No announcement yet.

G1 Sniper 2 Review

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • G1 Sniper 2 Review

    G1 Sniper 2 Review


    MSRP: $359.99 USD


    Avalibility: Early September (Limited Edition in USA/CAN)
    I also have this review up with sections here.


    If you want you can watch this video too:

    Today I am very excited to bring you my latest review of the G1 Sniper 2, GIGABYTE’s new Z68 based G1 Killer series gaming board. The story begins back about a year ago around this time, GIGABYTE started brainstorming on ways to improve their products and hit new market segments that previously hadn’t been attended to. They asked gamers what they would want in a board, and even brought in one of the best overclockers (Hicookie) to help them develop new boards. They concluded that gamers and overclockers want very different things in a motherboard but end up buying the same boards, as they are the best built and have the most features. GIGABYTE set off to experiment with prototypes of gaming boards and overclocking boards to see if they could better target each niche market. First they released the G1 killer X58 series (which was late in the game for X58) and then the X58A-OC (which was way too late in the game). Now we all know the story about those boards, both series G1 and OC were plagued with really late release as well as the fact that Sandy Bridge made X58 look pretty puny. Now at that time it seemed like the general consensus was that users wanted these boards but with the LGA1155 chipset, and today that is exactly what GIGABYTE has brought to the table.





    The G1 Sniper 2 bring us technologies that allow it to survive and keep pace with future generation processors, the case and point is Ivy Bridge, which is set to release next year, but use the same LGA1155 socket and chipsets. The major difference is that Ivy Bridge has a PCI-E 3.0 controller and it’s a 22nm chip. Well motherboard manufacturers have used the magic balls and looked into the future, and are now bringing out boards with PCI-E 3.0 capabilities. The G1 Sniper 2 is a perfect example of this, as it has native PCI-E 3.0 as well as full 22nm BIOS support, so this board is already 100% ready for Ivy Bridge.


    There are other features in my opinion that also make this board a hardcore gaming motherboard. The best feature of this board in my opinion is the Creative CA20K2 Audio processor and 5 built in audio amplifiers, with a whole array of high quality audio capacitors and even an on-board EMI shield. Next there is the Bigfoot NPU, one of the most high-tech consumer NICs there is. The next feature is one I think all of GIGABYTE’s motherboards should entail and that is full control of every fan connector on the motherboard individually, which this motherboard has, along with 5 matching temperature sensors in same individual regions. On top of all that a gaming board of this caliber should have excellent heatsink and PCI-E spacing; meaning the CPU’s heatsink should clear the RAM DIMMs, and the GPU should have enough room to breathe. This board has all of those things and a few extras; a front panel with an auto OC button and USB 3.0, some unique looking heatsinks, and all around nice aesthetics.


    So join me as we venture into the new G1 Sniper 2, GIGABYTE’s new G1 Killer. I will also compare against the old G1 Assassin and the Z68X-UD7-B3.


    Introduction


    Box + Accessories


    Layout and Design


    Audio and NIC + Testing


    Deep look at ICs and Features


    BIOS


    Voltage Regulator Module Testing (LLC, Ripple, PWM Freq., and Power Testing)


    Overclocking and Debugging Report and Test Setup


    Benchmarks


    Heatsink and LEDs


    Conclusion




    Box + Accessories


    Front and back of the box look like an ammo box, also has a handle and nice explanation of features. It’s actually a very big box, here are some comparison to old G1 Assassin which was ATX-XL.

    You see the Assassin’s box is a bit bigger.

    Up close of the chipset and future Ivy Bridge support.

    After you open the box you have a plethora of accessories and manuals, and even stickers.

    A nice front panel, it has USB 3.0 and OC button, with an extra USB 2.0/eSATA connector as well.
    Sin’s Take: A front panel like that actually will cost you $20-30, the UD7 also has one, but no OC button.

    Four SATA6GB/s cables, driver CD, and HDMI to DVI connector.
    Sin’s Take: Now you don’t actually need to use the HDMI output for D-MODE Virtu, its enabled by default after you install the Virtu software. I like how everything is nice and black to match everything else, no random blue cables
    There are the stickers, there is also a poster underneath.
    Sin’s Take: I like having the stickers, even if I don’t actually use them all, I used the clear GIGABTYE sticker, and the “G1-Killer” sticker on my thermos for LN2 OCing. The poster is double-sided, one side has bullet holes and the other is a typical G1 Killer 3 military guys-pointing-gun- at-you-type-thing.


