NVIDIA GeForce GTX 760 vs NVIDIA GeForce GTX 960 Comparison

The NVIDIA GeForce GTX 760 and NVIDIA GeForce GTX 960 represent two distinct generations of mid-range graphics cards, with the former based on the Kepler architecture and the latter on Maxwell 2.0. While both cards target similar performance segments and share the same 2 GB GDDR5 memory capacity, benchmark data reveals significant shifts in compute API efficiency and raw throughput. The GTX 960 emerges as the overall winner in head-to-head testing, taking two of three benchmarks, yet the GTX 760 demonstrates a surprising advantage in one specific API workload that highlights architectural trade-offs.

Head-to-Head Benchmarks

The most decisive victory for the GTX 960 comes in the Geekbench Metal test, where it scores 8,773 points against the GTX 760's 4,524 points. This represents a 48.4% delta in favor of the newer card, a massive gap that underscores the Maxwell 2.0 architecture's superior execution of Metal-optimized workloads. The GTX 960's advantage is nearly twofold, suggesting that its architectural improvements deliver substantial gains in low-level graphics API performance.

In Geekbench OpenCL, the GTX 960 again takes the lead, scoring 18,925 points versus 11,299 points for the GTX 760. The 40.3% delta in this test reinforces the pattern of the GTX 960 being significantly faster in compute-oriented benchmarks. OpenCL performance is critical for general-purpose GPU computing, and the data shows the GTX 960 holds a commanding edge, likely due to its higher boost clock and more efficient instruction scheduling.

However, the GTX 760 claws back one victory in the Geekbench Vulkan test, where it scores 12,551 points against the GTX 960's 9,231 points. This 36% delta swings in the opposite direction, a surprising result given the GTX 960's wins elsewhere. The Vulkan result suggests that Kepler's execution units handle Vulkan's driver model more effectively, or that the GTX 760's wider memory bus (256-bit versus 128-bit) provides a bandwidth advantage that Vulkan workloads exploit more heavily.

Looking at average benchmark scores, the two cards are remarkably close overall. The GTX 760 posts an average score of 9,458 across all its benchmarks, while the GTX 960 averages 9,273. This 185-point difference is less than 2%, meaning the overall performance picture is nearly identical despite the divergent API-specific results. The GTX 760's average places it just 0.2% behind the NVIDIA GeForce GTX TITAN BLACK, a much higher-tier card, while the GTX 960 sits 0.2% behind the NVIDIA GeForce GTX 465. Both cards fall in the 45th-46th percentile of all GPUs, indicating they occupy similar mid-range territory.

The nearest rival data shows the GTX 760 is 1.6% ahead of the AMD Radeon R7 M380 and 1.7% ahead of the NVIDIA GeForce GTX 850M, while the GTX 960 is 0.6% ahead of the AMD Radeon Vega 8. These small deltas confirm that neither card dramatically outperforms its direct competitors, making the GTX 960's wins in Metal and OpenCL all the more notable.

Architecture Differences

The GTX 760 uses the GK104 chip built on the Kepler architecture, fabricated on TSMC's 28 nm process node. It packs 3,540 million transistors into a 294 mm² die, yielding a transistor density of 12.0 million transistors per mm². In contrast, the GTX 960 employs the GM206 chip with the Maxwell 2.0 architecture, also on 28 nm, but with fewer transistors at 2,940 million and a smaller 228 mm² die. This gives the GTX 960 a higher transistor density of 12.9 million per mm², indicating a more compact and efficient design.

The clock speeds differ substantially, with the GTX 960 running a base clock of 1127 MHz and a boost clock of 1178 MHz, compared to the GTX 760's 980 MHz base and 1032 MHz boost. This 147 MHz base clock advantage and 146 MHz boost advantage for the GTX 960 contributes significantly to its higher performance in compute benchmarks. Memory clocks also favor the GTX 960, which operates at 1753 MHz (7 Gbps effective) versus the GTX 760's 1502 MHz (6 Gbps effective).

Shader configuration shows a reversal in raw counts, with the GTX 760 featuring 1,152 shading units and 96 texture mapping units (TMUs), while the GTX 960 has 1,024 shading units and 64 TMUs. Both cards have 32 ROPs. Despite having fewer shading units, the GTX 960 achieves higher pixel throughput at 37.70 GPixel/s versus 24.77 GPixel/s for the GTX 760, demonstrating that Maxwell's architecture extracts more work per shader. Texture rate tells a different story, with the GTX 760 reaching 99.07 GTexel/s against the GTX 960's 75.39 GTexel/s, reflecting the former's larger TMU count.

Memory configuration differs notably in bus width, with the GTX 760 using a 256-bit interface and the GTX 960 a 128-bit interface. This results in the GTX 760 having significantly higher memory bandwidth at 192.3 GB/s compared to the GTX 960's 112.2 GB/s. The GTX 960 compensates with higher effective memory speed, but the bandwidth deficit remains a key architectural difference. Floating-point performance is nearly identical, with the GTX 960 posting 2.413 TFLOPS against the GTX 760's 2.378 TFLOPS, a marginal 1.5% difference.

