AMD Radeon R9 M395X vs NVIDIA GeForce RTX 2080 SUPER Comparison

FAQ

Q: How do the two GPUs compare in the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce RTX 2080 SUPER scores 99,226, while the AMD Radeon R9 M395X scores 17,829. The RTX 2080 SUPER wins this test by 82%, as indicated by the -82% delta percentage for the AMD part.

Q: What is the average benchmark score for each GPU?

A: The AMD Radeon R9 M395X has an average benchmark score of 25,891, while the NVIDIA GeForce RTX 2080 SUPER averages 24,170. Despite this, they sit in similar percentile positions: 71st for AMD and 69th for NVIDIA.

Q: Which GPU has the higher memory bandwidth?

A: The NVIDIA GeForce RTX 2080 SUPER has a memory bandwidth of 495.9 GB/s, which is over three times higher than the AMD Radeon R9 M395X's 160.0 GB/s. Both use an 8 GB frame buffer, but the RTX 2080 SUPER uses GDDR6 memory while the AMD part uses GDDR5.

Q: What are the transistor counts and process nodes for these chips?

A: The AMD Radeon R9 M395X uses a 28 nm process with 5,000 million transistors on a 366 mm² die. The NVIDIA GeForce RTX 2080 SUPER uses a 12 nm process with 13,600 million transistors on a 545 mm² die, giving it a transistor density of 25.0M / mm² versus 13.7M / mm² for AMD.

Q: Which GPU supports ray tracing and tensor cores?

A: The NVIDIA GeForce RTX 2080 SUPER includes 48 RT cores and 384 tensor cores. The AMD Radeon R9 M395X has no RT cores or tensor cores listed in its specifications.

Q: What is the launch MSRP of the NVIDIA GeForce RTX 2080 SUPER?

A: The launch MSRP is 699 USD. No launch MSRP is listed for the AMD Radeon R9 M395X.

Architecture Differences

The AMD Radeon R9 M395X is built on GCN 3.0 architecture with the Amethyst chip, manufactured on a 28 nm process at TSMC. It belongs to the Gem System (R9 M300) generation. The NVIDIA GeForce RTX 2080 SUPER uses the Turing architecture with the TU104 chip, built on a 12 nm process, also at TSMC, and belongs to the GeForce 20-series. This process difference is significant: the NVIDIA chip packs 13,600 million transistors into a 545 mm² die, while the AMD chip contains 5,000 million transistors on a 366 mm² die. The resulting transistor density is 25.0M / mm² for NVIDIA versus 13.7M / mm² for AMD.

The shading unit counts differ substantially. The RTX 2080 SUPER has 3,072 shading units, 192 texture mapping units, and 64 ROPs. The R9 M395X has 2,048 shading units, 128 TMUs, and 32 ROPs. The Turing GPU also adds dedicated hardware not present on the AMD mobile part: 48 RT cores for ray tracing and 384 tensor cores for AI-accelerated workloads. The AMD chip has no such specialized units.

Memory architecture diverges as well. Both cards have 8 GB of memory on a 256-bit bus, but the RTX 2080 SUPER uses GDDR6 at 1937 MHz (15.5 Gbps effective) achieving 495.9 GB/s of bandwidth, while the R9 M395X uses GDDR5 at 1250 MHz (5 Gbps effective) for 160.0 GB/s. The NVIDIA card operates at a base clock of 1650 MHz with a boost of 1815 MHz; the AMD card has no base or boost clock listed, only memory clock data.

API support differs. The RTX 2080 SUPER supports DirectX 12 Ultimate (12_2), while the R9 M395X supports DirectX 12 (12_0). Both support OpenGL 4.6, but Vulkan versions differ: 1.4 for NVIDIA versus 1.2.170 for AMD. The NVIDIA card also lists specific display outputs (1x HDMI 2.0, 3x DisplayPort 1.4a, 1x USB Type-C), while AMD's output configuration is described as portable device dependent.