    Layout and Design
    First I am going to show you the G1 Sniper 2’s actual size compared to a normal ATX board like the Z68X-UD7 and an ATX-XL board, because it’s a hybrid of both.

    Shorter than ATX-XL.

    But a bit wider than normal ATX. Now this still qualifies for normal ATX size, so it won’t be a problem fitting into your case.


    The board has a busy looking PCB, its actually all black, no brown copper trace seep through, and you can see the heatsinks try to resemble a AR15 rifle.


    Sin’s Take: It’s something you need to get used to, I initially didn’t know what to think of how they looked, but over the months I really go used to it, and actually like it. Apart from the gun clip as the PCH heatsink, it looks pretty normal. I like the LEDs in the main heatsink, even if it doesn’t actually cool anything directly. Also this board is very pleasing on these eyes in general, I think GIGABYTE did a great job with the aesthetics; of course everyone is entitled to their own opinion.

    This is a different picture, just on its side.


    I am going to show you how the socket spacing compares to the Z68X-UD7 which uses standard socket and DIMM spacing. The G1 Sniper 2 has its spacing augmented.

    See Augmented.


    Sin’s Take: They did a great job adding extra space to the DIMMs, I had no trouble with my DDR3 2133 RAM (PSC). I don’t have hypers so I don’t know how they will clock, but this is a gaming board, not a subzero OC board. The increased space between the CPU and DIMMs might slightly hurt memory overclocker or performance, but we will take a look at that to see if there is any truth to it.


    The backpanel now:




    OC button has a cap, and HDMI has a plug.

    The memory area has no buttons around it, but there is plenty space to even change your RAM while a long GPU is in place.

    Here we have 7(6 angled, 1 straight) SATA connectors ((3)Black= Intel SATA3G, (2)White=Intel SATA6G, and (2)Gray=Marvell SATA6G). We then have 2 fan connectors, OC button connector, front panel headers, clear CMOS header, TPM, COM, 3 USB 2.0 ports which are out of view, and one USB 3.0 port which is on the side of the board next to the SATA ports.


    Sin’s Take: This is very important!!!!!! You see where they put the USB 3.0 connector (bottom of the picture) and the OC button one(white small lower left)? This is a HUGE improvement over the previous G1 Assassin. Before, on the Assassin, the OC button connector was right under the 3rd GPU slot, so when you did SLI the GPU would literally sit on the connector. The USB 3.0 connector is in the best place possible in my opinion for the board with a provided front panel, as that is less length to the cable, thus easier cable management, and better signal quality.

    You have two PCI-E 16x slots; the second one is actually an 8X slot. You not only have triple slot spacing, but that first 1x PCI-E slot isn’t blocked by the heatsink. Two PCI slots were added extra with an extra chip, as it’s not native to this platform.


    Sin’s Take: BRAVO!!!! GIGABYTE for not blocking the first 1x port, this is a big improvement over the UD7’s blocked first PCI-E 1X slot. I actually just checked again to make sure that is correct, it is NOT blocked! It is looking like this board really has been analyzed over and over and improved, as this is one of GIGABYTE’s best LGA1155 boards to date, and that is just by judging its layout and design.

    Here is the audio, we actually have some sort of EMI shielding on this board opposed to the G1 Killer X58 series. Same number of capacitors from Nichicon and same exact parts. Performance is even better than expected, it got “excellent” on every RMAA test except frequency response.. I will show you guys everything there is to the audio on this board, from the power in to the audio out. Please just go below if you want to see it!


    I also stripped the board nude for you if you want to see what is under the heatsinks.



    Audio Design and Testing!

    The Creative XF-I suite is also included, and it allows total user control. GIGABYTe went with the TOSLINK outputs to save space on the back panel, which is a good thing. They also went out of their way to use very high quality DACs and ADCs, which are found in some of the best Creative XF-I standalone PCI-E cards. The audio playback is really great, but there is no way to show you other than to run some tests. I have an oscilloscope but I felt that maybe users want to compare against other tests on the internet, so I used RMAA 5.5, which is basically a standard among review sites.


