AMD Radeon RX 5700 XT vs NVIDIA GeForce GTX 965M Comparison

The AMD Radeon RX 5700 XT and the NVIDIA GeForce GTX 965M occupy very different corners of the GPU landscape: one is a desktop graphics card built on the Navi 10 chip, the other a mobile MXM module derived from Maxwell 2.0 silicon. The database records both as end-of-life products, yet their benchmark profiles diverge in ways that go well beyond raw compute, and the head-to-head data shows the two trading wins depending on the API being exercised. What follows is a step-by-step walkthrough of the recorded specs, benchmark results, and what each card actually delivers.

FAQ

Q: Which GPU is faster overall according to the database?

A: The average benchmark scores are 16361 for the RX 5700 XT versus 14404 for the GTX 965M, a gap in AMD's favor. However, the percentile figures tell a more nuanced story: the RX 5700 XT sits at the 59th percentile against all GPUs, while the GTX 965M sits at the 56th, a surprisingly narrow spread given the hardware differences.

Q: Which GPU won the head-to-head benchmarks?

A: They split one apiece. The GTX 965M won Geekbench OpenCL with 14509 points against 9524, a 34.4 percent margin in NVIDIA's favor. The RX 5700 XT won Geekbench Vulkan with 68310 points against 14299, a 377.7 percent margin, which is one of the largest deltas the database recorded between these two.

Q: How much more memory does the RX 5700 XT have?

A: The RX 5700 XT ships with 8 GB of GDDR6 on a 256-bit bus delivering 448.0 GB/s of bandwidth. The GTX 965M ships with 2 GB of GDDR5 on a 128-bit bus delivering 80.19 GB/s. That is a fourfold capacity advantage and more than a fivefold bandwidth advantage for the AMD card.

Q: Are either of these GPUs still in production?

A: No. Both are flagged as end-of-life in the database. The RX 5700 XT was released in 2019 with a launch MSRP of 399 USD, while the GTX 965M was released in 2015 and has no launch MSRP recorded.

Q: Do either of these GPUs have ray tracing or AI acceleration hardware?

A: No. Both list null entries for RT cores and tensor cores. Neither architecture, RDNA 1.0 nor Maxwell 2.0, includes dedicated hardware for those workloads.

Architecture Differences

The two GPUs come from different eras of semiconductor design. The RX 5700 XT is built on TSMC's 7 nm process, packing 10,300 million transistors onto a 251 mm² die for a density of 41.0M transistors per square millimeter. The GTX 965M uses TSMC's 28 nm process with 5,200 million transistors spread across a larger 398 mm² die, yielding just 13.1M per square millimeter. The AMD part roughly doubles transistor count on a die nearly 150 mm² smaller, which illustrates how far process technology moved between these two generations.

Architecturally, the RX 5700 XT uses RDNA 1.0 on the Navi 10 chip, the first generation of the Radeon RX 5000 series, succeeding Vega and preceding Navi II. The GTX 965M uses Maxwell 2.0 on the GM204 chip, part of the GeForce 900M mobile lineup, sitting between the GeForce 800M predecessor and GeForce 10 Mobile successor.

Compute resources scale accordingly. The RX 5700 XT carries 2560 shading units, 160 TMUs, and 64 ROPs; the GTX 965M carries 1024 shading units, 64 TMUs, and 32 ROPs. That is a consistent 2.5x to 2x advantage across the rendering pipeline for the AMD card. Theoretical throughput backs this up: 9.754 TFLOPS FP32 against 1.946 TFLOPS, along with 19.51 TFLOPS FP16 on the RDNA part, a capability the database records no equivalent figure for on the Maxwell part. Pixel and texture fill rates follow the same pattern, at 121.9 versus 30.40 GPixel/s and 304.8 versus 60.80 GTexel/s.

Platform integration differs as well. The RX 5700 XT is a dual-slot PCIe 4.0 x16 desktop card with fixed display outputs, one HDMI 2.0b and three DisplayPort 1.4a. The GTX 965M is an MXM-B (3.0) module with no power connectors and portable-device-dependent display outputs, confirming it was designed exclusively for laptop integration rather than desktop use.

The Verdict

The data points to the RX 5700 XT for anyone whose workload rewards raw throughput and modern graphics APIs. Its Vulkan result of 68310 is nearly five times the GTX 965M's 14299, and its broader benchmark coverage (eleven recorded tests versus the NVIDIA card's two) shows a card engineered for desktop gaming and compute. Its nearest rivals in the database include the AMD Radeon Pro 5600M, NVIDIA RTX PRO 6000 Blackwell, AMD Radeon PRO W7500, and NVIDIA GeForce RTX 5090 D V2, all within roughly one percent of its average score.

