NVIDIA GeForce GTX 660 Ti vs NVIDIA GeForce GTX 770 Comparison

NVIDIA
GEFORCE

NVIDIA GeForce GTX 660 Ti

CORE STATE GK104
VRAM 2 GB
CLOCK SPEED 980 MHz
TDP 150 W
BUS WIDTH 192 bit
ARCHITECTURE Kepler
nm
PROCESS 28 nm
LAUNCH DATE 2012
VS
NVIDIA
GEFORCE

GeForce GTX 770

CORE STATE GK104
VRAM 2 GB
CLOCK SPEED 1085 MHz
TDP 230 W
BUS WIDTH 256 bit
ARCHITECTURE Kepler
nm
PROCESS 28 nm
LAUNCH DATE 2013

PERFORMANCE BENCHMARKS

geekbench_opencl
15,113
15,707
geekbench_vulkan
15,012
19,166
geekbench_metal
N/A
9,369

Analysis: NVIDIA GeForce GTX 660 Ti vs NVIDIA GeForce GTX 770

The NVIDIA GeForce GTX 660 Ti and the NVIDIA GeForce GTX 770 are both built on the same underlying GK104 chip and Kepler architecture, yet they occupy different positions in the GeForce 600 and GeForce 700 generations. The GTX 770 is the newer, more fully enabled part, and the benchmark data reflects a clear performance hierarchy, but the GTX 660 Ti retains strengths in specific workloads that make it a distinct option rather than a simple downgrade. The data shows a 2–0 win record for the GTX 770 in the head-to-head benchmarks, yet the average benchmark scores tell a more nuanced story, with the GTX 660 Ti actually posting a higher aggregate score.

Where Each One Wins

The GTX 770 wins decisively in the two benchmarks where both cards were tested directly. In the Geekbench OpenCL test, the GTX 770 scores 15707 against the GTX 660 Ti’s 15113, a 3.8% advantage. The gap widens dramatically in the Geekbench Vulkan test, where the GTX 770 reaches 19166 compared to the GTX 660 Ti’s 15012, a 21.7% lead. This makes the GTX 770 the clear choice for compute-heavy workloads that leverage Vulkan’s low-level API, as the newer card extracts substantially more performance from the same architectural foundation.

The GTX 660 Ti, however, wins in the aggregate benchmark average. Its average benchmark score is 15063, while the GTX 770 averages 14747. This is a 2.1% difference in favor of the older card, according to the nearestRivals data for the GTX 770. This apparent contradiction stems from the fact that the GTX 770’s average is pulled down by its Geekbench Metal score of 9369, a test that was not run on the GTX 660 Ti. When only the shared OpenCL and Vulkan tests are considered, the GTX 770 wins both; but when the full benchmark suite is averaged, the GTX 660 Ti’s consistent performance across its two recorded tests gives it a higher mean.

In terms of raw specification advantages, the GTX 770 leads in every computational resource. It has 1536 shading units versus 1344, 128 texture mapping units versus 112, and 32 ROPs versus 24. Its base clock is 1046 MHz versus 915 MHz, and its boost clock is 1085 MHz versus 980 MHz. These translate directly to higher pixel and texture rates: 34.72 GPixel/s versus 27.44 GPixel/s, and 138.9 GTexel/s versus 109.8 GTexel/s. The FP32 throughput also favors the GTX 770 at 3.333 TFLOPS versus 2.634 TFLOPS. The GTX 660 Ti’s only wins are in the average benchmark score and, by extension, its percentile ranking relative to all GPUs, where both cards share the same 57th percentile.

The Verdict

From the data, the choice depends on which benchmark suite one prioritizes. If the workload is centered on Vulkan compute, the GTX 770 is the unequivocal pick, as its 21.7% advantage in that test is the single largest performance gap between the two cards. The GTX 770 also wins OpenCL, but by a smaller margin of 3.8%, indicating that the newer card is not universally faster across all APIs.

For users who rely on a broader set of benchmarks, the GTX 660 Ti’s higher average score of 15063 versus 14747 suggests that it delivers more consistent performance in the tests where both are measured. The GTX 660 Ti’s lower power draw of 150 W versus 230 W, and its less demanding suggested PSU of 450 W versus 550 W, make it a more practical choice for systems with modest power supplies. The GTX 770 requires a 1x 6-pin + 1x 8-pin power connector setup, while the GTX 660 Ti uses 2x 6-pin, further reinforcing the GTX 660 Ti’s lower system integration requirements.

The GTX 770 is longer at 267 mm versus 241 mm, and it is the only one of the two with listed height and width dimensions: 111 mm and 38 mm, respectively. Both are dual-slot cards with identical display outputs (2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2) and identical API support (DirectX 12 (11_0), OpenGL 4.6, Vulkan 1.2.175). The GTX 770’s launch MSRP is 399 USD, while the GTX 660 Ti’s is 299 USD. The GTX 770 also has a larger memory bus at 256 bit versus 192 bit, and higher memory bandwidth at 224.4 GB/s versus 144.2 GB/s, despite both having 2 GB of GDDR5 memory.