    The capacitors are very high quality from Nichicon, here is a close up of the capacitors and the EMI shield:

    Audio Testing:


    For this we use RMAA 5.5 with a loop cable, to test recording and playback, I tested against the Z68X-UD7 with a Realtek ACL889, which is actually a good audio IC:



    BigFoot LAN Testing:

    The Bigfoot Killer e2100 sells for about $90 USD. It is a combination of the e2100 Network processing Unit (NPU) some 1GB of DDR2 memory, and a Marvell 88E1118R Physical Layer Device(PHY). Now you can’t control your network speed outside of your router, but you can control some characteristics from your router to your mouse, such as bypassing the windows network stack, and set application control priority, which is what the NPU does. The newest software actually allows the NPU to designate streams like videos and give them more priority so playback is smoother. It is a very nice design, but it requires very smooth power and many different voltages, which is why it has two separate phases and a bunch of LDOs.


    Here we will test with AIDA32 Network Performance Benchmark. I also had tested before but, because of a lot of controversy over testing it correctly, I opted just to do this test. Now it’s a master/Slave test, you use two computers on the same network, and set one to master and one to slave and transfer data, its actually simple to setup.



    Now it is important to note that we didn't run any ping tests, just a test that will show up intranet bandwidth, so while its a benchmark it doesn't give you the whole picture. The thing is i don't want to venture into that land because like I said there is a lot of networking stuff I am just not good at explaining.


    Deep look at ICs and Features



    This part is all about teaching readers how the motherboard works, and what ICs make it tick and tock. A motherboard is nothing but a compilation of ICs, copper, and some SMD components. We will take a look at those right now.


    We will start with the Voltage Regulator Module:

    Next we will look at what each component does:

    So I hope that helps understand what they do.

    It’s a 12 phase VRM, 8 for the CPU cores, 2 for the System Agent/Uncore, and 2 for the Internal GPU.


    Sin’s Take:It might not seems like a lot compared to the 24 phases of the UD7, but it is a gaming board, and Sandy Bridge doesn’t require much power. Also the PWM controls are also there with 10 Level LLC (which work excellent). We will test in the VRM section, and then later I will test the UD7’s VRM for comparison.


    Here are some very unique fan and temperature controls on this board






    Thanks to GIGABYTE for making this review possible.
    Last edited by sin0822; 08-18-2011, 04:44 PM.

  • #2
    Re: G1 Sniper 2 Review

    Now all the peripheral ICs:



    The PCI-E subsystem:



    Now I want to show you the new PCI-E 3.0 switches:



    I will finish this section off with a show of the DRAM VRM, almost same as the UD7:



    Now one last shot of the CPU Socket VRM area:



    BIOS
    Now some of my readers asked for BIOS shots. Let me do this for you since pictures of the actual BIOS are pretty hard to do right because you have to take a picture of the screen, so I will take screenshots of the TouchBIOS, which has all the same settings. Some interesting things are pointed out as well:
    vDRAM/vDIMM, CPU PLL, and QPI/VTT are now 5mv per step!!!!!!!!!!!!!!!!!!!!!!!!!!!!


    Also the PWM frequency and overcurrent protection make a difference, I will show you in the next section.









    Voltage Regulator Module Testing


    (Read points, LLC, Ripple, PWM Freq., and Power Testing)


    Here is how the Multi-Level LLC works, as well as how the PWM frequency setting works. LLC is actually really great. I also am showing you how the other voltages read compared to what is set in the BIOS:
    [URL=http://img263.imageshack.us/i/voltagescopy.png/][IMG]http://img263.imageshack.us/img263/9550/voltagescopy.png[/IM
    Sin’s Take: LLC level 7 is really great, it seems that level 6 and below all droop under load(which is good for 24/7), while level 7 gives you what you set at idle and raises a few millivolts under load. Then level 8-10 all increase under load, and actually level 10 gives you the least change in voltage, but it also offsets voltage from 10mv of what you set higher. Notice that software seems to find certain points like 1.488 and rounds the voltage to that, so it’s neither the most accurate nor the most precise, which is typical for software VCore readings.


    I recommend using Level6 for 24/7 overclocks, and Level 7 for high overclocks on air and water, and Level 10 if you want something more extreme. Level 6 and 7 in reality give you close to what you set, so you aren’t feeding your CPU more voltage than you think. I am glad that on this board LLC is working as it should; the 10 Levels are actually as good if not better than the old 3.


