|Overclocking and tweaking What's that smell?|
G1.Sniper 3 Review (PLX8747vsNF200, Circuit Analysis, LN2OC, Core3D...)
Z77 High-End Finally Has Arrived: GIGABYTE G1.Sniper 3
Every launch we see a couple one of a kind motherboards; in terms of features, innovation, and performance they just blow away everything else out of the water. GIGABYTE’s current crème de la crème is the G1.Sniper 3 and if its legacy has anything to say about it, then it really will take the spot at the top of the Z77 totem pole. Following G1 tradition, GIGABYTE has equipped the 3rd Generation G1 Killer: G1 Sniper 3 with a real Creative audio chip and Killer NIC, however GIGABYTE has also added a lot of first time things to the Sniper 3 which no previous G1 series motherboard had. User feedback helped GIGABYTE form what they think is a more appropriate G1 motherboard, with more conservative heatsinks, a lower price, better multi-GPU capabilities, and true overclocking features you can safely assume that the G1.Sniper 3 resembles an OC board just as much as a pure gaming board. The G1 Sniper 3 really is a one-of-a-kind Z77 board. Many people believe in the concept of the third outcome being the best, and they might not be wrong in that respect. Without any further delay let’s take a look into the history of the G1.Killer family, and see if that concept has any merit to it.
I will also do some testing with the PLX Technology PLX8747, and test it against native Z77, native Z68, the G1.Sniper M3, and finally the Z68X-UD7 which has an NF200. So PLX8747 versus NF200, that comparison will be done here with 2-way SLI. Also to top it off, I will take the Sniper 3 for some LN2 runs and report back, as well as share some OC profiles at the end for any owners and potential owners to use.
So let’s start with a brief history of the G1 Sniper 3 series, following the Table of Contents:
Here is a brief history of the G1.Killer series, if you are interested in buying this board, perhaps you want to know what the original and second generation G1.Killer boards offered and resembled.
Box, Accessories, and Board:
The box has a very cool green militaristic tint to it, much different than the white GIGABYTE boxes we have been seeing lately. If you have been a G1 series fan then you will probably notice that there are a bunch of new features on the front and back of the board.
The accessories package is extremely large, we have stickers, two DVDs (one for the drivers for the motherboard, the other for the WIFI/BT card); we have the WIFI+BT card and its USB cable, as well as two antennas, one for WIFI and the other for Bluetooth. You might also notice that with X79 we saw these same WIFI/BT cards, but they had blue PCBs, now they have black PCBs.
The backpanel I/O as well as 6 black SATA6GB/s cables, a USB 3.0 front panel bay, 4-way SLI bridge, 3-way SLI bridge, 2-way SLI bridge, and an internal SATA to backpanel eSATA to SATA conversion kit. The conversion kit provides the ability for the user to make eSATA backpanel ports out of internal SATA ports it also provides the ability to move a MOLEX connector externally and provides eSATA to SATA cables, so you can make your internal SATA drive and external SATA drive.
Total Accessories Package:
Sin’s Take: This eSATA conversion kit is very cool, but considering the board has no eSATA, I think its inclusion is well justified. I think it’s really cool that GIGABYTE includes the stickers and poster, I am sure stuff like that can cost more than a cable or two, especially because they are custom made, but I have heard that a lot of guys enjoy having the stickers as there are a variety of different designs. I used some of them, but not on a case, rather my LN2 thermos needed some decorating (G1 Style).
This is the majestic black and green G1.Sniper 3. The next step should be making the ports UV reactive. The heatsink style is very conservative and not as stylish as the Z68 GIGABYTE (non-G1) boards, but when the Sniper 3 is installed in a case you won’t really notice the difference. The lack of heatsinks molded out of gun parts is welcomed by many (I could have gone either way honestly). The gold skull and knife on the PCH heatsink brings together the motherboard. If you see the same board without the little emblem it would look a bit plain, but with it, it looks much more finished. It is connected to the heatsink with a very strong adhesive, so you can pry it off with a crowbar if you don’t like it (and void your 3-year warranty).
The G1.Sniper 3 has 5 fan ports, all 4-Pin, and they are controlled in sets of two with SMART, so the CPU and the system fan connector located right on the other side of the socket can be controlled together.
