AMD Radeon HD 6770 vs NVIDIA GeForce GTX 460M Comparison

Where Each One Wins

The benchmark data presents a clear split between these two mobile and desktop graphics solutions. The NVIDIA GeForce GTX 460M wins the only recorded head-to-head benchmark, the Geekbench OpenCL test, with a score of 4282 against the AMD Radeon HD 6770's 3649. That is a 17.3% advantage for the NVIDIA part, a substantial margin in raw compute throughput.

The GTX 460M's win is not marginal. It places the NVIDIA chip in the 25th percentile of all GPUs in the database, while the AMD card sits at the 22nd percentile. This means the GTX 460M outperforms a larger share of the overall GPU population than the HD 6770 does, even though both are long end-of-life products. The AMD card's nearest rivals include the NVIDIA GeForce GTX 1050 at 3629 (0.6% behind) and the NVIDIA GeForce GT 735M at 3616 (0.9% behind), which shows the HD 6770 is competitive with much newer entry-level parts in OpenCL compute workloads.

Where the AMD card would win is in any scenario that depends on pixel throughput, texture fill, or memory bandwidth. The HD 6770 has a pixel rate of 13.60 GPixel/s versus the GTX 460M's 5.400 GPixel/s, which is a 2.5x advantage. Its texture rate of 34.00 GTexel/s beats the NVIDIA chip's 21.60 GTexel/s by roughly 57%. Memory bandwidth also favors AMD at 76.80 GB/s versus 60.00 GB/s, a 28% lead. These numbers suggest that for older games or applications that are fill-rate limited rather than compute limited, the HD 6770 could close the gap or even take the lead, but the database's single recorded benchmark does not capture that scenario.

Architecture Differences

The two GPUs come from fundamentally different architectural philosophies. The NVIDIA GeForce GTX 460M uses the Fermi architecture, built on the GF106 chip, while the AMD Radeon HD 6770 uses TeraScale 2, built on the Juniper chip. Both are fabricated on a 40 nm process at TSMC, so the node is identical, but the implementation differs drastically.

The GTX 460M packs 1,170 million transistors onto a 238 mm² die, yielding a transistor density of 4.9M per mm². The HD 6770 uses 1,040 million transistors on a smaller 166 mm² die, giving a higher density of 6.3M per mm². This means AMD fit nearly as many transistors into a much smaller area, which is a sign of a more compact design.

The shading unit counts tell a striking story. The AMD card has 800 shading units, while the NVIDIA card has only 192. Yet the NVIDIA card still wins the OpenCL benchmark by 17.3%. This indicates that Fermi's shading units are far more efficient per unit than TeraScale 2's, or that the OpenCL workload favors NVIDIA's scheduler and memory architecture. The AMD card compensates with 40 texture mapping units versus 32, and 16 ROPs versus 24 for NVIDIA. The AMD card's higher ROP count is surprising given its lower overall pixel rate, but the clock differences explain this: the HD 6770's memory runs at 1200 MHz (4.8 Gbps effective) versus the GTX 460M's 625 MHz (2.5 Gbps effective), which directly impacts fill rates.

The FP32 compute figures are lopsided. The HD 6770 delivers 1,360.0 GFLOPS, while the GTX 460M manages only 518.4 GFLOPS. This is a 2.6x raw compute advantage for AMD, yet the OpenCL benchmark still favors NVIDIA. This suggests the benchmark is not purely arithmetic-bound, or that driver optimization and memory latency play a larger role than raw FLOP counts.

The Verdict

For anyone choosing between these two today, the data points to the NVIDIA GeForce GTX 460M as the stronger overall performer in compute workloads. Its 17.3% lead in the Geekbench OpenCL score, combined with a higher percentile rank (25th versus 22nd), makes it the safer pick for general-purpose GPU computing or any task that leverages OpenCL acceleration.

However, the AMD Radeon HD 6770 is not without merit. Its 2.5x advantage in pixel rate and 57% lead in texture rate mean it will handle older, fill-rate-heavy games better. The HD 6770 also has a higher FP32 throughput at 1,360.0 GFLOPS, which could matter for specific compute kernels that are not captured by the single benchmark. The AMD card's memory bandwidth advantage of 28% (76.80 GB/s versus 60.00 GB/s) also helps in bandwidth-sensitive scenarios.

The practical advice: if your workload is OpenCL compute, take the GTX 460M. If you are dealing with legacy DirectX 11 games from that era, the HD 6770's higher fill rates might serve you better, but the database does not provide game benchmarks to confirm this. The GTX 460M is also a mobile MXM module with no power connectors and a 50 W TDP, while the HD 6770 is a dual-slot desktop card requiring a 300 W power supply and a single 6-pin connector. If power draw or form factor matters, the NVIDIA part is far easier to accommodate.

FAQ

Q: Which GPU has a higher OpenCL benchmark score?