Power and physical specifications also diverge. The R9 M395X has a TDP of 75 W and uses an MXM module slot with no power connectors. The RTX 2080 SUPER has a 250 W TDP, is dual-slot, requires 1x 6-pin + 1x 8-pin power connectors, and suggests a 600 W PSU. The NVIDIA card measures 267 mm in length, 116 mm in height, and 35 mm in width.

Head-to-Head Benchmarks

The only directly comparable benchmark between the two GPUs in the data is Geekbench OpenCL. In this test, the NVIDIA GeForce RTX 2080 SUPER scores 99,226 against the AMD Radeon R9 M395X's 17,829. This represents a delta percentage of -82% for the AMD part, meaning the NVIDIA GPU is roughly 4.5 times faster in raw OpenCL compute performance. The head-to-head tally shows NVIDIA winning this single test, with zero wins for AMD.

Beyond this direct comparison, each GPU has its own benchmark footprint. The AMD Radeon R9 M395X records a Geekbench Metal score of 33,953 and a Geekbench OpenCL score of 17,829, yielding an average benchmark score of 25,891. The NVIDIA GeForce RTX 2080 SUPER has a broader test suite: 3DMark Steel Nomad DX12 at 1,882, Geekbench OpenCL at 99,226, Geekbench Vulkan at 111,284, Passmark DirectX 10 at 140, Passmark DirectX 11 at 165, Passmark DirectX 12 at 75, Passmark DirectX 9 at 227, Passmark G2D at 920, Passmark G3D at 19,490, and Passmark GPU Compute at 8,290. Its average benchmark score is 24,170.

Looking at percentile placement, the AMD part sits at the 71st percentile versus all GPUs, slightly above NVIDIA's 69th percentile. The nearest rivals for the AMD card are the AMD FirePro W7100 (25,856 average score, 0.1% delta), AMD FirePro D700 (25,842, 0.2% delta), AMD Radeon RX 6750 XT (26,011, -0.5% delta), and NVIDIA GeForce RTX 3080 Ti Mobile (25,740, 0.6% delta). The nearest rivals for the NVIDIA card include the NVIDIA GeForce GTX 780 Ti (24,236, -0.3% delta), NVIDIA GeForce GTX 1630 (24,277, -0.4% delta), AMD Radeon RX 6800S (24,063, 0.4% delta), and AMD Radeon RX 6600 XT (24,442, -1.1% delta).

These rival placements are telling. The AMD R9 M395X's average score places it near modern mid-range GPUs like the RX 6750 XT and the RTX 3080 Ti Mobile. The NVIDIA RTX 2080 SUPER's average score, despite being lower in absolute terms, sits near older high-end cards like the GTX 780 Ti and the GTX 1630. The discrepancy between the NVIDIA card's dominant OpenCL win and its lower average score suggests that its benchmark performance is uneven across different test types.

The Verdict

The data presents a clear compute performance hierarchy. In the single shared benchmark, the NVIDIA GeForce RTX 2080 SUPER outperforms the AMD Radeon R9 M395X by 82%, a decisive margin. The NVIDIA card's architecture advantages—more shading units, higher clock speeds, greater memory bandwidth, and dedicated RT and tensor cores—align with this result. However, the average benchmark scores tell a different story: the AMD part averages 25,891 versus 24,170 for NVIDIA, and the AMD card holds a higher percentile ranking (71 versus 69). This suggests that while NVIDIA dominates in OpenCL compute, the AMD card performs competitively across its own benchmark set, potentially reflecting differences in driver optimization or workload characteristics.

For users relying on OpenCL-heavy applications, the RTX 2080 SUPER is the clear choice based on the 82% advantage. For those considering overall benchmark consistency, the AMD card's higher average score and percentile position indicate it holds its own in its tested workloads. The RTX 2080 SUPER also offers modern features like ray tracing support through its 48 RT cores and tensor core acceleration, which the AMD part entirely lacks.

The production status of both GPUs is end-of-life, so neither represents a current purchase recommendation. The RTX 2080 SUPER was released later (2019-07-22 versus 2015-05-04) and succeeded the GeForce 10 series, while the AMD card succeeded the Solar System generation and was followed by Polaris Mobile. The NVIDIA card's launch MSRP was 699 USD.