Sin’s Take: In my opinion the 3 on the top half of the board are located very well, but instead of having the two at the bottom right next to each other, I think it would have been better if one was moved north, on the left side of the PLX bridge heatsink right above the first PCI-E 16x slot. I also want back individual fan port control, but this grouped control supposedly is much easier for the end-user to use, I guess I am just too advanced.
Here is the I/O panel, everything you need is here, however some people don’t like the fact that there are so many video outputs. The reason for all the video outputs is for Lucid MVP which is done best with the output being the motherboard instead of the GPU, as MVP’s main benefit is allowing the iGPU to do the output buffering. I personally do not use anything other than USB for my connectivity, but some people use 1394A and eSATA, for that GIGABYTE has provided two internal headers for 1394A as well as an eSATA conversion kit to provide eSATA ports, as there are a lot of internal SATA headers.
Sin’s Take: I also think that GIGABYTE possibly could use less outputs for the video, and instead provide two or three adapters, but then again the Z77 iGPU can support up to 3 monitors which some people will take advantage of.
Layout and Design:
The G1 Killer series motherboards are always well laid out. They GPU spacing is always at optimal levels, and since the G1.Assassin(X58) GIGABYTE has made sure to not make any of them ATX-XL which would require a special sized case. Instead boards like the G1.Sniper 3(Z77) and G1.Assassin 2(X79) are designed in the Extended-ATX (EATX) form factor. The EATX form factor is the same length as the ATX form factor but slightly wider, which shouldn’t pose a problem to 99% of ATX cases as most of the time, present included, the EATX boards don’t require extra screw holes.
The VRM is made up of 15 phases, 12 phases provide VCore for the CPU, 1 phase provides the VCCIO(VTT) and VCCSA(IMC) outputs for the CPU system agent and memory controller, and 2 phases provide the iGPU voltage. The G1.Sniper3 also has a PLX8747 which takes 16X PCI-E 3.0 lanes from the CPU’s PCI-E controller and provides the GPU/PCI-E slots with 32X PCI-E 3.0 lanes. As you can see there are 3 fan headers in the upper half of the Sniper 3, the two on either side of the CPU socket are tied together in the BIOS for PWM fan control and work off of the CPU temperature. This is great if you are going to use both fan headers to power cooling fans for the CPU’s heatsink, which most people will no doubt take advantage of. The VRM heatsinks are of a one piece design, and this limits their styling, but also helps make them much better at cooling components. The memory DIMMs are color coded to help users determine where to put their modules (use green slots first). There is also an OC area which will be discussed a bit later. We also have one more thing of interest, the USB 3.0 internal connector is well situated right behind the ATX 24-pin connector, which works great with the front panel bay, as it is in an optimal position.
Here we see the PCI-E slots, which are laid out for 4-way SLI/CF and can do 4-slot spacing for 2-way SLI/CF. There are 10 total internal SATA ports; two white are Intel SATA6GB’s, four black are Intel SATA3GB’s, and 4 gray are Marvell SATA6GB/s. GIGABYTE has also provided an SATA power connector, this power connector provides the PCI-E 16x slots with extra power for larger multi-GPU arrays. We also see an mSATA connector, as well as a bunch of internal headers. There is also a PCI slot which is not native to the Z77 PCH, but is provided for users who have PCI sound cards. There is a dual BIOS switch, which can be used to switch between two BIOSes.
Sin’s Take: The position of the front panel headers should be switched with the position of the TPM module connector, making it easier for the jumpers to reach, as I know many more people use the front panel headers than the TPM header.
I made this picture for those who want to know what the G1.Sniper 3 has to offer in terms of overclocking features, and there are many. Just like the UD5H and UD3H the G1.Sniper 3 possesses buttons (power, reset, and clear CMOS), voltage read points, and a POST Code display. The location of these buttons is optimal for any extreme overclocker as the features won’t be covered by GPUs and most benchers bench with their boards positioned horizontally, not vertically which means this little OC area is in a perfect position as the user is closer to the CPU for extreme cooling while also being able to monitor each GPU. The Dual BIOS switch is in a bit of an odd position; I would prefer it closer to the rest of the OC features. The PCI-E power connector (SATA power) is in a very good position as well, easily accessible and it doesn’t get in the way of the GPUs. The strong VRM on the Sniper 3 is identical to that of the UD5H.
