of all 4 cores combined
- DeltaT average of all 4 cores combined (core temp. minus the ambient temp.)
- Individual average of all 4 cores
- DeltaT individual average of all 4 cores (core temp. minus the ambient temp.)
The ambient room temperature was monitored using a UEi DT200 Digital Thermometer. The ambient room temperature stayed within a 22C to 23C range. A value of 22.5C was used to calculate the DeltaT.
Results
Stock CPU Speed
In the first set of graphs we're looking at our overall average temperature, at idle and full load, at stock settings. In the first graph we can see the Archon, with the factory base and fan, is 4.7C hotter under full load than our coolest competitor the Corsair H70. When compared to its modified self it is a full 2C hotter. When we upgrade our fan options we can see the Archon performing closer to the rest of pack, but the Noctua NH-D14 is the clear winner.
Our next two graphs are the average temperature per core at an idle and full load. These graphs start to tell us about the performance in a little more detail. Our first graph below shows us the data used to construct the first graph above and the second graph below correlates to the second graph above. If we focus on the Archon in factory base and fan configuration, we had an overall average temperate of 56.8C. But our new graph is telling us the actual average temperatures were 59.7C, 56.1C, 57.4C and 54.0C. And to complicate the whole issue a little more I'll make another point, these newer more specific values are still just more averages from a much larger set of numbers. Regardless of the detail, these new values are still telling us the same basic message our last graphs did.
Our next four graphs below correlate to our four graphs above, showing us our DeltaT values. In theory, if your system is configured like mine, you could add your ambient temperatures to these values and calculate your approximate temperature with each heatsink.
Overclocked CPU Speed
At our overclocked CPU speed of 4.0Ghz we can start to get a clearer picture of how well these heatsinks are actually performing. Unlike our previous testing, at this speed under full load, the heat generated is more likely to cause system errors if it is not adequately transferred to the heatsink for dissipation into the atmosphere.
In our first graph we can see that the Archon, with the base lapped and the stock fan attached, completed the testing with an average full load temperature of 76.4C. While this is the hottest value we see in our group, it is at an acceptable level. Upgrading the fans on the Archon proves to be recommendable, providing an additional 3C improvement. While it comes as no surprise to me, the TRUE Copper is starting to flex its muscles. At these heat levels the copper's ability to absorb heat is matching the NH-D14's ability to dissipate its heat via its aluminium finned twin tower design.
Our second graph is showing us the more detailed data used to construct our previous graph. In our previous graph the Archon, with a lapped base and the stock fan, was at an acceptable level. This graph is showing the Core_0 of the Archon at an average temperature of 79.7C under full load, I would classify this as an absolute maximum temperature level for continuous use. The Archon, in the two fan configuration, is performing at a more acceptable level.
And our final two graphs are showing resulting the DeltaT values from our previous graphs.
Conclusion
Pros
- Attractive, Stylish
- Slim Profile, provides clearance for tall memory heatsinks.
- Six Heatpipe Design
Cons
- Price
- Off-Sized 154mm x 118mm Fin Array
- Convex Base
- Inadequate Fan
I've had this review finished, with the exception of my final conclusions, for at least a week now. I had such high hopes for this heatsink, I even paid $80 for it from my favorite online heatsink retailer. But the truth is, straight out of the box, the Archon is no better than a $35 heatsink. If your willing to spend $20 on sandpaper and $20 on better fans the Archon will perform as well as a $50 heatsink. While Thermalright claims the Fin Array is 150mm x 125mm, my Digital Calipers say it's 155mm x 118mm in an ever so slight V configuration. If the Fin Array was resized to 140mm x 140mm, the fan or any 140mm fan would provide more air flow to the array. In it's present design 15mm of fin area, of the heatsink, has no airflow and almost 20mm of possible fin array could be exposed to the airflow.
In its present design the Archon is a over-priced, poorly designed hunk of metal with no special qualities that can't be improved upon. I'm awarding the Thermalright Archon 1 Star which is rather generous in my opinion.
