G.Skill ECO PC3-12800 CL7 Memory Review

[Eldonko]

G.Skill ECO PC3-12800 CL7 Memory Review​

Price: $125 - $135 CDN
Manufacturer Product Page: GSkill ECO PC3-12800 CL7
Manufacturer's Part Number: F3-12800CL7D-4GBECO
Warranty: Lifetime Warranty
Price: Click here to compare prices
Buy from: NCIX | DirectCanada | BestDirect



Established in 1989, Taipei, Taiwan based G.Skill International Co. Ltd. is a manufacturer of extreme performance memory and solid-state storage. Throughout the 1990s G.Skill was relatively unknown in North American markets; however over the past 10 years they have moved from relative obscurity to become a household name among enthusiasts. While only a few years ago G.Skill was tough to find in Canada, today G.Skill memory is available at major retailers including NCIX, Newegg, and Direct Canada.

Back in October of last year, G.Skill announced the release of their ECO low voltage, 1.35V dual channel DDR3 memory for Intel Lynnfield Core i5 and Core i7 platforms. In the press release, G.Skill stated the ECO memory has more power efficiency due to the modules needed 18% less VDIMM which in turn contributes to environmental friendliness and ultimately saves G.Skill customers money. In addition, G.Skill states that lower voltages decrease a system’s operating temperature, and according to their lab tests have revealed that 1.35V memory modules produce up to 16% lower temperatures than standard 1.65V memory.

From a pricing perspective, the ECO memory is positioned around the middle of G.Skill’s price spectrum at about $130 for a 4GB PC3-12800 kit. This price is above the value kits and below the top end enthusiast kits which use more expensive Elpida Hyper chips. G.Skill’s ECO line shouldn’t eat into their market share from their enthusiast lines such as Ripjaws, PI, and Trident since G.Skill is targeting a niche market that is interested in saving some power without sacrificing much performance.

This all sounds very intriguing so we couldn’t wait to get our hands on a kit of ECO memory and run it through some rigorous tests. Well that time has come and Hardware Canucks is pleased to present a review of one of the top products from the ECO family: the DDR3 1600 4GB (2GBx2) CL7-8-7-24 1.35V kit. In this review we will see for ourselves how the ECO clocks at the low 1.35V, how it scales with voltage above that, how much speed you have to sacrifice to run low voltages and how much energy is actually saved in the end.

 

[Eldonko]

Specifications

Specifications

Before getting into all of the pictures and testing, let’s take a brief look at the specifications for the G.Skill ECO PC3-12800 CL7 memory. Specs for the particular kit being tested in this review are as follows:

Main Board: Intel
System: Desktop
System Type: DDR3
M/B Chipset: Intel P55
CAS Latency: 7-8-7-24-2N
Capacity: 4GB (2GB x2)
Speed: DDR3-1600 (PC3 12800)
Voltage: 1.35 Volts
Registered / Unbuffered: Unbuffered
Error Checking: Non-ECC
Type: 240-pin DIMM
Warranty: Lifetime

G.Skill also provides a list of motherboards tested to be compatible with ECO PC3-12800 CL7. This does not mean boards that are not on the list are not compatible but rather these are just the boards specifically tested to be compatible by G.Skill. The most common issue you will run into with a board not on the list is the minimum VDIMM available in the BIOS may be 1.55v. This means without the latest BIOS and a supported board users will not be able to use the spec voltage of 1.35v.

Now you may notice there are no AMD boards on the list. Again, this does not suggest the memory will not work on an AMD platform, on the contrary it will be fine on most. ECO PC3-12800 CL7 is just optimized to run on Intel LGA 1156 platforms.

