Unlike the companies that used GeForce GTX 1080 PCBs and coolers for their 1070 Ti cards, Gigabyte pulled parts from the other side of its portfolio. It instead revised the design of its GeForce GTX 1070 G1 Gaming to accommodate Nvidia's higher-end solution. Naturally, the board itself is more than ample. It's the heat sink and fan that we need to look at more closely.
Of course, we've seen the competition name their cards as if 1070 Tis were differentiated by their out-of-box performance, garnishing them with software overclocking options as a consolation prize. Gigabyte scraps the Aorus label altogether for this one, though. If it can't apply factory overclocking, it's not a flagship. Period. Seems like sound logic to us.
Since the performance of all third-party GeForce GTX 1070 Ti cards is similar, given fixed clock rates mandated by Nvidia, actual benchmark results depend on the GPU Boost frequencies each vendor can sustain. As you might imagine, then, cooling and chip quality play the most important roles. That's why we're putting our emphasis on the actual implementation of each model. To that end, a lot of equipment goes into thoroughly documenting a graphics card's behaviors.
At this point, it's not possible to reach a meaningful conclusion by comparing the performance of one card over another. The differences are too small. Therefore, we're not going to label manufacturers in our benchmarks, particularly since one of those two variables (chip quality) cannot be held constant, even between retail samples off the shelf. Just know that every GeForce GTX 1070 Ti fares similarly out of the box.
Prior to launch, Gigabyte had an overclocked BIOS ready to go. It was initially even approved by Nvidia. As a result, around 25,000 cards shipped out from the factory with matching labels (the following two labels show the difference between our first sample and the model that eventually hit retail). If you were lucky enough to receive one of the first cards, you're probably seeing much higher frequencies than everyone else with a 1070 Ti.
However, since Nvidia banned this first edition of the card and refused to allow the overclocked BIOS' use, we're honoring Gigabyte's request to only use the updated version of its card for our review.
Gigabyte's bundle is limited to a software DVD and manual.
The fan shroud is made of anthracite-colored matte plastic with eye-catching orange highlights. A weight of just 757g makes this one of the lighter cards we've reviewed, and that helps simplify installation in compact cases.
A total of three 80mm fans are meant to ensure sufficient airflow. It's hard to tell that those fans are on the small side just by looking at pictures of the 1070 Ti Gaming 8G, since the cooler is long and narrow.
Around back, the PCB is covered with a single-piece, blackened aluminum plate with a white Gigabyte logo in the middle.
Plan for an additional 5mm of clearance for the backplate. This extra requirement means the Gigabyte GeForce GTX 1070 Ti Gaming 8G isn't particularly well-suited for multi-GPU configurations. Stability suffers without the backplate installed, so you don't want to try removing it.
Up top, that Gigabyte logo sports LED back-lighting. There's also a fan-stop indicator and an eight-pin auxiliary power connector toward the end.
From the back, this card is completely closed off. That makes sense, as the cooling fins are oriented vertically, blocking air from moving side to side and instead guiding it up (out the top) and down (towards your motherboard). But the design certainly won't help temperatures inside your case, since exhaust is recirculated.
The I/O bracket features five familiar outputs, four of which can be used simultaneously in multi-monitor configurations. In addition to the single DVI connector, you get one HDMI 2.0 port (particularly useful for VR HMDs) and three DisplayPort 1.4-ready interfaces. Ventilation holes cut into the plate suggest help with exhaust, but they really serve no purpose since the fins direct air vertically.
Since the technical specifications are required to be the same for all GeForce GTX 1070 Ti, the following table applies to all cards collectively:
|GPU||GeForce GTX 1080 (GP104)||GeForce GTX 1070 Ti (GP104)||GeForce GTX 1070 (GP104)|
|Base Clock||1607 MHz||1607 MHz||1506 MHz|
|GPU Boost Clock||1733 MHz||1683 MHz||1683 MHz|
|GFLOPs (Base Clock)||8228||7816||5783|
|Texel Fill Rate||277.3 GT/s||244.3 GT/s||201.9 GT/s|
|Memory Data Rate||10 Gb/s||8 Gb/s||8 Gb/s|
|Memory Bandwidth||320 GB/s||256 GB/s||256 GB/s|
|Transistors||7.2 billion||7.2 billion||7.2 billion|
|Die Size||314 mm²||314 mm²||314 mm²|
Test System & Measurement Methods
We introduced our new test system and methodology in How We Test Graphics Cards. If you'd like more detail about our general approach, check that piece out. In this case, only the CPU, RAM, motherboard, and cooling system are different.
|Hardware||Intel Core i7-6900K @ 4.3 GHz|
MSI X99S XPower Gaming Titanium
Corsair Vengeance DDR3-3200
1x 1TB Toshiba OCZ RD400 (M.2, System)
2x 960GB Toshiba OCZ TR150 (Storage, Images)
be quiet! Dark Power Pro 11, 850W Power Supply
Windows 10 Pro (Creators Update)
|Cooling||Alphacool Eisblock XPX|
Alphacool Eiszeit 2000 Chiller
2x be quiet! Silent Wings 3 PWM
Thermal Grizzly Kryonaut
|Case||Lian Li PC-T70 with Extension Kit and Mods|
Configurations: Open Benchtable, Closed Case
|Power Consumption Measurement||Contact-free DC Measurement at PCIe Slot (Using a Riser Card)|
Contact-free DC Measurement at External Auxiliary Power Supply Cable
Direct Voltage Measurement at Power Supply
2x Rohde & Schwarz HMO 3054, 500MHz Digital Multi-Channel Oscilloscope with Storage Function
4x Rohde & Schwarz HZO50 Current Probe (1mA - 30A, 100kHz, DC)
4x Rohde & Schwarz HZ355 (10:1 Probes, 500MHz)
1x Rohde & Schwarz HMC 8012 Digital Multimeter with Storage Function
|Thermal Measurement||1x Optris PI640 80Hz Infrared Camera + PI Connect|
Real-Time Infrared Monitoring and Recording
|Acoustic Measurement||NTI Audio M2211 (with Calibration File, Low Cut at 50Hz)|
Steinberg UR12 (with Phantom Power for Microphones)
Creative X7, Smaart v.7
Custom-Made Proprietary Measurement Chamber, 3.5 x 1.8 x 2.2m (L x D x H)
Perpendicular to Center of Noise Source(s), Measurement Distance of 50cm
Noise Level in dB(A) (Slow), Real-time Frequency Analyzer (RTA)
Graphical Frequency Spectrum of Noise
MORE: Best Graphics Cards
MORE: All Graphics Content