    Here is power consumption of the system, notice how much Virtu cuts down the power, but you have to remember that the Sniper 2 has extra Creative and Bigfoot subsystems that use a lot of power:

    Now we do the ripple testing, here is where I soldered (you can also read vcore from there)

    Ripple Testing:


    First we use the trigger and wait for the highest rising or falling edge at a certain point, to measure max Vpp (Vpp is ripple) peak and then take a screenshot of that. Then we use something called a pass fail test set at 40mv up and 40mv down for a total of 80mv Vpp that will pass. The test is run during the entire benchmark. If for instance we have over 80mv Vpp, or more than 40mv up or down then it will fail. So if it goes 60mv below the center line then it will fail, but it might not go above the line 60mv, so it’s not 80mv Vpp. In other cases it will go up about 30mv and then down about 39mv and pass but have a 69mv Vpp. Every test was done at 20mv/division (vertical has 8 divisions) and then 1ms/div (horizontal has 10 divisions). What does that mean? Basically from bottom to top its 160mv, and from left to right is 10ms, for those of you who don’t know how to read a scope shot. Please also realize that noise is also picked up inadvertently.


    To read more on how we test with the oscilloscope please go here: VRM, Voltage, and Oscilloscope Testing - Page 4


    I test at stock settings, and then at a good 4.5ghz overclock with at least 1.5v(exactly 1.5v in this case). So first we test wPrime 1024, SuperPI XS 1.5 32M, and then 3DMark Vantage CPU Tests 1 and 2 separately.


    Ripple Testing:

    Stock with optimized defaults loaded:
    wPrime:

    Vpp: ~30mv


    Fail %: 0%


    SuperPI:

    Vpp: ~36.9mv


    Fail %: 0%


    Vantage CPu test 1:

    Vpp: ~30.6mv


    Fail %: 0%


    Vantage CPU Test 2:

    Vpp: ~46.3mv


    Fail %: 0%


    Stress test:


    Now for some stress on the CPU, we clock to 4.5ghz with 1.5v. We also let the VRM heatsink get hot to the touch.


    wPrime:

    Vpp: ~63.8mv


    Fail %: 0%


    SuperPI

    Vpp: ~56.9mv


    Fail %: 2.3%


    Vantage CPU Test 1:

    Vpp: ~50mv


    Fail %: 0%


    Vantage CPU Test 2:



    Vpp: ~65mv


    Fail %: 1.9%


    Sin’s Take: Since I did the same tests on the 990FXA-UD7 and it also has an 8 phases just for CPU power, we can analyze what these results mean. First of all ripple was pretty good for only 12 phases. It seems that increase the switching frequency (in BIOS as PWM frequency) does help, as I expected ripple over 100mv, but it never happened. We saw much greater ripple and noise on the 990FXA-UD&, but it had no real impact on the OC. There is a large difference though, as the 990FXA-UD7 had a CPU with much larger current draw as well as two more cores, thus using a lot more power. So we except this to be better. Soon enough I will test the UD7 against this, so hopefully it will do better, but I don’t know. One thing is for sure though, the ripple control on this board is very good, we have good EMI shielding, and a lot of output capacitance. In retrospect the Z68X-UD7 has much less capacitance to the amount of phases, but larger current capacity. We also do not have to set PWM frequency or OCP for the Z68X-UD7 as its already at its highest for a board of that caliber.


    Voltage read points:


    (Be careful with the Vcore read point, you can also read on the back of the board as shown in the VRM testing above) Now in the past I recommended reading at the choke, but that is not a good idea, it’s very easy to do, but that voltage at the choke is much different than that being supplied. Plus on this board the chokes are through-hole, not SMD, so their outputs are on the back of the board. Please also realize that you need to have a good ground point, not all grounds will give you the same reading. Don’t use a MOLEX, for the CPU and QPI/VTT use the 8-pin CPU connector, and for the vDIMM and CPU PLL try and use the 24 pin connector’s ground.








    Overclocking and Test Setup
    First I do all testing with a 2600K, along with 1600mhz CAS 9 rated Kingston HyperX memory. I use these settings for 5 GHz:


    50x, BLCK 100.2(100.0 in windows), CPU PLL Voltage C3/C6 States disabled, EIST disabled, C1E disabled, Real-Time Ratio Change in OS enabled, PWM frequency +42%, overcurrent +60%. Multi-Level LLC set to Level 7 with 1.5v for Vcore. QPI/VTT at 1.15v with memory at 1600mhz timings 9,9,9,24 1T. I also set the CPU PLL at 1.66v, but you can set what you want, higher than 1.8v might even help. But don’t touch CPU PLL really.


