AMD Radeon 680M vs NVIDIA GeForce GTX 965M Comparison

FAQ

Q: Which GPU wins in raw benchmark performance?

A: The AMD Radeon 680M wins both head-to-head benchmark tests. It scores 23,468 in Geekbench OpenCL versus 14,509 for the NVIDIA GeForce GTX 965M, a 61.7% advantage. In Geekbench Vulkan, the 680M scores 21,965 against 14,299, a 53.6% lead.

Q: How do these GPUs rank against all other GPUs in the database?

A: The AMD Radeon 680M sits at the 57th percentile of all GPUs, while the NVIDIA GeForce GTX 965M sits at the 56th percentile. Their average benchmark scores are 15,270 and 14,404 respectively, placing them nearly adjacent in the overall hierarchy.

Q: What are the closest rivals for each GPU?

A: The 680M's nearest rivals are the NVIDIA GeForce GTX 580 (15,283, 0.1% behind), NVIDIA GeForce RTX 2060 (15,290, 0.1% behind), NVIDIA GeForce RTX 3050 OEM (15,199, 0.5% ahead), and AMD Radeon RX 7600 (15,171, 0.7% ahead). The GTX 965M's nearest rivals include the AMD Radeon RX Vega 11 (14,385, 0.1% behind), NVIDIA GeForce GTX TITAN (14,373, 0.2% behind), and AMD Radeon Vega 11 (14,352, 0.4% behind).

Q: What is the architecture generation gap between these two?

A: The AMD Radeon 680M uses RDNA 2.0 architecture on TSMC's 6 nm process, while the NVIDIA GeForce GTX 965M uses Maxwell 2.0 on a 28 nm process. The 680M is from the Navi II IGP generation released in 2023, whereas the GTX 965M is from the GeForce 900M generation released in 2015.

Q: Does the GTX 965M have any advantage in specifications?

A: Yes. The GTX 965M has more shading units (1,024 versus 768), more texture mapping units (64 versus 48), and dedicated 2 GB GDDR5 memory with an 80.19 GB/s bandwidth. The 680M uses system-shared memory with bandwidth that is system dependent.

Q: Which GPU supports ray tracing?

A: The AMD Radeon 680M has 12 ray tracing cores and supports DirectX 12 Ultimate (12_2). The NVIDIA GeForce GTX 965M has no ray tracing cores and only supports DirectX 12 (12_1).

Where Each One Wins

The AMD Radeon 680M wins decisively in every benchmark category where both GPUs were tested. The data shows a clean sweep: two wins for the 680M, zero for the GTX 965M.

The largest margin comes in Geekbench OpenCL, where the 680M scores 23,468 against the GTX 965M's 14,509 — a 61.7% advantage. That gap is substantial enough to place the 680M in a different performance tier for compute-heavy workloads like OpenCL acceleration. The Geekbench Vulkan result is similarly lopsided at 53.6% in favor of the 680M, with scores of 21,965 versus 14,299.

The GTX 965M does hold the specification advantage in raw hardware counts — more shading units, more TMUs, and dedicated VRAM — but benchmark results indicate those resources are not translating into competitive scores. The 680M's newer architecture and higher clock speeds (2,000 MHz base, 2,200 MHz boost versus 924 MHz base, 950 MHz boost) more than compensate.

For users choosing between these two, the data is unambiguous: the AMD Radeon 680M wins every measured performance comparison. The GTX 965M's only practical advantage is its older driver ecosystem and the fact that it was designed for MXM modules in gaming laptops of its era, but that is not reflected in any benchmark win.

Architecture Differences

The architectural gap between these two GPUs is generational. The AMD Radeon 680M is built on RDNA 2.0 architecture fabricated on TSMC's 6 nm process node, packing 13,100 million transistors into a 208 mm² die. That yields a transistor density of 63.0 million per mm². The NVIDIA GeForce GTX 965M uses Maxwell 2.0 architecture on a 28 nm process, containing 5,200 million transistors on a much larger 398 mm² die, resulting in just 13.1 million transistors per mm².

The 680M integrates 12 ray tracing cores, making it DirectX 12 Ultimate (12_2) capable. The GTX 965M has no ray tracing support and tops out at DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4, but the 680M's newer feature set gives it access to modern rendering techniques that the older Maxwell part cannot match.

Memory architecture differs fundamentally. The 680M uses system-shared memory with bandwidth dependent on the host system, while the GTX 965M has dedicated 2 GB GDDR5 on a 128-bit bus with 80.19 GB/s bandwidth. The 680M's memory clock is listed as "System Shared," meaning its performance scales with system RAM speed.

The 680M's FP32 compute throughput is 3.379 TFLOPS with FP16 support at 6.758 TFLOPS (2:1). The GTX 965M delivers 1.946 TFLOPS FP32 and has no listed FP16 capability. Pixel rates also favor the 680M at 70.40 GPixel/s versus 30.40 GPixel/s, and texture rates at 105.6 GTexel/s versus 60.80 GTexel/s.

