What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

G.Skill Ripjaws F3-16000CL9D-8GBRM 2x4GB Memory Review

Status
Not open for further replies.

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Hardware Canucks has just rung in the new year, and how fitting to do so with a 2x4GB kit of memory from G.Skill specified for the Intel P55 platform. Towards the end of 2010, memory manufacturers put on their density hats and made huge leaps and bounds in the high performance/high density DDR3 segment of the market. PC3-16000 kits of memory that run at DDR3-2000 with 2GB modules are quite common, and now 4GB modules able to run at these high frequencies are everywhere.

Hardware Canucks have done all kinds of investigative testing into high frequency memory performance on overall system speed, and have found that there just isn't a huge benefit to going with high clocks. There is also a pretty decisive limit in how high memory clocks can even go on a 24/7 system. This limit was seen more so with Bloomfield Intel CPU's due to memory controller limitations, but even with P55 CPU's that can handle higher memory clocks, there are still limits. It only makes sense then to increase density, as there is really no other place to go with DDR3 memory at this point.

What this realization has led to is a massive drop in price in high density DDR3 memory that can still run high frequency. Today we look at an extremely well priced, and well spec'd 2x4GB kit of DDR3 memory from the G.Skill Ripjaws family. Rated at DDR3-2000 with 9-10-9 timings, this kit is not ground breaking in the sense of the typical marketing number game. Instead, this kit offers high density and high speed at a very reasonable price. Add in the stock 1.55v voltage rating and we have ourselves a decent little kit of memory in the G.Skill F3-16000CL9D-8GBRM.</p><center><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/index-1.jpg" alt="G.Skill Ripjaws"></center>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Specifications

Memory Specifications

<p style="text-align: justify;">Memory specifications don't really need a whole page, but the G.Skill F3-16000CL9D-8GBRM Ripjaws have XMP profiles so we'll take a look at those and see how this kit sets itself up when dropped in a compatible motherboard.</p><center><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/specs-1.jpg" alt="G.Skill Ripjaws"></center><p style="text-align: justify;">G.Skill has specified this kit to run at DDR3-2000 9-10-9-28 2T which is actually kind of pedestrian considering some other 4GB kits that have been released, but G.Skill has binned these modules for operation at just 1.55v. G.Skill also offers a full lifetime warranty on this kit. Let's now take a quick look at how the XMP profiles boot this kit on a Gigabyte P55A-UD7.</p><center>
specs-2.png
specs-3.png
</center><p style="text-align: justify;">What's interesting is the discrepancy between the XMP profile that CPU-Z reads, and the actual timings that the XMP profile sets. In the XMP profile on the left tRCD is listed at 11, which is obviously incorrect according to G.Skill’s specs, but the memory was set with the correct tRCD when the XMP profile was booted; as seen in the screen above to the right. This could very well be just an issue of CPU-Z reading the XMP profile incorrectly somehow.</p><center><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/specs-4.png" alt="G.Skill Ripjaws"></center><p style="text-align: justify;">Looking at the voltages that the XMP profile sets is nice to see 1.300v for VTT voltage. Often XMP profiles far overshoot VTT which can be dangerous, 1.30v is a good value that most Lynnfield CPU's should be able to run DDR3-2000 at. The vDIMM is set to 1.60v however, despite the kit being rated for 1.55v. This could again be the motherboard not co-operating with the XMP profile. Keep in mind, the P55A-UD7 from Gigabyte isn't technically on the qualified motherboard list. Before we look at the heat sinks, let's take a quick look at what is underneath the heat sinks first.</p><center>
specs-5.jpg
</center><p style="text-align: justify;">The Hynix <b>h5tq2g83bfr -h9c</b> reference sheet can be found at the Hynix site here. Aside from a bunch of information that is of no interest to us, we found that these IC's are rated for operation at up to 1.575v. This tells us that these IC's are not higher voltage binned to run at lower volts, but designed for low voltage operation. The other key piece of information, designated by the H9 on the IC, is that they are rated for operation at 1333 with timings of 9-9-9. Considering G.Skill specifies these modules to run at DDR3-2000 with only a slight adjustment in tRCD to 10, it is likely that there is substantial binning of these IC's for this particular kit. It should be interesting to see how they overclock, and handle higher voltage.