Head-to-Head Benchmarks

The first head-to-head test is Geekbench OpenCL, where the GTX 770 scores 15707 against the GTX 660 Ti’s 15113. The deltaPct is -3.8%, meaning the GTX 660 Ti trails by 3.8%. This is a modest lead for the GTX 770, reflecting its higher core count and clock speeds. The GTX 770 has 14.3% more shading units (1536 versus 1344), 14.3% more TMUs (128 versus 112), and 33.3% more ROPs (32 versus 24). Its base clock is 14.3% higher (1046 MHz versus 915 MHz), and its boost clock is 10.7% higher (1085 MHz versus 980 MHz). Yet the OpenCL score advantage is only 3.8%, suggesting that OpenCL performance is not purely scaling with raw resource counts, possibly due to memory bandwidth constraints or driver optimizations.

The second head-to-head test is Geekbench Vulkan, where the GTX 770 scores 19166 against the GTX 660 Ti’s 15012. The deltaPct is -21.7%, indicating the GTX 660 Ti trails by a substantial margin. This is the GTX 770’s biggest win and represents a 27.7% raw score improvement over its own OpenCL result, while the GTX 660 Ti’s Vulkan score is nearly identical to its OpenCL score (15012 versus 15113). The GTX 770’s Vulkan advantage is likely tied to its higher memory bandwidth of 224.4 GB/s versus 144.2 GB/s, which is a 55.6% difference. The 256-bit bus versus 192-bit bus, combined with the faster 7 Gbps effective memory speed versus 6 Gbps, gives the GTX 770 a clear edge in bandwidth-sensitive workloads. The GTX 660 Ti’s memory clock is 1502 MHz, while the GTX 770’s is 1753 MHz, further widening the gap.

In the aggregate, the GTX 770’s two head-to-head wins give it a 2–0 record. However, the GTX 660 Ti’s average benchmark score of 15063 is higher than the GTX 770’s 14747. This is because the GTX 770’s average includes its Geekbench Metal score of 9369, which is significantly lower than its OpenCL and Vulkan scores. The GTX 660 Ti has no Metal score in its benchmark list, so its average is based solely on its OpenCL and Vulkan results. This discrepancy means that the GTX 770’s overall average is dragged down by a single weak test, even though it wins every test where both cards are present.

FAQ

Q: Which card has the higher average benchmark score?

A: The GTX 660 Ti has a higher average benchmark score of 15063, compared to the GTX 770’s 14747. The GTX 770’s average is lowered by its Geekbench Metal score of 9369, which is not present in the GTX 660 Ti’s benchmark list.

Q: How much faster is the GTX 770 in Vulkan?

A: The GTX 770 scores 19166 in Geekbench Vulkan, while the GTX 660 Ti scores 15012. This gives the GTX 770 a 21.7% lead, the largest margin in any head-to-head test.

Q: What is the memory bandwidth difference?

A: The GTX 770 has a memory bandwidth of 224.4 GB/s, while the GTX 660 Ti has 144.2 GB/s. This is a 55.6% difference, driven by the GTX 770’s 256-bit bus versus 192-bit and its 7 Gbps effective memory speed versus 6 Gbps.

Q: Do both cards have the same power connector requirements?

A: No. The GTX 660 Ti uses 2x 6-pin power connectors and has a TDP of 150 W with a suggested PSU of 450 W. The GTX 770 uses 1x 6-pin + 1x 8-pin and has a TDP of 230 W with a suggested PSU of 550 W.

Q: Are the physical dimensions identical?

A: No. The GTX 660 Ti is 241 mm long, while the GTX 770 is 267 mm long. The GTX 770 also has listed height (111 mm) and width (38 mm) dimensions, which the GTX 660 Ti does not have.

Q: Which card has more shading units?

A: The GTX 770 has 1536 shading units, while the GTX 660 Ti has 1344. The GTX 770 also has more TMUs (128 versus 112) and more ROPs (32 versus 24).

Architecture Differences

Both cards are built on the GK104 chip using the Kepler architecture, manufactured by TSMC on a 28 nm process node. The die size is identical at 294 mm², and the transistor count is the same at 3,540 million, resulting in the same transistor density of 12.0M / mm². The GTX 770 is from the GeForce 700 generation, while the GTX 660 Ti is from the GeForce 600 generation, but the underlying silicon is the same. The GTX 770 has a higher base clock of 1046 MHz versus 915 MHz, and a higher boost clock of 1085 MHz versus 980 MHz. The GTX 770 also has a higher memory clock of 1753 MHz versus 1502 MHz, translating to 7 Gbps effective versus 6 Gbps.