Process node advantages are stark: 6 nm versus 28 nm. This allows the 680M to deliver roughly 2.5 times the transistor density while using a fraction of the die area. The 680M's TDP is rated at 50 W as an integrated graphics processor, while the GTX 965M's TDP is not specified in the data.

Specification Differences

The two GPUs differ across nearly every measurable specification. Starting with process node, the 680M uses 6 nm versus the GTX 965M's 28 nm. Transistor count is 13,100 million versus 5,200 million. Die size is 208 mm² versus 398 mm². Transistor density is 63.0M per mm² versus 13.1M per mm².

Clock speeds show a massive difference: the 680M runs at 2,000 MHz base and 2,200 MHz boost, while the GTX 965M runs at 924 MHz base and 950 MHz boost. Memory configuration differs entirely — the 680M uses system-shared memory, while the GTX 965M has 2 GB GDDR5 with a 128-bit bus and 80.19 GB/s bandwidth.

Compute resources are split: the 680M has 768 shading units, 48 TMUs, and 32 ROPs, while the GTX 965M has 1,024 shading units, 64 TMUs, and 32 ROPs. The 680M adds 12 ray tracing cores; the GTX 965M has none. The 680M achieves 70.40 GPixel/s pixel rate and 105.6 GTexel/s texture rate, versus 30.40 GPixel/s and 60.80 GTexel/s for the GTX 965M.

FP32 throughput is 3.379 TFLOPS for the 680M versus 1.946 TFLOPS for the GTX 965M. The 680M also lists FP16 performance at 6.758 TFLOPS (2:1), which the GTX 965M lacks entirely. Power delivery differs: the 680M is an IGP with 50 W TDP and no power connectors, while the GTX 965M is an MXM module with no power connectors and unspecified TDP.

Bus interface and form factor separate them further: the 680M uses PCIe 4.0 x8 as an integrated processor, while the GTX 965M uses MXM-B (3.0) as a removable module. Production status shows the 680M as Active while the GTX 965M is End-of-life.

Head-to-Head Benchmarks

The head-to-head data covers two tests, and the AMD Radeon 680M wins both by wide margins.

In Geekbench OpenCL, the 680M scores 23,468 against the GTX 965M's 14,509. That is a delta of 61.7% in favor of the 680M. This test measures general-purpose compute performance across a variety of workloads, and the 680M's newer RDNA 2.0 architecture with its higher clock speeds and FP32 throughput delivers a crushing advantage. The GTX 965M's higher shading unit count (1,024 versus 768) cannot overcome the 680M's per-clock efficiency and 2,200 MHz boost clock.

Geekbench Vulkan shows a similar story: the 680M scores 21,965 versus 14,299 for the GTX 965M, a 53.6% delta. Vulkan is a low-level graphics API, and the 680M's support for DirectX 12 Ultimate features alongside its ray tracing cores gives it architectural advantages that show up even in compute-oriented workloads. The GTX 965M's Maxwell 2.0 architecture, while competent for its era, lacks the modern feature set and throughput of RDNA 2.0.

The average benchmark score reinforces this: the 680M averages 15,270 across all its benchmarks, while the GTX 965M averages 14,404. The 680M's nearest rivals include the RTX 2060 and GTX 580, both within 0.1% of its average score. The GTX 965M's nearest rivals cluster around the RX Vega 11 and GTX TITAN, all within 0.6% of its score.

The Verdict

The benchmark data delivers a clear verdict: the AMD Radeon 680M is the superior GPU in every measured category. It wins both head-to-head tests with margins exceeding 53%, and its average benchmark score of 15,270 places it alongside discrete desktop GPUs like the RTX 2060 and GTX 580.

Users should pick the AMD Radeon 680M if they need modern API support, ray tracing capability, and substantially higher compute throughput. Its 61.7% lead in OpenCL and 53.6% lead in Vulkan are not marginal differences — they represent a generational leap in performance. The 680M also benefits from being an active product with a 2023 release date, while the GTX 965M is end-of-life.

The NVIDIA GeForce GTX 965M retains one practical advantage: dedicated GDDR5 memory with 80.19 GB/s bandwidth. For applications that are sensitive to memory latency or require dedicated VRAM, this could matter. Its higher shading unit count (1,024 versus 768) and TMU count (64 versus 48) suggest it was designed with raw throughput in mind, but the benchmark results show those resources are not being utilized effectively compared to the 680M's modern architecture.

The 680M's 57th percentile ranking versus the GTX 965M's 56th percentile is close, but that metric masks the magnitude of the head-to-head deltas. In direct comparison, the 680M outperforms the GTX 965M by over 60% in one test and over 50% in another. The GTX 965M's nearest rivals — the RX Vega 11 and GTX TITAN — are all within 0.6% of its score, confirming it sits at a lower performance tier.

For anyone choosing between these two today, the data supports only one recommendation: the AMD Radeon 680M delivers decisively better performance, modern feature support, and active production status. The GTX 965M is a legacy product whose specifications no longer translate into competitive benchmark results.