We'll now take a look at the Ripjaws heat sink and a closer look at the modules.</p>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
A Look at the G.Skill Ripjaws Modules

A Look at the G.Skill F3-16000CL9D-8GBRM Ripjaws

<p style="text-align: justify;">I normally like to get upset with the manufacturer for putting modules in a package that can't sustain their weight, but I just can't do that with G.Skill.</p><center>
package-1.jpg
package-2.jpg
</center><p style="text-align: justify;">The Ripjaws don't have enough mass to their heat sinks in order to warrant anything other than a standard plastic blister pack. Unlike memory with heavy heat sinks that seem to fall out of the blister pack the second you open it, or often before you even open the package, Ripjaws sit securely in their blister pack. Heat sinks like the Ripjaws were the norm back when the blister pack became fashionable for memory, so it is no wonder the G.Skill Ripjaws modules are a suitable candidate for this type of package. Wow, I can't remember the last time I didn't complain about a memory package.</p><center>
package-3.jpg
package-4.jpg
</center><p style="text-align: justify;">As mentioned, the G.Skill Ripjaws do not have a bulky heat sink on them. They are pretty standard to be honest, but this isn't necessarily a bad thing. The thin aluminum heat sinks simply cover the eight IC's on each side and join at the top. The cooling they provide is going to be minimal, but they do offer a rather subtle but effective aesthetic quality.</p><center>
package-5.jpg
package-6.jpg
</center><p style="text-align: justify;">I am not sure what exactly a ripjaw is, but if one was to guess based on the teeth like extensions at the top of these heat sinks, one could guess that they were in fact meant to illustrate a tooth filled mouth. Obviously there isn't much else to say as aesthetics is a subjective thing, but I would prefer to see a heat sink such as this on memory instead of something over engineered, heavy, and difficult.</p>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
A Closer Look at the G.Skill Ripjaws

A Closer Look at the G.Skill F3-16000CL9D-8GBRM Ripjaws

<p style="text-align: justify;">Let's now take a closer look at the Ripjaws heat sinks, and what is underneath them.</p><center>
heatsinks-1.jpg
heatsinks-2.jpg
</center><p style="text-align: justify;">The first order of business is to identify the PCB's used on these modules. In addition to be being black, which serves nothing more than aesthetic purposes, the PCB is a 6 layer Brainpower design. Most contemporary performance memory is built around a quality PCB and that is definitely the case here. Brainpower has been the standard in memory PCB design for a lot of years now.</p><center>
heatsinks-3.jpg
</center><p style="text-align: justify;">Moving back toward the heat sinks, the photo above illustrates yet another benefit of the simple heat sink design. The last two memory reviews here at Hardware Canucks that involved large elaborate heat sinks, OCZ's Flex-EX and Corsair's Dominator GTX modules, both had mounting issues. That is definitely not the case with the G.Skill Ripjaws. Both sides of both modules have a near perfect application.</p><center>
heatsinks-4.jpg
heatsinks-5.jpg
</center><p style="text-align: justify;">To get the Ripjaws naked was pretty straight forward. The thermal pad used is adhesive, but not extremely sticky or tough to separate from the IC's. Again, another throw back to when heat sinks on memory was fresh and exciting. The thermal pad does transfer heat, but it isn't the most efficient method for cooling memory IC's. The key aspect of these heat sinks is that memory generally doesn't need additional cooling to perform at spec. Heat sinks are primarily for branding more than anything, and the Ripjaws do a good job of that, without sacrificing anything in return.</p>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Install & Test Setup / Methodology

