AMD Radeon HD 6850 X2 vs NVIDIA GeForce 920M Comparison

Head-to-Head Benchmarks

The only direct benchmark comparison recorded in the database is the Geekbench OpenCL test, and it shows a clear, though not overwhelming, victory for the AMD Radeon HD 6850 X2. The AMD part scores 3977 points against 3725 for the NVIDIA GeForce 920M, a delta of 6.8%. This is a meaningful margin in synthetic compute workloads, but it is far from a generational gap; both cards operate in the same performance tier according to their overall database percentiles.

Looking at the broader context, the AMD Radeon HD 6850 X2 sits at the 24th percentile of all GPUs in the database, while the NVIDIA GeForce 920M sits at the 20th percentile. This four-point percentile gap reinforces the OpenCL result: the AMD card is modestly ahead, but neither part is a performance leader. The AMD card's average benchmark score is 3977, while the GeForce 920M's average is 3287, a difference of 690 points. That average gap is larger than the single head-to-head OpenCL delta because the GeForce 920M's Vulkan score of 2849 drags its average down significantly. The AMD card has no Vulkan score recorded in the database, so its average is identical to its OpenCL result.

The nearest rivals for each card further contextualize the results. The AMD Radeon HD 6850 X2's closest competitors in the database are the NVIDIA Quadro K2000 at 3964 (0.3% behind), the NVIDIA GeForce 830M at 3957 (0.5% behind), the AMD Radeon R5 M420 at 3956 (0.5% behind), and the NVIDIA GeForce GT 745M at 3953 (0.6% behind). This cluster of scores within a 24-point spread indicates that the HD 6850 X2 is essentially tied with a group of mid-range mobile and workstation parts from a slightly later era. Its 6.8% lead over the GeForce 920M is therefore not a sign of class dominance, but rather a reflection of where the 920M sits relative to this broader pack.

The NVIDIA GeForce 920M's nearest rivals tell a similar story from the other direction. The GeForce GT 730M scores 3316, which is 0.9% ahead of the 920M. The Intel HD Graphics 530 scores 3332, 1.4% ahead. The GeForce GT 640 scores 3210, which the 920M beats by 2.4%. The Intel HD Graphics P4600 scores 3389, 3% ahead. The 920M's average of 3287 places it in a tight battle with integrated graphics and low-end discrete parts. Its OpenCL score of 3725 is notably higher than its average, meaning the Vulkan result of 2849 is the outlier that pulls it down. In OpenCL alone, the 920M is much closer to the AMD card than the average scores suggest.

The head-to-head delta of 6.8% in OpenCL is the single most important number for direct comparison. It shows the AMD card winning, but the margin is within the range of what a driver optimization or a different workload mix could shift. Neither card has a second recorded benchmark that aligns with the other's, so this single test remains the only apples-to-apples comparison available in the database.

The Verdict

The data points to the AMD Radeon HD 6850 X2 as the stronger compute performer. Its 3977 OpenCL score beats the GeForce 920M's 3725 by 6.8%, and its 24th percentile standing versus the 920M's 20th percentile confirms a consistent, if modest, advantage. For anyone running OpenCL workloads, the AMD card is the safer choice based strictly on the recorded measurements.

However, the verdict is not a clean sweep. The GeForce 920M has a Vulkan score of 2849, an API the AMD card does not support according to the database (its Vulkan field is null). This is a significant functional difference. If the target application relies on Vulkan, the GeForce 920M is the only option of the two that can run it. The AMD card's API support tops out at DirectX 11.2 (11_0), OpenGL 4.4, and no Vulkan. The 920M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The NVIDIA card is therefore more future-proof in terms of software compatibility, even if it loses on raw compute performance.

The 920M's power characteristics also matter for real-world deployment. The AMD card draws 254 W according to the TDP field, requires dual-slot cooling, and needs two 8-pin power connectors plus a 600 W suggested PSU. The 920M is an IGP-class IGP with a 33 W TDP and no power connectors. This is a massive difference in thermal and power requirements. The AMD card is a desktop-only design with a 40 nm process node, while the 920M is built on 28 nm. The 920M's 33 W TDP makes it suitable for laptops and portable devices, whereas the HD 6850 X2 cannot fit that use case at all.

For an end user, the choice is straightforward. If the workload is OpenCL compute and desktop installation are the only requirements, the AMD card wins. If power efficiency, compact form factor, or Vulkan support are priorities, the GeForce 920M is the correct pick. The database shows a 6.8% compute lead for AMD, but a modern API set and a 221 W TDP difference for NVIDIA. There is no overall winner; there is only the right tool for the specific need.

Architecture Differences

The two GPUs come from different design philosophies and different eras. The AMD Radeon HD 6850 X2 uses the Barts chip built on TSMC's 40 nm process. It packs 1,700 million transistors into a 255 mm² die, yielding a transistor density of 6.7 million per mm². The architecture is TeraScale 2, belonging to the Northern Islands generation (HD 6800 series). Its predecessor is Evergreen and its successor is Southern Islands. The release date recorded in the database is September 2011.

The NVIDIA GeForce 920M, by contrast, uses the GK208B chip on TSMC's 28 nm process. It contains 1,020 million transistors on a much smaller 87 mm² die, giving a higher density of 11.7 million per mm². The architecture is Kepler 2.0, from the GeForce 900M generation. Its predecessor is GeForce 800M and its successor is GeForce 10 Mobile. The release date is March 2015.

