Installation and Testing:
 

ECS NF650SLIT-A Test System

Installation was a breeze and the motherboard immediately powered up without a glitch. The OCZ memory with the Platinum XTC spreaders goes kind of nicely with the brushed and polished aluminum features on the motherboard. The motherboard didn't fail any boot sequences when switching out memory, one set right after the next. The system didn't require a CMOS Clear or motherboard reset to boot each time as some motherboards seem to require these days. Most the time this is due to the BIOS. Successful changes and boot ups on the fly are a very good traits for a motherboard. This also means a much greater chance of success for other end users.
 

As I mentioned previously, the 650-SLI motherboards are currently the most cost effective choice. The features it lacks are rarely needed by the common user. What it doesn't lack is dependable use. Every 650-SLI I've worked with has yet to give me any issues during installation and setup. You'll be happy to know that the ECS NF650SLIT-A is no exception and installed flawlessly. Here are a few BIOS shots of what to expect as you browse the motherboard. The BIOS is comparable to Abit's FP-IN9 Fatality, except for where voltage options are concerned. The NF650SLIT-A only has a couple of options when it comes to the CPU Link, Memory, and Chipset. The voltage range is a bit limited, especially in regards to the DDR2 Memory. The settings only go up to 1.95 volts! Which left me wondering what happens to the system when memory modules require 2.35v like Patriot's PC2-9200 Extreme.

Advanced Features	CPU Multiplier Settings	FSB & Memory Config
Memory Timings	Limited Voltage Options	PC Health Status

I found that leaving the BIOS settings to Auto yielded the best results for stability. Tinkering with the settings had sporadic results when it came to adding extra voltage under overclocked settings. The OCZ PC2-7200 Platinum 4-4-4 modules operated to spec on the platform up to 920MHz before the system would error loading the boot file. Interestingly, the OCZ memory needs 2.1v by spec to run properly and the maximum manual setting of 1.9v just wouldn't do. So again, leaving the BIOS setting on AUTO allowed me to use the memory up to its specified 900MHz 4-4-4 timings. While the voltage options aren't there, the motherboard is at least capable of handling itself competently. It would have been nice to push the memory to 1000MHz with 4-4-4 timings, but it just isn't completely possible. Your results may vary.

CPUZ @ 1066FSB	800MHz @ 4-4-4	Mainboard
CPUZ @ 1300FSB	976MHz @ 5-5-5	Module SPD

In terms of overclocking, the motherboard reached a maximum of 1300FSB before instability set in which couldn't be offset with the lack of extra voltage settings. I've seen other 650-SLI motherboards achieve very similar results around the 1300 FSB area with only a little success above. At 1300 FSB, the memory had to be set to 5-5-5 timings and everything left on Auto. The results were close to the Abit Fatality FP-IN9 which I can't hold completely against the NF650SLIT-A because the FP-IN9 comes with a factory set "supercharged" bus which automatically scales higher as the system is loaded. The good thing is, the motherboard carried the Quad Core Q6600 to 2.92GHz per core, and held true for more than enough time to certify its stability. The Q6600 used in this test has been tested to 3.05-3.1GHz with very limited success above, so this is pretty good for a budget 650-SLI board quad capable board. The E6600 chip reached almost the same results hitting the same Front Side Bus maximum.

Benchmark scores are pretty close across the spectrum. While the FP-IN9 scores just a few points higher in one test, it drops behind in a couple others. The NF650SLIT-A pulls ahead in the SANDRA Mathematical scores across the board where it doesn't pull completely ahead in the Multi-Media tests. Still, the two boards are nearly in a dead heat. Next up, you can see the real world 3D performance.

Got News? Send 'em in!