Install & Test Setup

<p style="text-align: justify;">Before we take a look at the setup and do some overclocking, let's have a quick check to make sure that the Ripjaws heat sink fits underneath a Thermalright Ultra-120 eXtreme.</p><center>
install-1.jpg
</center><p style="text-align: justify;">As we anticipated, the Ripjaws have plenty of space under the TRUE. This is likely the largest benefit of a simple heat sink; unlike tall Flex-EX, or Dominator heat sinks, the Ripjaws will not likely interfere with any CPU cooling solution. I am all for memory cooling, but if it starts to interfere with CPU heat sink mounting, what is the point?</p><center><table cellpadding="0" cellspacing="0" width="735px"><tr><td align="left">
setup-1.jpg
</td><td align="left">
setup-2.jpg
</td><td align="right">
setup-3.jpg
</a></td><td align="right">
setup-4.jpg
</a></td></tr></table><br /><table border="0" bgcolor="#666666" cellpadding="5" cellspacing="1" width="735px"><tr><td colspan="4"><b><font color="#ffffff">P55 Test Platform:</font></b></td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Memory:</b></td><td align="left" bgcolor="#ececec" width="75%">G.Skill 2x4GB PC3-16000 9-10-9-28 2T (F3-16000CL9D-8GBRM)</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Motherboard:</b></td><td align="left" bgcolor="#ececec" width="75%">Gigabyte P55A-UD7 (BIOS F7)</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Processor:</b></td><td align="left" bgcolor="#ececec" width="75%">Intel i7 860 (925B478)</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Processor Cooling:</b></td><td align="left" bgcolor="#ececec" width="75%">Chilly1 SS (Tuned by Ruffus)</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Thermal Paste:</b></td><td align="left" bgcolor="#ececec" width="75%">Arctic Silver Ceramique</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>PCH Cooling:</b></td><td align="left" bgcolor="#ececec" width="75%">Stock</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>PWM Cooling:</b></td><td align="left" bgcolor="#ececec" width="75%">Stock</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Power Supply:</b></td><td align="left" bgcolor="#ececec" width="75%">Corsair HX1000W</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Video Card:</b></td><td align="left" bgcolor="#ececec" width="75%">Biostar 8600GTS</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Additional Fans:</b></td><td align="left" bgcolor="#ececec" width="75%">Scythe Ultra Kaze 120MM 2000RPM 87.6CFM (DFS123812L-2000)</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>Hard Drives:</b></td><td align="left" bgcolor="#ececec" width="75%">Seagate 7200.9 80GB SATAII 8MB cache</td></tr><tr><td align="center" bgcolor="#cc9999" width="25%"><b>OS:</b></td><td align="left" bgcolor="#ececec" width="75%">Windows 7 Ultimate x64</td></tr></table></center><p style="text-align: justify;">Our setup in use today is very straight forward, aside from the CPU cooling. I happen to run a single stage phase change unit on this setup all the time. Some may argue that it will skew memory overclocking results because of the colder IMC on the CPU. To that I say, I am testing memory not the CPU's IMC; if this were a processor review, then absolutely, colder than average temperatures will skew results. What we have done here—or at least to the best of our ability—is eliminated the limitation of the CPU in memory clocking. This will allow us to fully test what the memory is capable of. Let's quickly go over our overclocking methodology now.</p>

Methodology

<p style="text-align: justify;">A debate as to what is and isn't stable for memory and CPU's will go on forever, so we are not here to perpetuate that discussion. Instead, we simply want to outline what we feel is adequate testing in order to provide accurate overclock results to you the reader.<p style="text-align: justify;">All of the overclocks listed on the next page went through the stress testing outlined above. This may or may not equate to 100% stability on your system, but it is a very accurate gauge of what we could expect out of this particular kit of memory in a 24/7 machine.

Keep in mind, however, that the overclocking results of our kit being tested today relies on a number of variables including the memory itself, the setup, and the person behind the controls. Just because we reach certain overclocks with this kit, doesn't mean every pair of modules will achieve the same overclocks. Every stick of memory is different, we are presenting a single set of results from a single kit of memory.</p>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Overclocking Results

Overclocking Results

<p style="text-align: justify;">As we mentioned at the end of the last section, overclocking involves a lot of variables and every sample of memory off the production line is going to behave differently. With that said, we still like to really test the overclocking capabilities of the sample of memory we receive in order to give us a ball park, or at least an idea, of what a single kit of a given model will do. Our sample may be ahead of the curve, it may be behind the curve, it is hard to say. We just ask that you keep this in mind when looking at the results below.</p><center><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/overclock-1.png" alt="G.Skill Ripjaws"></center><p style="text-align: justify;">The first item of our agenda is the voltage we went up to. With 24/7 clocks, the question is always, how much is too much? With the last 2x4GB kit of memory we had through Hardware Canucks, we actually had a voltage limit due to a very low temperature tolerance. That wasn't the case with this kit as the thermal testing will show, but we still wanted to stay reasonable and stuck with the same 1.72v as the upper limit.

For the timing sets that we tested, we basically went with what the kit wanted to do. Unlike the OCZ Flex-EX, this G.Skill Ripjaws sample ran 7-8-7 relatively strongly easily running over 800MHz. From there, we see an almost uniform climb up to 8-9-8, and on to 9-10-9 with 100MHz jumps in stable frequencies with each timing adjustment. At 9-10-9 we wanted to test if tRCD was a limiting factor and the 40MHz we gained going from 10 to 11 indicated it was. The final 10-11-10 timing set was just to let it all hang out and see how high we could clock this kit.