Power consumption and API support also separate the two cards. The GTX 960 has a TDP of 120 W and requires a single 6-pin power connector with a suggested 300 W PSU, whereas the GTX 760 draws 170 W and needs dual 6-pin connectors with a 450 W PSU. The GTX 960 supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 760 is limited to DirectX 12 (11_0) and Vulkan 1.2.175. Display outputs also differ, with the GTX 960 offering HDMI 2.0 and three DisplayPort 1.2 outputs, while the GTX 760 has HDMI 1.4a and a single DisplayPort 1.2.

FAQ

Q: Which card performs better in Metal-based applications?

A: The GTX 960 is decisively ahead, scoring 8,773 in Geekbench Metal versus 4,524 for the GTX 760, a 48.4% advantage. This makes the GTX 960 the clear choice for Metal-optimized workloads.

Q: How do the two cards compare in Vulkan performance?

A: The GTX 760 wins this benchmark, scoring 12,551 against the GTX 960's 9,231, a 36% delta. This is the only head-to-head test where the GTX 760 takes the lead.

Q: What is the overall average benchmark score difference?

A: The GTX 760 averages 9,458 points, while the GTX 960 averages 9,273. The GTX 760 is ahead by 185 points, or approximately 2%, despite losing two of three individual benchmarks.

Q: Which card has higher memory bandwidth?

A: The GTX 760 has 192.3 GB/s of bandwidth courtesy of its 256-bit bus, versus 112.2 GB/s for the GTX 960 on a 128-bit bus. This is a 71.5% advantage for the GTX 760.

Q: Does the GTX 960 consume less power?

A: Yes, the GTX 960 has a TDP of 120 W and requires only a single 6-pin connector with a 300 W PSU, whereas the GTX 760 draws 170 W and needs dual 6-pin connectors with a 450 W PSU.

Q: Which card supports newer API versions?

A: The GTX 960 supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 760 only reaches DirectX 12 (11_0) and Vulkan 1.2.175. The GTX 960 also offers HDMI 2.0, compared to the GTX 760's HDMI 1.4a.

Specification Differences

The two cards differ in nearly every core specification except for memory size (both 2 GB GDDR5), ROP count (both 32), process node (both 28 nm), foundry (both TSMC), bus interface (both PCIe 3.0 x16), slot width (both dual-slot), and physical length (both 241 mm). Key differences include:

• Chip and Architecture: GK104 on Kepler versus GM206 on Maxwell 2.0
• Transistors: 3,540 million versus 2,940 million
• Die Size: 294 mm² versus 228 mm²
• Transistor Density: 12.0M / mm² versus 12.9M / mm²
• Base Clock: 980 MHz versus 1127 MHz
• Boost Clock: 1032 MHz versus 1178 MHz
• Memory Clock: 1502 MHz (6 Gbps) versus 1753 MHz (7 Gbps)
• Memory Bus Width: 256-bit versus 128-bit
• Memory Bandwidth: 192.3 GB/s versus 112.2 GB/s
• Shading Units: 1,152 versus 1,024
• TMUs: 96 versus 64
• Pixel Rate: 24.77 GPixel/s versus 37.70 GPixel/s
• Texture Rate: 99.07 GTexel/s versus 75.39 GTexel/s
• FP32: 2.378 TFLOPS versus 2.413 TFLOPS
• TDP: 170 W versus 120 W
• Power Connectors: 2x 6-pin versus 1x 6-pin
• Suggested PSU: 450 W versus 300 W
• Display Outputs: 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 versus 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.2
• DirectX Support: 12 (11_0) versus 12 (12_1)
• Vulkan Support: 1.2.175 versus 1.4
• Release Date: 2013-06-24 versus 2015-01-21
• Generation: GeForce 700 versus GeForce 900
• Predecessor: GeForce 600 versus GeForce 700
• Successor: GeForce 900 versus GeForce 10
• Launch MSRP: 249 USD versus 199 USD

The Verdict

The data presents a nuanced picture. The GTX 960 wins two of three head-to-head benchmarks, including the heavily weighted Metal and OpenCL tests, with deltas of 48.4% and 40.3% respectively. Its higher clocks, newer architecture, and superior API support make it the stronger choice for modern compute workloads and applications that leverage DirectX 12 (12_1) or Vulkan 1.4 features. The GTX 960 also offers significantly lower power consumption, with a 120 W TDP versus 170 W, and simpler power requirements.

However, the GTX 760 is not without merit. Its 36% victory in Vulkan shows that Kepler still has strengths in certain low-level APIs, and its 256-bit memory bus provides 192.3 GB/s of bandwidth, which is nearly double the GTX 960's throughput. For workloads that are bandwidth-sensitive, such as high-resolution texture streaming or specific Vulkan titles, the GTX 760 may hold an edge. The GTX 760 also has a higher average benchmark score overall at 9,458 versus 9,273.

Users prioritizing raw compute performance, modern API support, and energy efficiency should select the GTX 960. Users with legacy applications that favor Vulkan 1.2 or that depend on high memory bandwidth may find the GTX 760 adequate, but its older DirectX 11_0 feature set and higher power draw make it a less future-proof option. The GTX 960's 199 USD launch MSRP also undercuts the GTX 760's 249 USD launch MSRP, reinforcing its position as the more balanced and cost-effective choice for most buyers, despite the GTX 760's narrow overall average score lead.