Specification Differences

The two GPUs differ across nearly every measured specification. The process node is 28 nm for AMD versus 12 nm for NVIDIA. Transistor count is 5,000 million versus 13,600 million, with die sizes of 366 mm² and 545 mm² respectively. Transistor density is 13.7M / mm² for AMD and 25.0M / mm² for NVIDIA.

Clock speeds: the AMD card has no base or boost clock listed, while the RTX 2080 SUPER has a 1650 MHz base and 1815 MHz boost. Memory clocks are 1250 MHz (5 Gbps effective) for AMD and 1937 MHz (15.5 Gbps effective) for NVIDIA. Memory type is GDDR5 for AMD and GDDR6 for NVIDIA, though capacity (8 GB) and bus width (256 bit) match. Bandwidth is 160.0 GB/s versus 495.9 GB/s.

Compute resources: 2,048 shading units, 128 TMUs, and 32 ROPs for AMD; 3,072 shading units, 192 TMUs, and 64 ROPs for NVIDIA. The NVIDIA card adds 48 RT cores and 384 tensor cores; AMD has none. Pixel rate is 23.14 GPixel/s for AMD versus 116.2 GPixel/s for NVIDIA. Texture rate is 92.54 GTexel/s versus 348.5 GTexel/s. FP32 performance is 2.961 TFLOPS for AMD versus 11.15 TFLOPS for NVIDIA. FP16 is 2.961 TFLOPS (1:1) for AMD versus 22.30 TFLOPS (2:1) for NVIDIA.

Power: 75 W TDP for AMD versus 250 W for NVIDIA. The AMD card is an MXM module with no power connectors; NVIDIA is dual-slot with 1x 6-pin + 1x 8-pin connectors and suggests a 600 W PSU. The NVIDIA card has specific dimensions (267 mm length, 116 mm height, 35 mm width); AMD lists none. Display outputs are portable device dependent for AMD versus 1x HDMI 2.0, 3x DisplayPort 1.4a, and 1x USB Type-C for NVIDIA. DirectX support is 12 (12_0) versus 12 Ultimate (12_2). Vulkan support is 1.2.170 versus 1.4. Both share OpenGL 4.6 and PCIe 3.0 x16 interfaces.

Where Each One Wins

The NVIDIA GeForce RTX 2080 SUPER wins decisively in raw compute throughput. Its 82% lead in Geekbench OpenCL, combined with substantially higher pixel rate, texture rate, FP32, and FP16 figures, makes it the superior choice for compute-intensive workloads. The 495.9 GB/s memory bandwidth is over three times higher than the AMD card's 160.0 GB/s, favoring bandwidth-sensitive tasks. The presence of 48 RT cores and 384 tensor cores opens up ray-traced rendering and AI-accelerated features that are entirely unavailable on the AMD part. The NVIDIA card also supports DirectX 12 Ultimate and Vulkan 1.4, enabling the latest graphics API features.

The AMD Radeon R9 M395X wins in efficiency metrics and certain benchmark aggregations. Its 75 W TDP is one-third of the NVIDIA card's 250 W, making it suitable for power-constrained mobile implementations. Its average benchmark score of 25,891 exceeds NVIDIA's 24,170, and its 71st percentile ranking is above NVIDIA's 69th. In its nearest rival comparisons, the AMD card sits within 0.5% of the AMD Radeon RX 6750 XT, a much newer part, while the NVIDIA card's nearest rivals include older cards like the GTX 780 Ti. The AMD card's Geekbench Metal score of 33,953 provides a data point for Apple ecosystem compatibility, though no direct NVIDIA Metal comparison exists in the shared benchmarks.

For portable device deployments, the AMD card's MXM form factor and lack of external power connectors make it architecturally suited for laptops and compact systems. The NVIDIA card's dual-slot design, 267 mm length, and 600 W PSU recommendation target desktop towers. The choice between these two end-of-life GPUs depends on whether the priority is maximum compute performance with modern features (NVIDIA) or lower power consumption with competitive average benchmark scores (AMD).