NVIDIA GeForce GTX 680 vs NVIDIA GeForce RTX 3050 OEM Comparison

# The Verdict

The NVIDIA GeForce RTX 3050 OEM is categorically the faster card in this comparison, and the benchmark data leaves no room for ambiguity. In the two head-to-head tests available, the RTX 3050 OEM wins both, with a 259.3% lead in Geekbench OpenCL and a 223.6% lead in Geekbench Vulkan. The GTX 680 is a legacy product from the Kepler era, and its 55th percentile ranking versus all GPUs trails the RTX 3050 OEM's 57th percentile. However, the RTX 3050 OEM's average benchmark score of 15199 places it nearly neck-and-neck with the AMD Radeon RX 7600 (15171, a 0.2% delta), while the GTX 680's 14150 average puts it just below the NVIDIA Tesla K10 (14029, 0.9% ahead). For any user choosing between these two, the RTX 3050 OEM is the only rational pick for modern workloads, though the GTX 680 retains relevance only in legacy DirectX 9 or compute scenarios where its specific feature set is acceptable.

# Architecture Differences

The architectural gap here is generational and profound. The RTX 3050 OEM uses the GA106 chip built on Samsung's 8 nm process, packing 12,000 million transistors into a 276 mm² die with a transistor density of 43.5M per mm². The GTX 680 uses the GK104 chip on TSMC's 28 nm node, with 3,540 million transistors on a 294 mm² die, yielding a density of just 12.0M per mm². That means the RTX 3050 OEM crams over 3.6 times more transistors into a slightly smaller physical area. The foundries differ too—Samsung for the 3050 OEM, TSMC for the 680—and the process node difference (8 nm vs 28 nm) explains much of the performance delta.

Core configuration diverges sharply. The RTX 3050 OEM features 2304 shading units, 72 texture mapping units, and 32 ROPs, alongside 18 ray tracing cores and 72 tensor cores. The GTX 680 has 1536 shading units, 128 TMUs, and 32 ROPs, with no ray tracing or tensor cores at all. The TMU count actually favors the older card, but that advantage is swamped by the newer architecture's raw shader throughput. The RTX 3050 OEM also carries hardware support for DirectX 12 Ultimate (12_2), while the GTX 680 is limited to DirectX 12 (11_0). Both support OpenGL 4.6, but Vulkan support differs: the 3050 OEM supports Vulkan 1.4 versus the 680's 1.2.175.

Memory architecture is another chasm. The RTX 3050 OEM uses 8 GB of GDDR6 on a 128-bit bus with 224.0 GB/s bandwidth. The GTX 680 uses 2 GB of GDDR5 on a 256-bit bus with 192.3 GB/s bandwidth. The newer card has four times the capacity and 16.5% more bandwidth despite the narrower bus. Clock speeds also differ significantly—the 3050 OEM runs at 1515 MHz base and 1755 MHz boost, versus the 680's 1006 MHz base and 1058 MHz boost. Effective memory speed is 14 Gbps on the 3050 OEM versus 6 Gbps on the 680.

# Head-to-Head Benchmarks

The head-to-head results are lopsided in favor of the RTX 3050 OEM. In Geekbench OpenCL, the 3050 OEM scores 60740 against the GTX 680's 16906, a delta of 259.3%. This is not a marginal win; it is a 3.6-fold advantage. The Vulkan test shows a similar story: 57103 for the 3050 OEM versus 17646 for the 680, a 223.6% lead. These are the only two shared benchmark scores in the data, but they cover both a compute-centric API (OpenCL) and a modern graphics API (Vulkan), giving a consistent picture.

The RTX 3050 OEM's average benchmark score of 15199 sits 7.4% above the GTX 680's 14150. In the context of nearest rivals, the 3050 OEM is essentially tied with the AMD Radeon RX 7600 (15171, 0.2% delta) and the AMD Radeon 680M (15270, -0.5% delta), while the GTX 680 is bracketed by the Tesla K10 (14029, +0.9%) and the GTX 1070 Ti (14277, -0.9%). The percentile rankings reinforce this: the 3050 OEM sits at the 57th percentile of all GPUs, the 680 at the 55th. For the 3050 OEM, its Passmark G3D score of 11857 and Passmark GPU Compute score of 5779 are strong showings for its class. The GTX 680 lacks those Passmark entries entirely, so direct comparison there is impossible, but the Geekbench deltas are decisive.

# Specification Differences

The two cards differ on nearly every measurable specification. Process node: 8 nm (Samsung) for the 3050 OEM versus 28 nm (TSMC) for the 680. Transistor count: 12,000 million versus 3,540 million. Die size: 276 mm² versus 294 mm². Transistor density: 43.5M/mm² versus 12.0M/mm². Base clock: 1515 MHz versus 1006 MHz. Boost clock: 1755 MHz versus 1058 MHz. Memory clock: 1750 MHz (14 Gbps effective) versus 1502 MHz (6 Gbps effective). Memory size: 8 GB versus 2 GB. Memory type: GDDR6 versus GDDR5. Bus width: 128-bit versus 256-bit. Bandwidth: 224.0 GB/s versus 192.3 GB/s.