For the most part this kit clocked about what we were expecting. Based on the previously mentioned Flex-EX kit we looked at a month ago, we were a bit surprised by the clocks at the low end, and a little disappointed by the clocking at the upper end. The IC's on this Ripjaws kit are different so the differences in behavior shouldn't surprise us. Overall though, the G.Skill Ripjaws F3-16000CL9D-8GBRM seem to have a decent amount of head room, as well as offering a wide range of usable timing sets to get the most out of your system; wherever that system is running at.</p><center><table align="center" bgcolor="#666666" cellpadding="5" cellspacing="1" width="90%"><tr><td align="center" valign="top" bgcolor="#ececec" width="25%">7-8-7
click for full size...
overclock-2.png
<br>840MHz 7-8-7-30<br>@ 1.70v</td><td align="center" valign="top" bgcolor="#ececec" width="25%">8-9-8
click for full size...
overclock-3.png
<br>940MHz 8-9-8-30<br>@ 1.70v</td><td align="center" valign="top" bgcolor="#ececec" width="25%">9-10-9
click for full size...
overclock-4.png
<br>1050MHz 9-10-9-30<br>@ 1.72v</td></tr><tr><td align="center" valign="top" bgcolor="#ececec" width="25%">9-11-9
click for full size...
overclock-5.png
<br>1090MHz 9-11-9-30<br>@ 1.72v</td><td align="center" valign="top" bgcolor="#ececec" width="25%">10-11-10
click for full size...
overclock-6.png
<br>1140MHz 10-11-10-30<br>@ 1.72v</td><td align="center" valign="top" bgcolor="#ececec" width="25%">9-10-9
click for full size...
overclock-7.png
<br>1002MHz 9-10-9-28<br><b>@ 1.56v (Spec)</b></td></tr></table></center><p style="text-align: justify;">Normally the overclocking section would end there, but since we had a couple of solid 2x4GB kits on hand we decided to see how they would play together. Below is a quick 8x32M run of SuperPi at a very nice 1060MHz @ 9-10-9-30 1T using the G.Skill Ripjaws F3-16000CL9D-8GBRM and OCZ Flex-EX OCZ3FXE2133LV8GK. The kits have different IC's but play quite well together.</p><center>
overclock-8.png
</center>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Bandwidth Benchmarks

Bandwidth Benchmarks

<p style="text-align: justify;">We won't be doing any exhaustive benchmarking with memory anymore. The impact on system performance is limited at very best. Instead we will focus on overclocking and the features that a particular kit offers. With that said, here is a list of Everest Bandwidth & Latency numbers that the overclocks on the previous page achieve.</p><center><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/benchmarks-1.png" alt="G.Skill Ripjaws"><br /><br /><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/benchmarks-2.png" alt="G.Skill Ripjaws"></center><p style="text-align: justify;">Bandwidth is as bandwidth does. CPU clocks can influence the bandwidth numbers so these results have to be taken with a grain of salt. We have no real direct comparison due to the nature of memory overclocking, but we can see that bandwidth scales pretty much as one would expect, but the fact of the matter is that these bandwidth numbers really don't translate to system wide performance.

The performance differences just aren't there. Modern enthusiast systems simply aren't bandwidth restricted so performance is slim to none in terms of memory overclocking. Obviously we have a very significant difference in bandwidth read and write when it comes to comparing the slowest and the fastest overclocks we achieved with this kit, but that just isn't going to equate to a lot of real world performance. Let's move on to have a look at the cooling abilities of the basic G.Skill Ripjaws heat sinks.</p>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Temperature Analysis

Temperature Analysis

<p style="text-align: justify;">Temperature analysis of memory modules has started to intrigue us after the temperature sensitivity of the OCZ Flex-EX modules was discovered. This section is very straight forward. We mount a K-type thermal probe to the side of an IC in the middle of the module, and then monitor temperatures during full HCI MemTest load with and without the heat sink, as well as with and without active cooling. We will use our top overclock of 1140MHz at 10-11-10-30 with 1.72v vDIMM from the previous section. Photos will be taken 15 minutes into an HCI run with an Extech TM200 digital thermometer will be used for measuring IC temperature as well as ambient temperature.</p><center>
temp-1.jpg
</center><p style="text-align: justify;">Above is a photo of the thermal probe mounted to the side of the IC. This will allow for consistent temperature measurement with and without the heat sink on. Ideally we would like to measure from the center of the IC, but then the thermal probe itself would interfere with the cooling ability of the heat sink.