Sin’s Take: This is the first G1 Killer motherboard which has OC features, and that is a big deal as GIGABYTE is blending the gaming and overclocking board a bit more than before. Only time will tell where this will lead us.
Here we have the specialized G1 Killer features, the Creative chip and the Bigfoot chip. Now if you can’t spot the creative chip, it is that little IC in the center of the picture, now unlike many other motherboards, the G1 Sniper 3 actually has a real Creative chipset, in this case the Recon3Di (CA0132). This little chip replaces the XF-I provided by the CA20K2 as well as all the other components like external DACs and ADCs. However GIGABYTE has provided on-board line drivers which are amplifiers. GIAGBYTE also provided a few Nichicon MUSE series audio capacitors and an EMI shield. Now it is safe to say that if a motherboard on the market currently for Z77 says that it has XF-I, then you can automatically assume they are referring to software emulation with a Realtek ALC892 or ALC898, the Z77X-UD5H and Maximus 5 Gene both have this software emulation with ALC898 ICs. However now that the G1 Killer series is using Core3D/Recond3Di, Core3D isn’t even in the XF-I family, and it has its own software suite.
BigFoot Networks(BFN) was bought out by Qualcomm/Atheros and since has come about a redesign, such that the BigFoot IC is a tiny IC with a “K” written on it. It is a miniaturized e2200, and is referred to as the e2201. The new IC reduces implementation and component costs and carries the same features and software.
Sin’s Take: A lot of people have some doubts about how the Core3D audio performs, so I have run some tests on the IC to see how it does in RMAA 5.5.
The only way to get 4-way GPU compatibility into an EATX form factor motherboard is to push the first 16X in the upper most position it can be, however that creates an issue for changing out memory modules while the first GPU is in the slot. On the G1.Sniper 3 this doesn’t seem to be an issue however, as I think GIGABYTE moved the DIMMs slightly north to compensate, and I can actually change out my DIMMs with a long GPU installed. Also users need not to worry about any interference with motherboard design and filling up all the GPU slots, as the board has been designed to accommodate GPUs in every slot.
The backside of the motherboard carries 1/3 of the MOSFETs used for the CPU and iGPU VRMs. The reason being because GIAGBYTE has used 3-MOSFETs per phase to increase power output and decrease temperatures, by putting an extra MOSFET on the back of the motherboard, and heat load is more evenly spread internally through the 2oz of copper in the PCB. Do not worry about these MOSFETs, they do not get hot, even under heavily overclocking. Users will be relieved to know that the heatsink hold down is done very nicely by screws and washers opposed to plastic push pins.
Let’s take a look at the heatsinks:
The heatsink design is very conservative, so as to attract a larger target audience. Perhaps GIGABYTE figured that the guns only attracted a smaller target audience, and thus adopted more widely accepted designs. You will notice that these heatsinks are the same style as those used on the Z77X-UD5H, except here they are a majority black, which makes them much more visually appealing. The neon green is a nice touch with the slots and DIMMs being color coordinated with the heatsinks. The gold/orange skull and knife actually match the color of the dual BIOS LEDs and the power button.
The paste on the PLX heatsink actually surprised me, maybe because it was actual paste, but it isn’t bad, and it is paste not a pad. Same goes for the PCH.
For the next segment of the review we have the board naked:
First of all here is the manual’s depiction of the motherboard’s connections:
A lot of people never open the manual, but there is much you can learn, GIGABYTE isn’t shy to people knowing how everything is connected.
The first thing I will talk about is my favorite, the voltage regulators. I don’t know why I like them so much, perhaps because there is so much variety and it reflects upon the build quality of the board. GIGABYTE has nothing to hide in this respect, the VRM is extremely powerful. Below is a list of the parts used:
On the Z77X-UD5H and the G1.Sniper 3, the IR3598 is used in doubler mode, effectively cutting maximum switching frequency to 600KHz, which is more than enough to facilitate WR overclocking. The GIGABYTE Z77 VRMs are extremely strong, run extremely cool, and have extremely good voltage regulation. GIGABYTE’s build quality shines through like rays of sunshine on a rainy day when it comes to the VRM.