The core configuration differs significantly: the GTX 770 has 1536 shading units, 128 TMUs, and 32 ROPs, while the GTX 660 Ti has 1344 shading units, 112 TMUs, and 24 ROPs. This means the GTX 770 has 14.3% more shading units and TMUs, and 33.3% more ROPs. Both cards share the same API support: DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. Neither card has ray tracing cores or tensor cores, as these features are absent from the Kepler architecture. The GTX 770 has a larger memory bus of 256 bit versus 192 bit, and a higher memory bandwidth of 224.4 GB/s versus 144.2 GB/s. Both cards have 2 GB of GDDR5 memory.

Specification Differences

The GTX 770 and GTX 660 Ti differ in several key specifications. The GTX 770 has a base clock of 1046 MHz and a boost clock of 1085 MHz, while the GTX 660 Ti has 915 MHz and 980 MHz, respectively. The memory clock is 1753 MHz for the GTX 770 versus 1502 MHz for the GTX 660 Ti. The GTX 770’s memory bus is 256 bit, compared to the GTX 660 Ti’s 192 bit, and its bandwidth is 224.4 GB/s versus 144.2 GB/s.

The compute resources are higher on the GTX 770: 1536 shading units, 128 TMUs, and 32 ROPs, versus 1344, 112, and 24 on the GTX 660 Ti. The pixel rate is 34.72 GPixel/s for the GTX 770 versus 27.44 GPixel/s, and the texture rate is 138.9 GTexel/s versus 109.8 GTexel/s. The FP32 performance is 3.333 TFLOPS versus 2.634 TFLOPS. The TDP is 230 W for the GTX 770 versus 150 W for the GTX 660 Ti. Power connectors differ: the GTX 770 uses 1x 6-pin + 1x 8-pin, while the GTX 660 Ti uses 2x 6-pin. The suggested PSU is 550 W for the GTX 770 and 450 W for the GTX 660 Ti. The GTX 770 is longer at 267 mm versus 241 mm, and it has listed height (111 mm) and width (38 mm) dimensions, which the GTX 660 Ti lacks. The launch MSRP is 399 USD for the GTX 770 and 299 USD for the GTX 660 Ti.

DETAILED SPECIFICATIONS

SPECIFICATION
GTX 660 Ti
GTX 770
Core Specs
Shading Units
1,344
1,536 +14.3%
Shaders
1,344
1,536 +14.3%
TMUs
112
128 +14.3%
ROPs
24
32 +33.3%
Clocks
Base Clock
915 MHz
1046 MHz
Boost Clock
980 MHz
1085 MHz
Memory Clock
1502 MHz 6 Gbps effective
1753 MHz 7 Gbps effective
Memory
Memory Size
2 GB
2 GB
VRAM (MB)
2,048
2,048 0.0%
Memory Type
GDDR5
GDDR5
Memory Bus
192 bit
256 bit
Bandwidth
144.2 GB/s
224.4 GB/s
Cache
L1 Cache
16 KB (per SMX)
16 KB (per SMX)
L2 Cache
384 KB
512 KB
Performance
Pixel Rate
27.44 GPixel/s
34.72 GPixel/s
Texture Rate
109.8 GTexel/s
138.9 GTexel/s
FP32 (TFLOPS)
2.634 TFLOPS
3.333 TFLOPS
FP64 (TFLOPS)
109.8 GFLOPS (1:24)
138.9 GFLOPS (1:24)
Power
TDP
150 W
230 W
TDP (W)
150
230 +53.3%
Suggested PSU
450 W
550 W
Power Connectors
2x 6-pin
1x 6-pin + 1x 8-pin
Architecture
Architecture
Kepler
Kepler
GPU Name
GK104
GK104
Generation
GeForce 600
GeForce 700
Process Size
28 nm
28 nm
Transistors
3,540 million
3,540 million
Die Size
294 mm²
294 mm²
Foundry
TSMC
TSMC
Density
12.0M / mm²
12.0M / mm²
API Support
DirectX
12 (11_0)
12 (11_0)
OpenGL
4.6
4.6
Vulkan
1.2.175
1.2.175
OpenCL
3.0
3.0
CUDA
3.0
3.0
Shader Model
6.5 (5.1)
6.5 (5.1)
Physical
Slot Width
Dual-slot
Dual-slot
Length
241 mm 9.5 inches
267 mm 10.5 inches
Height
—
111 mm 4.4 inches
Outputs
2x DVI1x HDMI 1.4a1x DisplayPort 1.2
2x DVI1x HDMI 1.4a1x DisplayPort 1.2
Bus Interface
PCIe 3.0 x16
PCIe 3.0 x16
Other
Launch Price
299 USD
399 USD
Production
End-of-life
End-of-life
Predecessor
GeForce 500
GeForce 600
Successor
GeForce 700
GeForce 900
View GeForce GTX 660 Ti Details View GeForce GTX 770 Details