    Here were results with the 2600K at max 5.2ghz:

    On hand we also have a superb 2500K overclocker, on this CPU I easily did 5.5ghz but only on one core, my air cooling couldn’t take much more:

    OC button = +7 multipliers!! (Much better than G1 Killer X58 (+4BLCK))


    Now I did a lot (a few weeks) of testing on this board, and found very little of the issues I had with early P67/Z68 boards. There are no bootloops, no cold boots, no BIOS flashbacks, nothing bad. The only thing I found, or I thought was wrong was that the Turbo Mode doesn’t work. I also noticed that Touch BIOS was smooth and works well.


    Turbo Mode is off on purpose because of OC button:


    I made an inquiry at GIGABYTE, and I was told that it doesn’t work because of the OC button’s potential, it gives and automatic 7x multiplier jump (for 2500K=4.1ghz for 2600K=4.2ghz) and it does it on the fly, so the C states are disabled, and ratio-change in OS is enabled. If you want Turbo to work properly, and it does by the way, enable C3/C6 states and disable Ratio Change in OS under advanced CPU core control(auto setting for CPU PLL overvoltage is enabled). A 7x multiplier jump on the fly is pretty cool if you ask me, you can hit it anytime you wish and it works, you can also disable it while doing anything.


    Anyways here are the test systems, I tossed some AMD systems in there that I recently tested, as well as the UD3H when applicable:










    All testing was done on BIOS F3.


    Sin’s Take: What is I see is the G1 Sniper 2 winning in many respects, but it should be winning in 3D benchmarks just because there is no added latency from the NF200. But it is only winning slightly. In the past I saw the G1 Assassin X58 beat out my X58A-UD5, because of the on-board NIC and Audio offloading CPU and memory resources, here we are seeing that same thing but to a lesser degree. I believe that it can do better, but because this BIOS is initial release it still needs to be tweaked for performance, as the BIOS can influence a lot of performance factors. I feel as if by launch when we have BIOS F4 or F5 the UD7 will get beat by even more in 3D benches. Right now we are putting a rookie up against a hall of famer, the UD7’s 2D performance is not only strong, but 3D performance isn’t bad either on the latest installment of the UD7. That might explain why the Sniper 2 doesn’t win in every 3D bench like it should be. BTW look at those wPrime scores, the Sniper 2 is winning, but then we move to SuperPI and its back to the UD7. Unigine has sound so the Sniper 2 wins hands down. 3DMark11 is a bit mixed, with the Sniper 2 barley winning, but that is where it should win hands down. It is interesting to see performance at 4.5ghz vs 5ghz on both boards, as sometimes at 5ghz one will win while at 4.5 the other will win. Very interesting results, either way the UD7 seems like the better benching board, and the Sniper 2 the better gaming. The gaming benchmarks are where we see a difference. CODMW2 shows great results ahead of the UD7 as does the BFBC2. BFBC2 was played online as well so the NPU should help us a bit there. You can also see that the UD7 wins in memory benchmarks almost always.


    Conclusion:

    With this board GIGABYTE has literally taken Z68 to a whole other level. Not only is the board aesthetically pleasing, it also has the stuff to back up its looks. It’s both beautiful on the inside and outside, something many boards aren’t. We take this board up against GIGABYTE’s other top of the line contender, the Z68X-UD7, which is really GIGABYTE’s crown jewel and at the end of the day we have a very hard decision to make as the boards are very close in price. Both boards are phenomenal, but the choice is pretty clear. If you are an overclocker or someone who needs 3-way SLI get the UD7, if you are a gamer get the G1 Sniper 2. Benchmark results show the boards are at each other’s throats. It is very nice to see the G1 Sniper 2 release at this point though, as it really looks like GIGABYTE is with this board; it has a very solid BIOS that could have only come from platform maturation and long hours of testing. Maybe that is why they released it so much later than other LGA1155 boards. In the other end of the field it looks like GIGABYTE really put those G1 killer X58 surveys to good use, as we see some very good improvements to the series, as minor as they are. I like that there is PCI-E 3.0, and I like the fact that GIGABYTE didn’t spend most of the past 3 months advertising it every day; they silently slipped it in and didn’t announce it until after reviewer’s had the board in hand. It was a pleasant surprise, but something that was handled very well. I wouldn’t buy a board because it has PCI-E 3.0, but its appealing to many. If I was planning on keeping this board for more than 2 years then maybe the PCI-E 3.0 would come in handy when we see GPUs that use it.