That is how 16X PCI-E 3.0 can be changed to 32X PCI-E 3.0 and divvied up between 4 slots to provided 2-way at 16x/16x PCI-E 3.0, 3-way at 16x/8x/8x PCI-E 3.0, and 4-way at 8x,8x,8x,8x PCI-E 3.0.
Ways to implement PLX8747 and their Pros and Cons.
There is more than one way to implement a PLX8747 and produce different performance results, there are three options or actually 4 that come to mind.
1.Direct connect: All CPU lanes to PLX, and then to all slots. Pros: simple and effective, produces best multi-GPU results. Cons: Single card performance lower than without PLX.
2.Half Connect: 8x of the CPU’s PCI-E lanes are connected directly to the first PCI-E 16x slot, with the rest going to the PLX (still outputting 32x lanes even from 8x). This method can be used to help single card performance and also increase marketing by allowing the manufacturer to state that the board will have 8x more lanes than other boards with PLX, that is because the PLX will output 32x lanes whether you provide it 8x or 16x. Pros: Single card performance helped. Cons: Multi GPU (3 or more) performance may be hurt because PLX might have more delay than normal because of fewer lanes provided to it.
3.Single lane bypass: This method can give you the best of both worlds, it allows for one 16x slot or the PLX to get all the 16x lanes from the CPU at once. So one slot, can take all the CPU’s lanes, and then all the other PCI-E 16x slots will be disabled, or the PLX takes all lanes. This brings back single card performance to its best, and also provides the best multi-GPU performance from the PLX. Pros: best of both worlds, Cons: space for the switches, cost, and layout considerations.
The G1.Sniper 3 uses the direct connect implementation where all lanes go to the PLX.
The PLX8747 is a very new IC, used on any Z77 board which has PCI-E 3.0 and more than 16X PCI-E 3.0 lanes for the CPU; it is a novice in its field. It hinders the BCLK OC by ~1 MHz the maximum, I was able to reach 114.38 MHz, opposed to slightly more than 115 MHz on the UD5H, UD3H, and Sniper M3. So it is more than enough for anyone, even extreme benchers, as 7GHz (which is like winning the lottery) can be hit with 63x, 62x, or 61x 114MHz.
It isn’t as mature as the NF200, so support for it is pretty new, yet seems to be fine. I was able to run SLI without issues (two GTX570s), and before I had run 7970s in multi GPU configurations without too much issue. However if you are going to OC the BCLK, then the 7970s need to be run in PCI-E 2.0 mode.
Below is a list of the more interesting ICs on the motherboard, dealing with the audio and NICs.
We can see that GIGABYTE listened to its users because a lot of users asked for Intel NIC instead of BigFoot, so GIGABYTE included both! That is their way to please everyone, and bring in a new crowd who want dual NICs.
Here is a list of all the connectivity ICs on this board, all 8 PCI-E 2.0 lanes from the PCH itself are being used, and the VIA 1394a controller works off of one of the PCI outputs from the iTE8892E which provides PCI from PCI-E. That way GIGABYTE saves an extra PCI-E 1x lane for something else, and doesn’t waste an extra PCI lane. That is smart thinking, as I have seen other manufacturers not be that efficient in their designs, for instance using a PLX 4x chip to help consolidate 6 PCI-E 2.0 1x connections into 4x PCI-E 2.0 lanes as they had run out of PCI-E 2.0 1x lanes.
Here is how the USB 3.0 is done, as you are probably wondering how GIGABYTE is taking 4 native USB 3.0 ports and making them 10.
Now you might be thinking, hubs? Is that going to hurt performance? The truth is I tested with an SSD and USB 3.0 external drive enclosure. Let’s just say that while the other Z77 boards I tested topped out around 205-210mb/s the Sniper 3 was faster at around 225MB/s and that was using the internal header which uses the hub. If you fill the ports up, I am sure it might slow down. I wouldn’t worry, as I believe the G1.Sniper 3 has faster USB 3.0 than any other GIGABYTE Z77 board, perhaps its BIOS is more attuned to it. USB 3.0 tests are shown later in this review.