The G/Skill ECO family of memory consists of six different kits, all 1.35v spec, but range from CL 7 to CL 9 and 1333 to 1600 Mhz. <img src="http://images.hardwarecanucks.com/image/eldonko/GSkillECO/Images/ECO1.jpg" alt="" style="float: right; margin: 4px 0px 4px 7px">


G.Skill ECO family of memory:

• G.Skill DDR3 1333 4GB (2GBx2) CL9-9-9-24 1.35V
• G.Skill DDR3 1333 4GB (2GBx2) CL8-8-8-24 1.35V
• G.Skill DDR3 1333 4GB (2GBx2) CL7-7-7-21 1.35V
• G.Skill DDR3 1600 4GB (2GBx2) CL9-9-9-24 1.35V
• G.Skill DDR3 1600 4GB (2GBx2) CL8-8-8-24 1.35V
• G.Skill DDR3 1600 4GB (2GBx2) CL7-8-7-24 1.35V

 

[Eldonko]

A Closer Look at G.Skill ECO PC3-12800 CL7

A Closer Look at the G.Skill ECO PC3-12800 CL7 Memory

In this section we will take a closer look at the packaging for the G.Skill ECO, the memory itself and how well the spreaders contact the memory chips. The ECO comes in a plastic package, common to many types of memory.

​

The marketing on the front package draws attention to the spec voltage of the memory (1.35v) and G.Skill specifically notes that the memory is designed for socket 1156 Intel i5 and i7 CPUs. On the reverse of the package we see highlights for power efficiency and environmental friendliness as selling points and also mention of lower temperatures and improved stability.

​

The plastic packaging fits snugly around the memory and you can clearly see on the label exactly which kit you are getting. G.Skill also provides contact information on the package in case you need support; however the phone numbers will have you calling Taiwan which isn’t optimal in our books. Other phone numbers for specific regions can be found on G.Skill’s website.

​

The ECO series seems to use a rather minimalist design; no flashy giant heatspreaders here! All we have is a basic grey heatspreader with G.Skill’s logo and the spreaders are about as small and compact as it gets.

​

The G.Skill logo and design on the heatspreaders has a metallic look with an edged texture. This finish gives the memory a modern looking silver shine and when held at the right angle, light catches the edges to provide a flashy contrast to the grey.

​

The label on the memory is typical of most and provides a part number, part description, timings, and spec voltage.

​

The heatspreaders really fit tightly to the memory chips on the G.Skill ECO PC3-12800 CL7 and we have no concerns about or heat getting trapped in gaps. Low voltage operation and tightly fitting heatspreaders give the ECO a very low operating temperature as you will see in a later section.

​

Since the ECO uses regular-sized heatspreaders, we see no issues with heatsink clearance and the memory should fit fine on any board you have. As you can see, there is tons of room around the memory on our test board; the EVGA P55 FTW.

​

We did not take the risk of damaging the memory by removing the heatspreaders to reveal the chips but we have a feeling we have a good idea of what is under the hood so to speak. More on that in the Overclocking section!

 

[Eldonko]

Test Setup and Methodology

Test Setup and Methodology


Test Setup

Out test setup consists of a socket 1156 Intel i7 platform with a EVGA P55 FTW motherboard and an Intel i7 860 CPU. Since the G.Skill ECO memory is made to be used on this platform we should have no compatibility issues.

​

Testing Methodology

The following section shows the maximum stable overclock achieved on G.Skill ECO PC3-12800 CL7 memory using a variety of voltages and timings. For testing methodology three main stability tests will be used and along with a variety of benchmarks. The first of the two main stability tests will be Linpack (LinX version 0.6.3) with memory usage set to 3,072MB and 25 loops run. In the enthusiast world, Linpack is a benchmark designed to measure performance on Intel CPUs in GFlops. However, it's also a very useful tool for checking the stability of a CPU and memory. LinX picks up memory errors very quickly and if you are able to complete a 25 loop test with the specifications above your system is likely stable or very close to it.

The second stability will be a minimum of three hours of HCI memtest using all memory available. This will pick up any memory error that LinX may have missed. On top of that, the third stability test will be 3 runs of 3DMark Vantage. This tests the 3D stability of the overclock as well as CPU, BCLK and memory. Once an overclock passes these tests but fails anything further, this is the point deemed as “stable” for the purposes of this review.

The EVGA P55 FTW (657) BIOS used for testing is E657_A51, released on December 29, 2009. This BIOS allows users to select VDIMM lower than 1.55v so worked great for our testing. Our i7 860 test CPU has been run well over 1000Mhz with other memory kits so we can confirm that the CPU's memory controller is not limiting our overclock with the ECO.