    What Sin Likes: I really like what is different (improved) from the X58 Assassin and the UD7. The spacing between the CPU and DIMMs is greater so that a larger cooler can fit, but this increase trace length and thus can impact performance, but we saw that GIGABYTE has done a good job of keeping the performance at a good level I really like the audio and the NIC, if anything I like control that the NPU provides. The audio sounds crisp and it shows in RMAA testing, the bass is pretty impressive as well. The amplifiers and the capacitors really do a good job, and provide what any off the shelf XF-I sound card does and possibility better. The best part about the NIC and Audio is that they work off of the PCH’s PCI-E lanes, meaning that the GPU’s PCI-E lanes are never touched, on X58 this was an issue, and one that still persists to Sandy Bridge, but this board has you covered. The fan control and temperature sensors are the things that really seal the deal for me though, as it’s a feature any 24/7 motherboard needs as it will be in a case and not on a bench-table. Now when I talk about attention to detail I talk about the little things, like moving the USB internal header to the edge of the board and making sure that the first PCI-E 1X slot isn’t blocked like on some other boards, and I can say that they did pay attention to the little things.

    What Sin Thinks Can Be Better: Honestly, if GIGABYTE put some buttons for start, reset, or even just clear CMOS I have a feeling many would buy this board without hesitation. Those things might not seem handy for gamers, but they honestly help anyone who is trying to overclock, and some gamers do overclock. Also for system building and diagnosing, it’s really a pain to use a screw driver to short two pins in the middle of 20 pins especially when you can’t see where you are shorting. Moving on I think they could have added a few more phases for the money that is being spent, but the phases on this board can handle a 5.2 GHz OC for years without a glitch, but of course people always want more. Efficiency would also be raised with more phases and heat would be lower. Now the Bigfoot application control caused issues with my ISP, so I have to set my STEAM games to priority 2 instead of 1 for them to work, same as the previous G1 Assassin I had, but that is Bigfoot’s issue, not GIGABYTE’s. The heatsink in the middle of the board that would usually sit on the Northbridge/IOH of the NF200 is really cooling nothing directly. It’s acting as a larger heatsink for half the VRM heatsink, which might not be a bad thing. It’s more than just kind of weird that it is there, but the LEDs in it make it looks like a well-rounded board. Anyways my last complaint is about the price, as I have heard rumors of it being $450 USD, which is just really a lot for a Sandy Bridge motherboard. On the other hand I just received confirmation that the board will be set MSRP of $359.99 USD for USA and Canada, but there is a catch, it is going to be a Limited Edition product in North America.

    Final word: As any top of the line GIGABYTE board, this motherboard is solid physically (design), psychologically (BIOS) mature, and emotionally (Performance) it’s doing its best at this point. That is why it is replacing the G1.Assassin X58 as my 24/7 motherboard. Before the Assassin I wasn’t really a gamer at all, but now after the Assassin I have played over 200 hours of Battlefield Bad Company 2(sin0822 if you want to look me up), the game I bought to help review the Bigfoot NIC back in March 2011. The limited edition part is what I find odd, I would think a board like this will sell very well, but we have to remember over $300 is a lot for a Sandy Bridge motherboard. Its price at $360 is going to make people debate over the Z68X-UD7 or the G1 Sniper, but if I was a gamer I wouldn’t have a hard time choosing. If I was a bencher then I wouldn’t either because the lack of simple on/off buttons is just not something I can deal with. On the other hand I would have a very hard time if I overclocked and gamed, as both boards fit the gig, but in the end it would come down to will I trade buttons, NF200, and 16 extra phases for Creative Audio and Bigfoot? That is a question you need to answer for yourself. With the G1 Sniper 2 GIGABYTE has come close to perfecting the art of Z68 board making, at $360 it might be one of the most expensive LGA1155 motherboards, but its features do fit its price for the most part.


    All in All I really enjoyed this board, found no major problems, everything functions as it should, and the extras are all there: 9.8/10 (Excellent)


    I would recommend this board to anyone looking for a hardcore gaming motherboard that will overclock great and last them for ages, they just have to come up with the funds to buy it. :)

    Comment


    • #3
      Re: G1 Sniper 2 Review


      Comment

      Working...
      X