Below is the Z77 PCH:
The Z77 PCH, BD82Z77 (SLJC7) Platform Controller Hub (PCH), has a 6.7W TDP which is 0.6W higher than the Z68 PCH! This would account for the higher PCH temperature readings from the Z77 PCH; however this is nothing to be alarmed about. This PCH seems very capable of dealing with a heatsink with no active cooling. Compared to the Z68 chipset, the new Z77 chipset has the same number of SATA 6GB/s ports. In fact most everything is the same except for the fact that the Z77 Chipset new supports 3 independent displays, USB 3.0, and some newer Intel technologies like Rapid Start and Rapid Connect. The Z77 PCH provides 4 native USB 3.0 ports, which Intel says cannot work in Windows XP or Vista as USB 3.0, so users much switch their xHCI mode to “auto” from “smart auto”. This is very simple to do in the UEFI under the integrated peripherals section in advanced mode.
Here are the random extra ICs that didn’t fit in on the other two charts.
The GIGABYTE 3D UEFI really has some a far way, it grew from a laggy, slow, buggy platform to a nice, stable, easy to navigate platform which I actually like. Its features have been expanded, for instance you can load and save profiles for the BIOS settings from any other board with the same make and model. That means you can use any of the OC profiles I share later in this review, and apply them to your board for your OC, that is the easiest in BIOS stable OC, as I and others can OC better than auto OC programs.
GIGABYTE UEFi OC Profile Sharing:
GIGABYTE’s original UEFi didn’t even have the ability to assign names to the 5 total BIOS profiles one could save. However after a recent BIOS update, we not only have naming, but us Z77 GIGABYTE owners are able to save up to 8 profiles on the BIOS, as well as save BIOS profiles on a USB drive. Now one would think, can you share your profile with a friend, and surprisingly the answer is yes! I can make a profile on my G1.Sniper 3 and you can load it on your G1.Sniper 3!
Here is a video of it in action and a how to:
Here is a list of the profiles I have made:
Here is where you can download the zip, and then open and pick a profile you want to use. Please beware you should only use the BENCH profiles if you are subzero, as the stock BCLK is 105mhz, and vcore is 1.45.
Here are the profiles: http://www.mediafire.com/?hq0abbswifc6oag
Air Overclocking(LLC Testing Included)
A note on LLC and Software monitoring accuracy: So last time I did my reviews I tested out the LLC and compared software(easytune, cpuz, hwmon) to hardware readings(voltage read points and multimeter). It showed some interesting discrepancies which are good to note as you don’t want to over-volt your CPU or use more LLC than you should. Software voltage monitoring isn’t really accurate enough for precision voltage monitoring or testing. I also had requests from some to use more practical voltages than 1.45v to test LLC, as no user is going to use 1.45v 24/7. I used 1.45v to show how strain has an effect on LLC, but the results at 1.4v are the same as the algorithms are set very strong as the PWM and VRM can handle it. Also my CPU frequency is increased to increase the current used by the CPU and the frequency will have a big effect on LLC as well. That means that at stock, LLC extreme will raise your VCore more than it will at 4 GHz and 4 GHz more than at 4.5 GHz. That means that LLC high will droop more at 4.5 GHz than at 3.5 GHz. So just be aware of this.
Very nice results mate? Yes indeed the G1.Sniper 3 has the same LLC and voltage regulation as the UD5H, but I painted the background green. I added an extra column to show you the percent difference between idle and load, minus means the voltage drooped while positive means the voltage rose. What is interesting is that software shows much more droop than hardware monitoring, and hardware monitoring is 99% accurate while software is about 75%. So what does this mean? This means that setting LLC extreme until about 1.6v will raise your voltage under load, not keep it constant as CPUz reports.
Percent Difference = ((Load-Idle)/((Load+Idle)/2)
Percent Error is calculated = (Fake Load-Real Load)/Real Load
So let’s take a look at how well the new Sniper 3 overclocks.
Here is the maximum water/air overclock of 5.3GHz on my 3770K:
Booya! 5300MHz or 5.3GHz, is my maximum overclock using a Zalman AIO water cooler:
I think it works pretty well and stays quiet, not as good as H100 though. I am not stable at 5.3 GHz, I can do 5 GHz stable with a lot of heat, or I can settle for 4.8GHz stable:
The voltage I used for 4.8 GHz is nice. I will share the profile with you guys, so don’t worry.