The GTX 965M's case is narrower but real. It won the only OpenCL head-to-head by 34.4 percent, which matters for OpenCL-driven compute workloads, and its own rival cluster (Radeon RX Vega 11, GTX TITAN, Radeon Vega 11, Iris Xe MAX) sits within a fraction of a percent of its average. It is also a mobile module, so it competes in a form factor the RX 5700 XT cannot enter at all. For a portable system where the OpenCL result is the relevant workload, the recorded data favors it. In every other measurable dimension, capacity, bandwidth, fill rates, and theoretical compute, the AMD card leads decisively.

Specification Differences

The fields where these two GPUs diverge are extensive:

• Process node: 7 nm (RX 5700 XT) versus 28 nm (GTX 965M), both on TSMC
• Transistors and die: 10,300 million on 251 mm² versus 5,200 million on 398 mm²
• Clocks: base 1605 MHz, boost 1905 MHz, game clock 1755 MHz for AMD; base 924 MHz, boost 950 MHz for NVIDIA. Memory clocks run at 14 Gbps effective versus 5 Gbps effective.
• Memory: 8 GB GDDR6, 256-bit, 448.0 GB/s versus 2 GB GDDR5, 128-bit, 80.19 GB/s
• Shader resources: 2560 shading units, 160 TMUs, 64 ROPs versus 1024, 64 TMUs, 32 ROPs
• Compute: 9.754 TFLOPS FP32 and 19.51 TFLOPS FP16 versus 1.946 TFLOPS FP32 with no FP16 figure recorded
• Fill rates: 121.9 GPixel/s and 304.8 GTexel/s versus 30.40 GPixel/s and 60.80 GTexel/s
• Power and form factor: 225 W TDP, dual-slot, 1x 6-pin plus 1x 8-pin connectors, 550 W suggested PSU versus an MXM module with no connectors, no recorded TDP, and no suggested PSU
• Interface and outputs: PCIe 4.0 x16 with one HDMI 2.0b and three DisplayPort 1.4a versus MXM-B (3.0) with portable-device-dependent outputs
• Dimensions: 272 mm long, 111 mm high, 36 mm wide versus null (module format)
• Release: 2019 for the RX 5700 XT with a 399 USD launch MSRP; 2015 for the GTX 965M with no MSRP recorded

Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, which is a point of API-level parity despite the hardware gap. Neither has RT cores or tensor cores.

Head-to-Head Benchmarks

Only two tests appear in both GPUs' recorded results, and they point in opposite directions.

First, Geekbench OpenCL: the GTX 965M scored 14509 while the RX 5700 XT scored 9524. That is a 34.4 percent win for the NVIDIA card, and it is the more curious result of the pair given the AMD card's massive hardware advantage. The database offers no driver or runtime explanation, but the recorded fact stands: in this OpenCL run, the mobile Maxwell module outscored the Navi desktop card.

Second, Geekbench Vulkan: the RX 5700 XT scored 68310 against 14299. That 377.7 percent margin is the dominant result of the comparison and aligns with everything the hardware specs suggest, from the 2.5x shading unit count to the more than fivefold memory bandwidth advantage.

Context from each card's full benchmark sheet sharpens the picture. The RX 5700 XT also recorded 74647 in Geekbench Metal, 2134 in 3DMark Steel Nomad DX12, 16277 in PassMark G3D, 922 in PassMark G2D, 7650 in PassMark GPU Compute, and DirectX ratings of 226 (DX9), 116 (DX11), 97 (DX10), and 64 (DX12) in the PassMark suite. The GTX 965M has only its two Geekbench entries recorded, so cross-test coverage strongly favors the depth of data available on the AMD card.

The result is a one-to-one split in wins, but with wildly asymmetric margins: NVIDIA's win is measured in tens of percent, AMD's in hundreds.

Where Each One Wins

RX 5700 XT territory. Vulkan workloads are the clearest case, with the 377.7 percent head-to-head margin leading the evidence. More broadly, any workload that scales with memory capacity or bandwidth (8 GB and 448.0 GB/s versus 2 GB and 80.19 GB/s) falls to the AMD card, and its texture rate of 304.8 GTexel/s versus 60.80 GTexel/s marks out high-resolution rendering as its home ground. Against its own peer set, the Radeon Pro 5600M, RTX PRO 6000 Blackwell, Radeon PRO W7500, and RTX 5090 D V2, it holds its average score within a single percentage point, indicating a competitive desktop-class profile overall.

GTX 965M territory. OpenCL compute is the recorded bright spot, thanks to the 34.4 percent win at 14509 versus 9524. Its practical role, though, is defined by form factor: as an MXM-B module with no power connectors and device-dependent outputs, it fits into laptops where a 225 W dual-slot desktop card is categorically unsuitable. Within its own rival cluster, it sits essentially even with the Radeon RX Vega 11, GTX TITAN, Radeon Vega 11, and Iris Xe MAX.

The final ledger: the RX 5700 XT wins on hardware and Vulkan; the GTX 965M wins the OpenCL measurement and the mobile form factor it exclusively occupies. Choose the AMD card for desktop graphics throughput; choose the NVIDIA module only where OpenCL performance in a portable chassis is the deciding requirement.