Shading units: 2304 versus 1536. TMUs: 72 versus 128. ROPs: 32 for both—one rare point of parity. RT cores: 18 versus none. Tensor cores: 72 versus none. Pixel rate: 56.16 GPixel/s versus 33.86 GPixel/s. Texture rate: 126.4 GTexel/s versus 135.4 GTexel/s—another spec where the older card wins narrowly. FP32 throughput: 8.087 TFLOPS versus 3.250 TFLOPS. FP16: 8.087 TFLOPS on the 3050 OEM, not specified on the 680. TDP: 130 W versus 195 W—the newer card draws less power. Power connectors: 1x 8-pin versus 2x 6-pin. Suggested PSU: 300 W versus 450 W. Bus interface: PCIe 4.0 x8 versus PCIe 3.0 x16. Display outputs: 1x HDMI 2.1 and 3x DisplayPort 1.4a versus 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. Dimensions: the 3050 OEM is 242 mm long and 112 mm tall; the 680 is 256 mm long, 111 mm tall, and 38 mm wide. Release dates: 2022-01-03 for the 3050 OEM versus 2012-03-21 for the 680. The 680 has a launch MSRP of 499 USD; the 3050 OEM has no listed launch MSRP.

# FAQ

Q: Which card is faster in compute workloads?

A: The RTX 3050 OEM wins decisively. In Geekbench OpenCL, it scores 60740 versus 16906 for the GTX 680, a 259.3% advantage. Its FP32 throughput of 8.087 TFLOPS more than doubles the 680's 3.250 TFLOPS.

Q: Does the GTX 680 have any advantage in texture processing?

A: Yes, in one narrow spec. The GTX 680 has 128 TMUs and a texture rate of 135.4 GTexel/s, while the RTX 3050 OEM has 72 TMUs and a texture rate of 126.4 GTexel/s. However, this does not translate into any benchmark win for the 680.

Q: How do the memory configurations compare?

A: The RTX 3050 OEM has 8 GB of GDDR6 on a 128-bit bus with 224.0 GB/s bandwidth. The GTX 680 has 2 GB of GDDR5 on a 256-bit bus with 192.3 GB/s bandwidth. The newer card offers four times the capacity and higher bandwidth despite the narrower bus.

Q: Which card supports modern graphics features?

A: The RTX 3050 OEM supports DirectX 12 Ultimate (12_2), Vulkan 1.4, and includes 18 ray tracing cores and 72 tensor cores. The GTX 680 supports only DirectX 12 (11_0) and Vulkan 1.2.175, with no ray tracing or tensor core hardware.

Q: What is the power consumption difference?

A: The RTX 3050 OEM has a TDP of 130 W and a suggested PSU of 300 W, using a single 8-pin connector. The GTX 680 has a TDP of 195 W and a suggested PSU of 450 W, requiring two 6-pin connectors. The newer card is more power-efficient.

Q: How do the two cards rank relative to all GPUs?

A: The RTX 3050 OEM sits at the 57th percentile of all GPUs with an average benchmark score of 15199. The GTX 680 sits at the 55th percentile with an average score of 14150. The 3050 OEM's nearest rival is the AMD Radeon RX 7600 (0.2% delta), while the 680's nearest rival is the NVIDIA Tesla K10 (0.9% delta).

# Where Each One Wins

The RTX 3050 OEM wins in every measurable benchmark category. Its 259.3% OpenCL lead and 223.6% Vulkan lead over the GTX 680 make it the clear choice for modern gaming, compute workloads, and any application that leverages DirectX 12 Ultimate features like ray tracing. The 8 GB memory capacity is essential for current game textures and larger datasets, while the 8.087 TFLOPS FP32 throughput handles compute tasks with headroom. The 130 W TDP also makes it easier to cool and power, requiring only a 300 W PSU and a single 8-pin connector.

The GTX 680 wins only in specific legacy scenarios. Its 128 TMUs and 135.4 GTexel/s texture rate edge out the 3050 OEM on paper, which could theoretically benefit older DirectX 9-era games that are texture-bound. Its 256-bit memory bus and 2 GB GDDR5 configuration is a relic of its era, but for very old titles with modest VRAM requirements, it may still function. The 680's dual 6-pin connectors and 450 W PSU requirement are drawbacks, not advantages. Its 33.86 GPixel/s pixel rate is substantially lower than the 3050 OEM's 56.16 GPixel/s, meaning even in fill-rate-bound scenarios, the newer card wins. The 680's Vulkan 1.2.175 support is functional but outdated, and its lack of ray tracing and tensor cores makes it unsuitable for any modern AI or RT workload. For anyone holding a GTX 680, the benchmark data says the upgrade to the RTX 3050 OEM is a 2.2x to 3.6x improvement depending on the API, with the added benefits of lower power draw and quadruple the memory.