A number of combinations of memory timings will be tested at voltages of 1.35v, 1.45v, 1.55v, and 1.65v. Actual voltages as measured by digital multimeter are as follows:

 

[Eldonko]

Overclocking Results

Overclocking Results

Introductions, specs, and pictures behind us we can now to get to the meat of this review and take a look at how the G.Skill ECO PC3-12800 CL7 overclocks! When beginning a testing session it is always useful to know what chips the memory is using so you know which timings are going to work and which will not. Right now memory for P55 chipset Intel i5 and i7 platforms is usually Elpida-based and uses either BBSE or the more expensive Hyper chips.

Looking at the price of this kit, we can rule out Hyper and with binning at 1.35v 800Mhz and timings of 7-8-7 we could be looking at a number of different options. The chips could be either low voltage binned BBSE, another flavour or not even Elpida for that matter, maybe some Samsung 40nm that are rated at 1.35v or even something like Micron or Hynix. Also from the range of results we have seen from other users of this kit, G.Skill could be using a variety of different ICs for the memory. Either way we will keep the chips used in the ECO as a mystery for you to figure out from our test results.

As we began testing we found that TRCD was the main timing that would get you more clocks. We were a little surprised to see increasing CAS or TRP over 6 did absolutely nothing for max stable speed on this memory and above 6 was counterproductive since subtimings would loosen with CAS. For voltage the ECO scaled great above 1.35v up as far as about 1.58v and then we had diminishing returns above that. We have seen some other tests of the same kit scale up to 1.65v but ours topped out at 1.58v max so we decided on a max of 1.55v for our testing. Moving up by increments of 0.1v, the graph below shows how the ECO scales for 1.35v, 1.45v, and 1.55v as we increased TRCD from 6 to 9.

As you can see from the results above the difference in speed you can achieve going from 1.35v to 1.55v is huge, 160Mhz in some cases. This makes you wonder why a user would want to run this memory at 1.35v and sacrifice so much speed so we will take a closer look at differences in power consumption and temperature between 1.35v and 1.55v in the Temperature and Power Consumption Analysis section.

We would also like to add that this memory is definitely benchable at higher speeds, but for LinX and HCI Memtest stability at full memory usage the above chart shows where this kit maxes out. Loosening secondary timings like TRFC, Rank to Read Delay, and Refresh to Refresh gains you a few extra Mhz as well (maybe 10 at the most), but the performance lost from looser secondaries makes it not worth it in the end.

Detailed screenshots of the ECO scaling at 6-6-6 are below with 1.35v on the left and 1.55v on the right (click the image to enlarge). TRCD = 6 really didn’t get us far, with a minor gain of 19Mhz going from 1.35v to 1.55v.

​

TRCD = 7 scaled much better than TRCD 6 and we managed to gain 71Mhz from 1.35v to 1.55v.

​

The memory is specified for TRCD 8 so we weren’t surprised to see the memory start to fly with a little extra voltage. We gained 126Mhz from 1.35v to 1.55v. We should note that the ECO does not run the 800Mhz at 1.35v at spec timings of 7-8-7 2T if you have BCLK overclocked past a stock speed of 133. We needed 1.45v to get to 800Mhz, 1.35 got us to only 754Mhz which is 50Mhz short. Since the specified speed of these modules is stable at stock a BCLK we see no issues but it is important to remember our issues if you are overclocking the ECO modules.

​

TRCD 9 really was the key to speed with the ECO modules and this is also where we saw the biggest difference in speed from 1.35v to 1.55v: a gain of 159Mhz! Increasing CAS and TRP got us nothing above 967Mhz so we stuck with 6-9-6.

​

We found it a little strange G.Skill would spec this memory at 7-8-7 when 6-8-6 would get us the same clocks but it is likely because a variety of chips are used for this kit and some may not clock to spec speeds with CAS 6. Overall the ECO memory is able to produce a decent overclock and both enthusiast and mainstream users should find some advantages with G.Skill ECO RAM.

 

[Eldonko]

Timing Comparisons

Timing Comparisons

Considering we are able to run CAS and TRP 6 to 8 without a problem, let's take a look at which is best and why. Using Super PI we will test 6-9-6 vs. 7-9-7 vs. 8-9-8 at our max speed of 967Mhz.