Now let’s look at some of my air cooled memory OC results with a normal ULV kit from Kingston:
It is a simple 1.35v 1600mhz 9-9-9-24 T2 8GB(2x4GB) kit.
Now this memory is based upon Micron memory, and it does about 2200MHz max clocks, I didn’t spend much time on it but you can tighten those timings up a bit:
Also 1600 MHz with XMP but T1 instead of T2 works just fine:
If you are wondering why I used SuperPI32M for stability of memory, it is because it is a good memory stability tester. I used stock voltage (1.35v for the run above) and 1.55v for the 2200 MHz.
Liquid Nitrogen Testing Results:
So part of my reviews I am taking every board for some extreme liquid nitrogen overclocking.
This is the max LN2 CPU Frequency Overclock:
Here is the maximum BCLK Overclock:
Here are some benchmarks:
The board is pretty efficient in 2D Benchmarks.
Some LN2 OC Notes:
First off the Sniper 3 has pretty good cold boot bug behavior, I only triggered it twice throughout 5 hours of benching, and of course I have no cold bug. Better than the other motherboards I have tested. The G1.Sniper 3 has better restart behavior as well, and I was actually able to change the BCLK a lot with full shut down and not trigger the CBB. Now the Sniper 3 has the PLX bridge, so OCing with the BCLK with cards like the 7970 needs to have the PCI-E mode set to 2.0 instead of 3.0, but the 7970 and the ATI 7000 series in general have this issue. It seems to not persist on the GTX680s. It is a pretty good board for LN2 OC, I thought it worked well, and the CPU OC is just as high as on the other boards I have tested.
Now let’s see the testing, here is a quick shot of the test setup:
Test Setup and Benchmarks
I think you already have an idea of how the PLX performs, but below you will see how it does in multi-GPU configurations.
PLX8747 vs. NF200 vs. Native Z77 and Z68
This is what everyone was waiting for, the PLX 8747 versus the NF200, the new kid on the block versus the old and trusted. However why are we changing now? The answer is because the NF200 doesn’t support PCI-E 3.0. This must be something that really angers NVIDIA, as they just lost all possible sales of the NF200 (maybe sapphire might try it out).
Above are 4 boards used (I forgot to take a picture of the 5th, Z77X-UD3H). Now each board represents a specific platform configuration.
Name of board= PCI-E Controller, PCI-E configurations, Any Special Gaming ICs? , PCI-E Mode capability
Z77X-UD5H= Native Z77 platform, 8x/8x and 16x/0x, Basic Z77 IC configuration, PCI-E 3.0 capable
G1.Sniper M3= Native Z77 platform, 8x/8x and 16x/0x, Enhanced Z77 IC configuration, PCI-E 3.0 capable
G1. Sniper 3= PLX 8747, 16x/16x, 16x/0x, 16x/8x/8x/0x, 8x/8x/8x/8x, Enhanced Z77 IC configuration, PCI-E 3.0 capable
Z68X-UD3H= Native Z68 Platform, 8x/8x and 16x/0x, Basic Z68 IC Configuration, PCI-E 2.0 capable
Z68X-UD7= NF200, 16x/16x, 16x/0x, 16x/8x/8x, Basic Z77 IC Configuration, PCI-E 2.0 capable
All boards on the latest BIOS, however the issue is that I could only test at 3.5GHz, so to reduce the bottleneck of the CPU I lowered resolution very low as to help un bottleneck. The only test in which it seems to still be bottlenecked is Battlefield 3, however new BIOS for the Z68 boards are helping to allow Ivy Bridge Multiplier OC, UD3H is already capable, but the UD7 BIOS is in route. Once those BIOS come out I will try to expand the tests and run them again, as well as try better GPUs. However GPU configuration here will be important, as these GPUs aren’t bandwidth hungry like the new ones, and that could make a big difference(in favor of the PLX). The more bandwidth a GPU needs, the more it will like the PCI-E bridge chips and perform better with them compared to native. However even with my PCI-E 2.0 GPUs, we can see improvement in 2-way SLI on PLX versus native, which surprised me as I thought the gains would be much less. I ran all these tests a minimum of 3 times, and then I took the average for the average FPS and used the median score for the benchmarks.
Are you guys ready?
The battle has begun!
