Scythe Ninja 3 - Testing Results

Written by Luke Ponio    Sunday, 25 July 2010 21:11
Article Index
Scythe Ninja 3
Packaging and Design
Testing Results
All Pages

Ambient temperature was kept at a consistent 20-20.5°C. Arctic Silver Ceramique thermal paste was used for all tests. Temperatures were analyzed using a temperature probe touching the Intel heatspreader and Core Temp for software measurements. The core temp reading in the charts is the average of all four core temperatures according to Core Temp. Four threads of Prime95 small FFT were ran simultaneously with Core Damage. The purpose of the test was to determine results that a home user would obtain. There was no scientific testing conducted, simply because I do not have the proper equipment.

Thermal Probe Placement

All thermal control options including Speedstep were disabled in the BIOS for all tests. All testing was done under Windows 7 highest performance power profile.

Test System Specifications

  • Intel Core 2 Quad Q9400 (8 x 333FSB = 2.66GHz @ VID 1.184v load 1.15v idle)
  • Asus P5W-DH Deluxe 3001 bios (1.65v MCH)
  • G-Skill 2x1GB DDR2 800 PC6400 F2-6400CL4D-2GBPK (2.1v)
  • XFX 8500GT Fanless PV-T86J-Y1S3
  • Kingwin Lazer LZ-850w


  • Scythe Ninja Plus Rev. B
  • Stock Intel Cooler

Test Procedures:

  • Idling at desktop for 15 minutes
  • Full load with 4 threads of Prime95 small FFT and Core Damage for 15 minutes


Scythe Ninja 3 Test Results

Obviously bringing the stock Intel cooler into this is like bringing a knife to a gunfight but it does help to show how much of an upgrade you can expect if you are an unfortunate soul still using the boxed cooler. The difference between the Ninja Plus and Ninja 3 surprised me so much that I re-ran the tests multiple times and did a couple remounts each. The numbers all stayed the same though. The difference between the high and low fan speeds also surprised me. Unfortunately my current thermometer doesn’t measure in tenths of a digit but that is still less than a 1°C difference from minimum RPM to maximum.

I also tried both coolers fanless because Scythe mentions it on the box but didn’t include the data because these were open air tests and the temperatures were sky high. After 15 minutes the Ninja 3 hit 86°C according to Core Temp with a probe reading of 67. The Ninja Plus hit 93°C core temp and 73 probe temp after only 8.5 minutes. I quickly turned it off after seeing those temps. If you want to run these heatsinks fanless you need to have airflow in your case. Scythe’s warranty says the Ninja 3 is designed for fanless usage such as e-mailing, internet browsing, word processing, and spreadsheet tasks at an ambient temperature of up to 25°C. It specifically warns against benchmarking or data processing and I can see why. The reason why I didn’t try fanless operation in a case was I have no accurate way of measuring the airflow inside my case and it is too large a variable to not consider. At idle I recorded probe temperatures of 41°C for the Ninja Plus and 36°C for the Ninja 3 so if you aren’t doing anything on your computer your CPU won’t burn up if you turn the fan off or it malfunctions.

Here is the Ninja Plus rev. B next to its younger brother the Ninja 3:

Scythe Ninja 3 Comparison

Overall I was very pleased with the performance of the Ninja 3. I did not expect such a big improvement over the older Ninja Plus Rev. B. I guess more fins and heatpipes really do add up to more efficient cooling.