It turns out that 6-9-6 and 7-9-7 are exactly the same in terms of efficiency in running SuperPI. For these runs, subtimings were kept identical to avoid any differences based on that.

​

Increasing CAS and TRP to 8 for 8-9-8 we saw a slightly slower PI time. The reason for this is that Read to Write timings (different DIMMs, different rank, same rank) also had to be increased from 8 to 10 to maintain stability. These settings have a relationship with CAS latency so you can't loosen one without the other following.

In the end we have determined that CAS and TRP have no effect on performance except when loosened to a degree where other secondary timings have to follow. Your best bet is to stick with CAS and TRP at 6 for this memory. However, we should also say that if you do purchase this product, you may have different ICs than the kit of ECO tested here and results may vary.

 

[Eldonko]

Memory Benchmarks

Memory Benchmarks

Sandra Memory Bandwidth and Memory Latency

SiSoftware Sandra (the System ANalyser, Diagnostic and Reporting Assistant) is an information & diagnostic utility. The software suite provides most of the information (including undocumented) users like to know about hardware, software, and other devices whether hardware or software. The name “Sandra” is a (girl) name of Greek origin that means "defender", "helper of mankind".

The software version used for these tests is SiSoftware SiSoftware Sandra Professional Business 2009.SP4 and the two benchmarks used are the Memory Bandwidth and Memory Latency. These benchmarks were chosen as they provide a good indication of memory performance and will tell us which settings are most effective. The bandwidth test shows performance of memory sub-systems and the memory latency test shows performance of the memory controller.

Results: Using the max clock speeds at 6-6-6, 6-7-6, 6-8-6, and 6-9-6 at 1.55v we see that both memory bandwidth and memory latency improve with speed. This tells us that maxing out your memory at 6-9-6 will yield better performance in most applications than using lower TRCD. Increasing CAS and TRP just leads to looser subtimings and does not gain any speed so 6-9-6 appears to be optimal for our test kit of G.Skill ECO PC3-12800 CL7.

 

[Eldonko]

Power Consumption and Temperatures

Temperature and Power Consumption Analysis

We are going to start this section with a quote from G.Skill’s Press Release about ECO memory, it goes as follows:

“Compared to the current DDR3 standard voltage of 1.65V, G.Skill’s ECO memory modules require 18% less VDIMM. This can have a great contribution to power efficiency, environmental friendliness and ultimately saves G.Skill customers money. In addition, the lower voltages decreases a system’s operating temperature, resulting in improved stability. G.Skill’s own internal lab tests have revealed, 1.35V memory modules produce 16% lower temperatures than standard 1.65V memory.”

There are quite a few interesting claims in there so let’s test them for ourselves and see what kind of power and temperature savings we are looking at over G.Skill Ripjaws modules.


Temperature Analysis:

In order to test out G.Skill’s temperature claims we used our ECO test kit at 1.35v and put it up against a kit of G.Skill Ripjaws at 1.65v.

Memory speeds and timings were kept exactly the same to ensure consistency and only the voltage was changed. To measure the memory temperature we used a Fluke digital thermometer and the temperature probe was placed inside the memory heatspreader right up against the edge of a memory chip. For consistency we ran LinX and took the temperature reading at the 10 minute mark for both 1.35v and 1.65v.

​

Marking at the 10 minute point in LinX for the ECO at 1.35v we got a temperature of 30.5C, which in our experience is very low for any memory. Switching over to the Ripjaws at 1.65v and again marking at the 10 minute LinX point we see a modest gain in temperature to 34.2C, a gain of 3.7 degrees or about 12%.

​

While we didn’t get the 16% lower temperature that G.Skill claimed, 12% is very close and it is not a stretch to believe their numbers are achievable. We are also unaware of how G.Skill measured the temperature delta so some other factors could be influencing the tests. Any way you look at it 30.5C is an exceptionally cool temperature for fully-loaded and overclocked memory to run.

Power savings will be another area where the ECO memory earns its place in a PC, let’s take a closer look.


Power Consumption Analysis:

The temperature analysis gave us a decent temperature delta between the ECO at 1.35v and the Ripjaws at 1.65v, now let’s take a look at how much power we are saving by running the ECO at 1.35v.