The memory subsystems could not be more different in type. The AMD card has 2 GB of GDDR5 running at 1050 MHz, translating to 4.2 Gbps effective, on a 256 bit bus for 134.4 GB/s of bandwidth. The GeForce 920M has 2 GB of DDR3 at 900 MHz, or 1800 Mbps effective, on a 64 bit bus for just 14.4 GB/s. The bandwidth gap is enormous: the AMD card offers 134.4 GB/s versus 14.4 GB/s for the NVIDIA part. That is a 120 GB/s difference, and it directly impacts fill-rate-bound workloads. The AMD card's pixel rate is 25.60 GPixel/s versus 7.632 GPixel/s for the 920M. The texture rates are closer: 38.40 GTexel/s for AMD versus 30.53 GTexel/s for NVIDIA.

Compute unit counts also favor the AMD card. The HD 6850 X2 has 960 shading units, 48 TMUs, and 32 ROPs. The GeForce 920M has 384 shading units, 32 TMUs, and only 8 ROPs. FP32 throughput is 1.536 TFLOPS for AMD versus 732.7 GFLOPS for NVIDIA, meaning the AMD card delivers more than double the single-precision floating-point performance. Neither card has RT cores or tensor cores recorded, and neither has FP16 data in the database.

The AMD card uses PCIe 2.0 x16, while the 920M uses PCIe 3.0 x8. The AMD card's display outputs are 2x DVI, 1x HDMI 1.3a, and 1x DisplayPort 1.1. The 920M's outputs are listed as "Portable Device Dependent," meaning it is designed for laptop panels. The AMD card is dual-slot and requires 2x 8-pin power connectors plus a 600 W PSU. The 920M is an IGP-class IGP with no power connectors and no PSU recommendation. The power envelope difference, 254 W versus 33 W, is the starkest architectural contrast on paper. The physical design choices follow directly: one is a power-hungry desktop board, the other is a low-power mobile part.

FAQ

Q: Which GPU has the higher OpenCL score?

A: The AMD Radeon HD 6850 X2 scores 3977 in Geekbench OpenCL, while the NVIDIA GeForce 920M scores 3725. The AMD card wins by 6.8%.

Q: Does the NVIDIA GeForce 920M support Vulkan?

A: Yes, the GeForce 920M has a Vulkan score of 2849 and lists Vulkan API version 1.2.175 support. The AMD Radeon, by contrast, has no Vulkan support listed in the database.

Q: How do the memory bandwidth figures compare?

A: The AMD card has 134.4 GB/s of bandwidth from 2 GB of GDDR5 on a 256 bit bus. The GeForce 920M has 14.4 GB/s from 2 GB of DDR3 on a 64 bit bus.

Q: What is the TDP difference between the two?

A: The AMD Radeon HD 6850 X2 has a TDP of 254 W and requires a 600 W suggested PSU. The GeForce 920M has a TDP of 33 W and has no PSU recommendation.

Q: Which GPU has more shading units?

A: The AMD card has 960 shading units, while the GeForce 920M has 384 shading units. The AMD card also has more TMUs (48 versus 32) and more ROPs (32 versus 8).

Q: What is the average benchmark score for each?

A: The AMD Radeon HD 6850 X2 has an average score of 3977, while the GeForce 920M averages 3287 across its two recorded tests.

Where Each One Wins

The AMD Radeon HD 6850 X2 wins in raw compute. Its OpenCL score is 6.8% higher, its FP32 throughput is 1.536 TFLOPS versus 732.7 GFLOPS, and its pixel rate is 25.60 GPixel/s versus 7.632 GPixel/s. For any workload that stresses shading units, texture units, or memory bandwidth, the AMD card is clearly ahead. The 134.4 GB/s of bandwidth versus 14.4 GB/s is a factor of over nine, making the AMD card the choice for high-resolution textures and heavy bandwidth utilization. Its 32 ROPs versus 8 ROPs also means it handles fill-rate-limited scenarios much better.

The NVIDIA GeForce 920M wins in efficiency and API coverage. At 33 W versus 254 W, it consumes a fraction of the power, and its 28 nm process with 11.7M transistors per mm² shows a more modern design despite lower absolute performance. The 920M supports Vulkan 1.2.175 and DirectX 12 (11_0), while the AMD card lacks Vulkan entirely and maxes out at DirectX 11.2. For any modern application built on Vulkan, the 920M is the only functional choice. The 920M also wins on physical integration: it is an IGP-class part with no power connectors, suitable for portable devices, while the AMD card needs dual-slot space and two 8-pin connectors.

The 920M's nearest rival data shows it is competitive with Intel integrated graphics like the HD Graphics 530 (1.4% behind) and the HD Graphics P4600 (3% behind), while beating the GeForce GT 640 by 2.4%. The AMD card sits in a cluster of mid-range parts like the Quadro K2000 and GeForce GT 830M, all within 0.6% of each other. The database does not record a single workload where the 920M beats the AMD card head-to-head; the only direct test is OpenCL, which AMD wins. But the 920M wins in portability, power, and software support, which are real advantages that the benchmark scores do not capture.

For a desktop user with a capable PSU and a need for compute throughput, the AMD card wins without question. For a laptop user or anyone running Vulkan-based software, the GeForce 920M is the only viable option, despite its lower scores. The data supports both, depending entirely on the deployment context.