AMD Radeon HD 6850 X2 vs NVIDIA GeForce GTX 650 Comparison

The AMD Radeon HD 6850 X2 and NVIDIA GeForce GTX 650 represent two very different approaches to graphics hardware from the early-to-mid 2010s. The HD 6850 X2 is a dual-GPU monster built on a mature 40 nm process, while the GTX 650 is a single-chip, power-efficient card on a newer 28 nm node. The data shows a clear split: the GTX 650 wins the only direct benchmark comparison, but the HD 6850 X2 retains higher raw theoretical throughput in several key areas. This analysis breaks down where each card holds an advantage, based strictly on the benchmark and specification data.

Where Each One Wins

The NVIDIA GeForce GTX 650 is the winner in the only head-to-head benchmark recorded. In Geekbench OpenCL, the GTX 650 scores 4545, which is 12.5% higher than the HD 6850 X2's score of 3977. This is a significant margin for a dual-GPU card against a single-chip solution. The GTX 650 also demonstrates broader API support, with scores in Geekbench Metal (2400) and Geekbench Vulkan (4524), while the HD 6850 X2 has no recorded scores beyond OpenCL.

However, the AMD Radeon HD 6850 X2 wins in pure compute specifications. It features 960 shading units, 48 texture mapping units (TMUs), and 32 raster operation units (ROPs), which are all higher than the GTX 650's 384 shading units, 32 TMUs, and 16 ROPs. The HD 6850 X2 also has a substantial advantage in fill rates: its pixel rate is 25.60 GPixel/s versus 8.464 GPixel/s for the GTX 650, and its texture rate is 38.40 GTexel/s versus 33.86 GTexel/s. The FP32 compute rating tells a similar story: 1.536 TFLOPS for the AMD card versus 812.5 GFLOPS for the NVIDIA card.

The practical takeaway is that the GTX 650 wins in real-world compute workloads as measured by OpenCL, but the HD 6850 X2 has higher theoretical peaks. The GTX 650 also wins on efficiency, with a 65 W TDP compared to the HD 6850 X2's 254 W TDP, and it requires only a 250 W suggested PSU versus 600 W for the AMD card.

Architecture Differences

The two cards are built on fundamentally different architectures. The AMD Radeon HD 6850 X2 uses the Barts chip based on TeraScale 2, part of the Northern Islands (HD 6800) generation. It is fabricated on a 40 nm process at TSMC, with 1,700 million transistors on a 255 mm² die. The transistor density is 6.7M per mm².

The NVIDIA GeForce GTX 650 uses the GK106 chip based on Kepler, part of the GeForce 600 generation. It is fabricated on a 28 nm process at TSMC, with 2,540 million transistors on a 221 mm² die. The transistor density is 11.5M per mm². This newer process allows the GTX 650 to pack more transistors into a smaller die, which explains its drastically lower power consumption despite having more transistors overall.

Memory configurations differ substantially. The HD 6850 X2 has 2 GB of GDDR5 memory on a 256-bit bus, yielding 134.4 GB/s of bandwidth. The GTX 650 has 1024 MB of GDDR5 on a 128-bit bus, yielding 80.00 GB/s. The effective memory clocks are 4.2 Gbps for the AMD card and 5 Gbps for the NVIDIA card. The GTX 650 also supports PCIe 3.0 x16, while the HD 6850 X2 is limited to PCIe 2.0 x16.

Feature support also diverges. The GTX 650 supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The HD 6850 X2 supports DirectX 11.2 (11_0) and OpenGL 4.4, with no Vulkan support. The GTX 650 also has a length of 147 mm (5.8 inches) and is single-slot, while the HD 6850 X2 is dual-slot.

Head-to-Head Benchmarks

The only direct benchmark comparison available is Geekbench OpenCL. The NVIDIA GeForce GTX 650 scores 4545, while the AMD Radeon HD 6850 X2 scores 3977. This represents a 12.5% advantage for the GTX 650, as indicated by the deltaPct of -12.5 from the perspective of the AMD card.

This result is surprising given the HD 6850 X2's higher theoretical compute. The AMD card has nearly double the FP32 throughput (1.536 TFLOPS vs 812.5 GFLOPS) and almost double the pixel rate (25.60 GPixel/s vs 8.464 GPixel/s). Yet in the actual OpenCL test, the GTX 650 comes out ahead. This suggests that the Kepler architecture's efficiency and the newer 28 nm process translate to better real-world execution, despite lower raw specs.

The GTX 650 also shows consistency across different APIs. Its OpenCL score of 4545 is nearly identical to its Vulkan score of 4524, with a Metal score of 2400. The HD 6850 X2 has no equivalent cross-API data, so its performance profile outside OpenCL is unknown.