To show the difference in power consumption we are measuring the wattage being drawn by the full system from the wall with a Blue Planet EM100 Energy Meter. First we measured the power consumption for the ECO at 1.35v at full load LinX and then switched over to the Ripjaws at 1.65v to see exactly how much power would be saved. System speeds and all other voltages were kept the same. Power consumption for ECO at 1.35v was 292W and Ripjaws at 1.65v gave us 300W for a savings of 5W or 3% of total system power consumption.

​

Sure “a great contribution to power efficiency and environmental friendliness” is open to an individual’s interpretation; and from an enthusiasts perspective 8W is nothing to worry about when he is shooting for the max overclock possible. However, for a mainstream user that uses features such as SpeedStep and runs a low power system at stock speeds, 8w is a big deal and the choice between 1.65v memory and 1.35v should be quite clear.

 

[Eldonko]

Conclusion

Conclusion

With the “go green” movement in full swing, G.Skill has been innovative in picking a relatively untapped market segment to focus on for a new line of memory. This is especially true with new power-efficient CPUs around the corner that will utilize memory power savings to a much greater extent. With the ECO line, G.Skill has the environmentally conscious market niche at their fingertips with no competition anywhere to be seen. Throughout the course of this review we took a close look at the memory, how it clocked, as well as power consumption and temperatures. Now let’s chat about what it all means for a perspective buyer of ECO memory.

In terms of looks, the ECO line has a rather minimalist design with simple grey heatsinks which aren't anything special to look at but G.Skill still managed to keep the memory looking sharp with the silver designs and logos. The positive side of these heatsinks is that the ECO kit should fit in about any setup since are no giant heatspreaders or fan systems to worry about interfering with your CPU heatsink.

Overclocking results were decent overall considering the market this kit is targeted at and with a little extra voltage the ECO kit can fly along at speeds nearing 1000Mhz. On the other hand, we were a little curious about the specified timings of 7-8-7 since 6-8-6 works just as well. This discrepancy could be because G.Skill is using a few different types of chips for this line and some of the other ICs may not function well at CAS 6. We did have a bit of an issue running 800Mhz at 7-8-7 using overclocked BCLK and it took 1.4v instead of 1.35v which isn't a cause for concern but it is an interesting wrinkle in the fabric of this kit. However keeping BCLK at a stock speed of 133, we were able to run this kit at stock speed and timings at 1.35v fully stable so we don’t see this as a big issue. When overclocking, if you have a problem running your ECO kit at 1.35v and 800Mhz you can always up voltage to 1.4v or up TRCD to 9 with minimal effort.

Memory temperatures for the ECO memory were nothing short of outstanding with full-loaded temperatures of 30.5C. To put that into perspective, human body temperature is 37C so ECO has 6.5C on you! Power savings were also decent when compared to a 1.65v kit at the same speeds, a difference of 8W. 8 watts over a long period will definitely add up in terms of savings to your power bill even though it isn't anything dramatic. Nonetheless, it is great to see G.Skill's claims are backed up by actual results.

Whether an efficient kit such as this one is worth it or not depends on what type of user you are. Enthusiasts are usually just after the maximum allowable overclock and power savings are the last thing on their minds. However, a mainstream user generally does not overclock in any way so fast memory at stock speed and low power consumption is a winning combination. People running Folding@home systems which sometimes crunch away 24/7 can also see some serious benefits with a kit like this.

Considering G.Skill's ECO memory is targeted towards mainstream, environmentally conscious consumers and G.Skill has other lines like Ripjaws which are built for the enthusiast market we have to say G.Skill really hits the mark with this one. ECO memory cashes in first on an emerging market and has little to no competition at this time. For that reason we are pleased to give G.Skill the Hardware Canucks HWC Dam Innovative award!


Pros

- Exceptionally cool-running memory
- Saves energy conscious users some power
- Decent overclocking headroom at 1.55v
- Sharp looking heatspreaders that wont interfere with large CPU heatsincs
- Lifetime warranty


Cons

- At 1.35v overclocking is limited

Thanks to G.Skill for making this review possible!​

G.Skill ECO PC3-12800 CL7 2x2GB Memory Review Comment Thread​