A: The NVIDIA GeForce GTX 460M scores 4282, which is 17.3% higher than the AMD Radeon HD 6770's 3649.

Q: Does the AMD card have more shading units?

A: Yes, the HD 6770 has 800 shading units, while the GTX 460M has 192. Despite this 4x difference, the NVIDIA card wins the OpenCL benchmark.

Q: Which card has better memory bandwidth?

A: The AMD Radeon HD 6770 has 76.80 GB/s of bandwidth, versus 60.00 GB/s for the GTX 460M, a 28% advantage for AMD.

Q: What are the power requirements for each card?

A: The GTX 460M has a 50 W TDP and uses an MXM module with no power connectors. The HD 6770 has a 108 W TDP, requires a 300 W power supply, and uses a single 6-pin power connector.

Q: Are both GPUs on the same manufacturing process?

A: Yes, both use a 40 nm process at TSMC. The GTX 460M uses a 238 mm² die, while the HD 6770 uses a smaller 166 mm² die.

Q: Which card has a higher pixel fill rate?

A: The AMD Radeon HD 6770 has a pixel rate of 13.60 GPixel/s, which is 2.5x higher than the GTX 460M's 5.400 GPixel/s.

Head-to-Head Benchmarks

The only direct comparison in the database is the Geekbench OpenCL test. The NVIDIA GeForce GTX 460M scores 4282, and the AMD Radeon HD 6770 scores 3649. The delta is 17.3% in favor of NVIDIA. This is a decisive margin, placing the GTX 460M above the AMD FirePro W2100 (4295, 0.3% behind) and the NVIDIA Quadro K3000M (4241, 1% behind) in the broader ranking. The HD 6770, meanwhile, sits just above the NVIDIA GeForce GTX 1050 (3629, 0.6% behind) and the NVIDIA GeForce GT 735M (3616, 0.9% behind), showing it is competitive with much newer entry-level parts.

The biggest wins for the GTX 460M are in the OpenCL compute score and the overall percentile ranking. The biggest wins for the HD 6770 are in pixel rate (13.60 versus 5.400 GPixel/s), texture rate (34.00 versus 21.60 GTexel/s), FP32 throughput (1,360.0 versus 518.4 GFLOPS), and memory bandwidth (76.80 versus 60.00 GB/s). These are not captured in the single benchmark, so they represent areas where the AMD card could outperform in workloads that are not compute-bound.

Specification Differences

The two cards differ in nearly every measurable specification except for the process node (both 40 nm) and the foundry (TSMC).

• Chip and Architecture: NVIDIA uses GF106 with Fermi architecture; AMD uses Juniper with TeraScale 2.
• Transistors: NVIDIA has 1,170 million; AMD has 1,040 million.
• Die Size: NVIDIA is 238 mm²; AMD is 166 mm².
• Transistor Density: NVIDIA is 4.9M per mm²; AMD is 6.3M per mm².
• Memory Speed: NVIDIA runs at 625 MHz (2.5 Gbps effective); AMD runs at 1200 MHz (4.8 Gbps effective).
• Memory Size: NVIDIA has 1536 MB; AMD has 1024 MB.
• Memory Bus: NVIDIA uses 192 bit; AMD uses 128 bit.
• Memory Bandwidth: NVIDIA has 60.00 GB/s; AMD has 76.80 GB/s.
• Shading Units: NVIDIA has 192; AMD has 800.
• Texture Mapping Units: NVIDIA has 32; AMD has 40.
• Render Output Units: NVIDIA has 24; AMD has 16.
• Pixel Rate: NVIDIA is 5.400 GPixel/s; AMD is 13.60 GPixel/s.
• Texture Rate: NVIDIA is 21.60 GTexel/s; AMD is 34.00 GTexel/s.
• FP32: NVIDIA is 518.4 GFLOPS; AMD is 1,360.0 GFLOPS.
• TDP: NVIDIA is 50 W; AMD is 108 W.
• Slot Width: NVIDIA is MXM Module; AMD is Dual-slot.
• Power Connectors: NVIDIA has none; AMD has 1x 6-pin.
• Suggested PSU: NVIDIA has none listed; AMD is 300 W.
• Display Outputs: NVIDIA is portable device dependent; AMD is 2x DVI, 1x HDMI 1.3a, 2x mini-DisplayPort 1.1.
• DirectX: NVIDIA supports 12 (11_0); AMD supports 11.2 (11_0).
• OpenGL: NVIDIA supports 4.6; AMD supports 4.4.
• Release Date: NVIDIA launched on 2010-09-02; AMD launched on 2011-01-20.
• Predecessor: NVIDIA's is GeForce 300M; AMD's is Evergreen.
• Successor: NVIDIA's is GeForce 500M; AMD's is Southern Islands.
• Launch MSRP: AMD has a launch MSRP of 129 USD; NVIDIA has no listed MSRP.
• Card Length: AMD is 198 mm (7.8 inches); NVIDIA has no listed length.