When looking at the nearest rivals, the GTX 650's average score of 3823 places it just above the Intel UHD Graphics 710 (3792, 0.8% delta) and slightly below the AMD Radeon R5 Graphics (3883, -1.5% delta) and NVIDIA Quadro 2000 (3898, -1.9% delta). The HD 6850 X2's average score of 3977 sits close to the NVIDIA Quadro K2000 (3964, 0.3% delta) and NVIDIA GeForce 830M (3957, 0.5% delta). The percentile rankings are close too: 24th percentile for the HD 6850 X2 versus 22nd for the GTX 650.

FAQ

Q: Which card performs better in OpenCL?

A: The NVIDIA GeForce GTX 650 scores 4545 in Geekbench OpenCL, which is 12.5% higher than the AMD Radeon HD 6850 X2's score of 3977.

Q: Does the HD 6850 X2 have any performance advantages?

A: Yes, in theoretical specifications. The HD 6850 X2 has 960 shading units versus 384, a pixel rate of 25.60 GPixel/s versus 8.464 GPixel/s, and FP32 compute of 1.536 TFLOPS versus 812.5 GFLOPS.

Q: How do power requirements compare?

A: The HD 6850 X2 has a 254 W TDP and requires a 600 W suggested PSU with 2x 8-pin connectors. The GTX 650 has a 65 W TDP and requires only a 250 W suggested PSU with a single 6-pin connector.

Q: What are the memory differences?

A: The HD 6850 X2 has 2 GB of GDDR5 on a 256-bit bus with 134.4 GB/s bandwidth. The GTX 650 has 1024 MB of GDDR5 on a 128-bit bus with 80.00 GB/s bandwidth. The GTX 650 has a higher effective memory clock at 5 Gbps versus 4.2 Gbps.

Q: Which card supports more modern APIs?

A: The GTX 650 supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The HD 6850 X2 supports DirectX 11.2 (11_0) and OpenGL 4.4, with no Vulkan support.

Q: How do their benchmark scores compare to competitors?

A: The HD 6850 X2's average score of 3977 is 0.3% above the NVIDIA Quadro K2000 (3964). The GTX 650's average score of 3823 is 0.8% above the Intel UHD Graphics 710 (3792) but 1.5% below the AMD Radeon R5 Graphics (3883).

Specification Differences

The following fields differ between the two cards:

• Process Node: 40 nm (AMD) vs 28 nm (NVIDIA)
• Transistors: 1,700 million vs 2,540 million
• Die Size: 255 mm² vs 221 mm²
• Transistor Density: 6.7M / mm² vs 11.5M / mm²
• Memory Size: 2 GB vs 1024 MB
• Memory Bus Width: 256 bit vs 128 bit
• Memory Bandwidth: 134.4 GB/s vs 80.00 GB/s
• Effective Memory Clock: 4.2 Gbps vs 5 Gbps
• Shading Units: 960 vs 384
• TMUs: 48 vs 32
• ROPs: 32 vs 16
• Pixel Rate: 25.60 GPixel/s vs 8.464 GPixel/s
• Texture Rate: 38.40 GTexel/s vs 33.86 GTexel/s
• FP32: 1.536 TFLOPS vs 812.5 GFLOPS
• TDP: 254 W vs 65 W
• Slot Width: Dual-slot vs Single-slot
• Power Connectors: 2x 8-pin vs 1x 6-pin
• Suggested PSU: 600 W vs 250 W
• Bus Interface: PCIe 2.0 x16 vs PCIe 3.0 x16
• Display Outputs: 2x DVI, 1x HDMI 1.3a, 1x DisplayPort 1.1 vs 1x DVI, 2x DisplayPort 1.2
• DirectX Support: 11.2 (11_0) vs 12 (11_0)
• OpenGL Support: 4.4 vs 4.6
• Vulkan Support: None vs 1.2.175
• Length: Not specified vs 147 mm (5.8 inches)
• Release Date: 2011-09-18 vs 2013-11-26
• Predecessor: Evergreen vs GeForce 500
• Successor: Southern Islands vs GeForce 700

The Verdict

The benchmark data is unambiguous: the NVIDIA GeForce GTX 650 wins the only direct comparison, beating the AMD Radeon HD 6850 X2 by 12.5% in Geekbench OpenCL. If your priority is raw compute performance in that specific workload, the GTX 650 is the clear choice. It also offers better API support, including Vulkan, and operates at a fraction of the power draw—65 W versus 254 W.

The AMD Radeon HD 6850 X2, however, should not be dismissed. Its specifications are far more impressive on paper: 960 shading units, 32 ROPs, 2 GB of memory on a 256-bit bus, and 1.536 TFLOPS of FP32 compute. These numbers suggest it could excel in workloads that leverage high fill rates or large memory bandwidth, though the benchmark data does not confirm this. Its 24th percentile ranking is slightly better than the GTX 650's 22nd percentile.

For a practical recommendation: choose the GTX 650 if you want confirmed performance, modern API support, and minimal power requirements. Choose the HD 6850 X2 only if your workload specifically benefits from its higher theoretical throughput and larger memory pool, and you can accommodate its 600 W PSU requirement. The data supports the GTX 650 as the more reliable performer in real-world testing.