Creative Core3D (CA0132) Audio Testing:
This is the software:
A lot of people have been asking for RMAA 5.5 shots, so I ran RMAA 5.5 on the Core3D, in three modes, 16/48, 24/48, and 24/96. The stock I believe is 16/48.
It performs pretty well up to 24/96. It is hard to configure the Creative cards for RMAA, as they don’t play well together, but it scores extremely well in its native configuration, but not so excellently in the 24-bit/96KHz, which can show some weakness of the IC.
Read Life Sound: I actually like this little IC, its features are kind of cool, like the scout mode actually works can be configured to turn on or off with hotkeys. The sound is pretty good for gaming, I really like it, and the amps really do help. With gaming I look for two things while I play battlefield bad company 2 online (BTW I rock at it!) anyways, first of all the sound quality and second I look for the ability to hear which direction and distance the enemy is advancing from. With this IC I was impressed with the second, as scout mode does this thing where the really soft sounds of faraway foot steps are slightly amplified and you can hear the enemy and their direction. The sound quality was also very good with my Sennheiser PX-250 headphones (they have bass tubes, I know).
Sin’s Take: it is better than ALC898 in most everything, but not as good as the CA20K2 which was previously used on the G1.Killer boards. Now the implementation of a single chip opposed to 10+ really saves a lot of cost, and those savings seem to have been passed back. It is better than 99% of motherboard audio implementations and thus still is very good, and you don’t pay too much more for it, so it actually isn’t a bad idea. However it isn’t CA20K2, but it does excellent for gaming purposes. It seems however that at higher audio quality, the results aren’t as good as they could be, and this could be the IC’s limitation. Same results occur on the G1.Sniper M3 with the same audio and slightly different layout but same parts. However its sub 96kHz results are all fabulous. I personally don’t change my audio mode, so I hear excellent sound according to RMAA, but others might, and your sound is still very good, just if you want extreme excellence, you might want to go buy an audio card, as there is NO other motherboard that has better integrated sound for Z77 other than the G1.Sniper 3 and M3. Some other boards might say XF-I but they use ALC898 from Realtek, and ASRock also has some Core3D integrations, but they don’t have the amps, caps, nor Emi shield the G1 series has.
USB 3.0 Tests:
USB performance is really good on this board compared to the other boards I have tested. I have no idea why, perhaps it is because of the BIOS. I use the same setup on all boards to test, a USB 3.0 to SATA 3GB/s using a SATA6GB/s drive without any data on a partition.
Included Software Package:
Here is the Sound Blaster software package, it is very well laid out, and it isn’t confusing. Very easy to use and pretty cool.
This is the Big Foot Networks software:
It is the same software as on the other G1 series boards, the e2201 software is the same as that of the e2200.
This is GTL an excellent piece of software; very good for overclocking in-Windows, however it is for benching as the settings you try won’t stick after reboot. For LN2 OC this software is a must!
MVP is included and pretty cool.
3D power is good for testing the Digital PWM settings through Windows; I use it for LLC testing.
While testing the G1.Sniper 3 I kept thinking, this really is a great Z77 board. I tried to find some things that aren’t so great about it and some that are extremely impressive, I found both and had a great time doing it.
The one biggest downside is that while the PLX8747 makes multi GPU configurations prosper, it also makes single GPU configurations hurt compared to native Z77 setups as the delay is still present as it was on the NF200. So bottom line, if you care about 3% then this board is great for multi-GPU setups. I believe that the PLX is an improvement as it is a new chip, it isn’t 7 years old like the NF200, and it supports PCI-E 3.0. The reason chips like that have some delay is because of the output bandwidth being double that of the input. The easy way to think about it is to think about the PLX and NF200 being water jugs 32oz in size. Let’s say a normal PCI-E controller outputs 16oz and a single slot needs 16oz and the time needed is 1 second to fill. To fill two slots at 8oz each, one would pour half into each slot within the same time. If you have 16oz slots and a bigger jug (32oz versus 16oz), you can do the same, but the issue is that the jug has to be filled twice thus adding delay. That is why there is a delay because while it allows more bandwidth to the GPUs the input is still only half that of the output. There are also peer to peer connections between GPUs which the extra 16x from the PLX helps with as well. I believe that the benefits will shine through when using PCI-E 3.0 graphics cards like the GTX680, as it is certified for PCI-E 3.0 and works with it.