In the photos below, the IC temp is T1 with ambient being T2. You will notice that ambient is quite low; this is not an error in reading, it is simply that cold in my testing area as I am going without heat this winter to raise awareness for homelessness in London, ON where I live. The only problem is, that while I was putting the heat sink back on the module after testing with the Ripjaws heat sink off, ambient temperature climbed a rather substantial 3C in about 15 minutes. Evidently the sun came up. Unfortunately we will just have to work with the results I achieved. We will first look at the temperatures utilizing active cooling from a 2000RPM 120mm Scythe Ultra Kaze fan overhead.</p><center>
temp-2.jpg
temp-4.jpg
</center><p style="text-align: justify;">Above to the left we have the temperature without the Ripjaws heat sink on the module reaching 22C with an ambient of 13C. We can see that with the Ripjaws heat sink mounted that the temperature climbs to 24C with an ambient of 16.4C in the image to the above right. With the discrepancy of ambient temp, it appears that with active cooling, the Ripjaws heat sink doesn't really provide any cooling benefit, but also doesn't appear to hinder cooling at all.</p><center>
temp-3.jpg
temp-5.jpg
</center><p style="text-align: justify;">Like above, the difference in ambient temperature appears to be providing the only difference between the temperatures with and without the Ripjaws heat sink attached when no active cooling is being used. This isn't really a surprising result as the Ripjaws heat sink is very basic, and as we mentioned, provides more of an aesthetic quality to the memory rather than a performance one.

What is a little surprising is how these Hynix IC's didn't fail with the high temperatures of testing without active cooling. For those that recall the Flex-EX review, the memory became unstable at around 37C. These IC's ran right through 15 minutes of HCI without spitting a single error at even 55C. These Hynix IC's are a little more versatile when it comes to temps so I am more than fine with the simple, yet effective, Ripjaws heat sinks on this kit. It is good to know that end users are not going to be paying G.Skill for an elaborate heat sink that is out-performed by a simple fan.</p>
 
Last edited by a moderator:

3oh6

Well-known member
Joined
Mar 18, 2007
Messages
1,049
Location
Edmonton, AB
Conclusion

Conclusion

<p style="text-align: justify;">We are starting to get a pretty good picture of what the various 2x4GB kits have to offer Intel P55 users. The G.Skill Ripjaws are the second 2x4GB kit we have looked at here at Hardware Canucks and we have to admit that we were pretty happy with what we saw. A little bit of a departure from the OCZ Flex-EX modules with their elaborate cooling solution, the Ripjaws are more subtle and simple, but not without some bite.

We are fully aware of the usefulness of a DDR3-2000 2x4GB kit when it comes to performance, but we are also aware that there is a market for such a kit. It is easier to justify purchasing a DDR3-2000 2x4GB kit when the price tag is reasonable, and at just over $150 CND, the F3-16000CL9D-8GBRM Ripjaws are just that: reasonable.</p><center><img src="http://images.hardwarecanucks.com/image/3oh6/gskill/f3-16000cl9d-8gbrm/conclusion-1.jpg" alt="G.Skill Ripjaws"></center><p style="text-align: justify;">The black PCB and subtle blue aluminum heat sinks lend a very simple yet elegant aesthetic presence to this G.Skill kit, and we like the fact that we aren't paying extra for an elaborate cooling solution. Our temperature testing, while not completely exhaustive, was still able to provide enough evidence to substantiate our claims of limited benefits from memory heat sinks. Our claims were only further enhanced when we were able to run the memory at much higher temperatures than anticipated.

The combination of appropriate cooling, solid overclocking ability that spreads out over a range of timing sets, DDR3-2000 spec with only 1.55v, and a price tag that is actually quite reasonable makes the G.Skill F3-16000CL9D-8GBRM Ripjaws a very attractive 2x4GB kit of memory for Intel P55 users in our opinion. The stated reasons above also allow us to give this kit our Hardware Canucks Dam Good award.</p>

<b>Pros:</b>
  • Low voltage for DDR3-2000 2x4GB kit
  • Versatile timing and clocking range
  • Can operate at high temperatures
  • Heat sink won't interfere with most CPU cooling solutions
  • Simple heat sink does its job, and doesn't increase cost of kit

<b>Cons:</b>
  • Overclocking wasn't quite equal with substantially more expensive 2x4GB kits
  • Limited availability in Canada with Newegg.ca being the only carrier we could find


<center><img src="http://images.hardwarecanucks.com/image/3oh6/dam_good-1.jpg" alt="Hardware Canucks Dam Good"></center>

<p style="text-align: justify;"><center><b><i>Hardware Canucks would like to thank G.Skill for making this review possible and supplying the memory used today.</i></b></center></p>


 
Last edited by a moderator:
Status
Not open for further replies.
Top