The build quality and features of the G1.Sniper 3 are the best of their class. The Intel NIC with the BigFoot NIC, the Creative Core3D, Digital PWM and VRM components, and overclocking features are of the highest quality and performance you can find on a Z77 motherboard. The layout and design of the motherboard is extremely optimal, not requiring ATX-XL but providing 4-way SLI and CrossFireX capabilities. The USB 3.0 performance is very good, audio performance is excellent at default but could be better, the 2D efficiency and multi GPU efficiency is very good as well. I believe that for Z77 for almost all 2-way, 3-way, 4-way GPU benchmarks, a PLX 8747 equipped motherboard might be the only way to go and the price of the G1.Sniper 3 makes it an easy choice in all respects. Its overclocking features are very useful and provide every overclocker everything they need to take their CPU to its maximum potential. The new 3D UEFI is improved over the X79 boards, and the advanced mode has so many good useful features and is very easy to navigate. I have used UEFI from the top 5 manufacturers and I have to say that the GIGABYTE rivals them, and in my opinion is one of the easiest to navigate. The 3D mode is unrivaled so far in offering very cool features, it is unlike a BIOS. When I use the 3D mode I feel like I am in an OS because of the ability to drag and configure windows. The profile sharing is going to bring overclocking to new levels, and in my opinion can redefine a novice overclocker’s experience of the sport.
Sin’s Take/Bottom line: This board is built for multi-GPU gaming and benching, and it does a great job with both. Its features are abundant and the accessories package is big enough to suite everyone’s needs. In the US the price is really unbeatable for the features and the quality is great over the top. I definitely recommend this board for anyone looking for take full advantage of their SLI/CF setups.
Thanks to GIGABYTE for making this review possible.
Last edited by sin0822; 05-30-2012 at 03:50.
Too bad it doesn't have a gun-shaped heatsink anymore.
Where courage, motivation and ignorance meet, a persistent idiot awakens.
Things people say: Massman is a god (05-02-2102), You're a genius (25-10-2011), Massman, overlord of overclocking (07-30-2010), brain dead corporate dude that deserves to be stillborn (09-18-2010)
Things I write: Windows 8 Results (Temporarily) Not Accepted At HWBOT - Benchmark Result Veracity Compromised, Raspberry Pi Overclocking Guide – Cheap, Fun, and Overclockable (with HWBOT Prime), HWBOT Integrates Intel Extreme Tuning Utility ("XTU") - Opportunity to Grow The OC Community!, Intel Haswell Overclocking Fully Disclosed - Theory For Core i7 4770K!, HWBOT Pro OC Cup, a new challenge for extreme overclockers., Critical thinking - Should Maximus V Extreme be banned from competitive overclocking?, HWBOT Vision 2012, The Industry’s Social Responsibilities – Support the Community., In Response to Chew's Goodbye Note (the unhealthy relation between company and community), The X58A-OC, the enthusiast community and a long-term vision, The Efficiency Rating, Hardware sharing, nostra culpa?, The paradox of a fair overclocking competition, HWBOT Memory Index v0.1, AMD Thuban Voltage and Temperature Scaling, Is LN2 on GPU really worth the trouble?
I like the mechanical part of the review but its missing the "fun" factor of overclocking. It's like you're subzero and just being serious lol.
here's an example of a fun review http://www.xtremesystems.org/forums/...ion&highlight=
another review that has some color: http://www.jonnyguru.com/modules.php...Story&reid=287
Last edited by Hondacity; 05-30-2012 at 07:40.
Oh God, Honda you want it to be really cheesy
LMAO, how cheesy is this one
yea i really like Jonny's reviews on PSUs, i like his writing style a lot.
Sometimes i am unsure whether to make it funny or make it serious. Maybe in the conclusion and intro next time I will make it funny for you. Just for you honda. lol As far as the subzero goes, i thought all you guys wanted was benchmark scores, now you want humor to go along with them too lol, sure honda for you next time. .
Im just suggesting. I remember writing 50+ reports in electronics lab[from electrons to components and applications]. My reports were mechanical. In this world called the internet. Catering to everybody and not just engineers is always good. Including abit of writing zest adds that extra "must read" factor.