AMD Radeon R9 M375X vs NVIDIA GeForce GTX 1070 Comparison

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 1070 is in a completely different performance class than the AMD Radeon R9 M375X. In the two benchmark tests where both cards were recorded, the GTX 1070 wins both, with deltas of 440.3% in Geekbench OpenCL and 164.1% in Geekbench Vulkan. The average benchmark score tells the same story: the GTX 1070 sits at 9780, while the R9 M375X averages 8325. That is a 17.5% gap in the aggregate, but the head-to-head tests reveal the true scale of the difference in compute workloads.

For users choosing between these two, the verdict is straightforward. The GTX 1070 is the pick for anyone running OpenCL or Vulkan applications, where it delivers 5.4x and 2.6x the raw score respectively. The R9 M375X is not competitive in those metrics. However, the R9 M375X holds a higher percentile rank in one sense: it sits at the 43rd percentile of all GPUs, while the GTX 1070 sits at the 47th. That is a narrow gap, but it reflects the fact that the GTX 1070's nearest rivals are workstation and compute cards, not gaming parts. The R9 M375X's rivals include the GTX 675MX and the MX330, which are older or lower-tier mobile parts. So the R9 M375X is closer to its own peers, while the GTX 1070 is closer to much more capable hardware.

The practical conclusion: if the workload is OpenCL or Vulkan, the GTX 1070 is the only rational choice. If the workload is something lighter, the R9 M375X may suffice, but the benchmark data offers no evidence that it wins any test against the GTX 1070. The GTX 1070 wins 2 out of 2 recorded head-to-head tests, and the R9 M375X wins 0.

Architecture Differences

The two GPUs come from different architectural generations and manufacturing processes. The GTX 1070 uses the GP104 chip on NVIDIA's Pascal architecture, built on a 16 nm process at TSMC. It packs 7,200 million transistors into a 314 mm² die, giving a transistor density of 22.9 million per mm². The R9 M375X uses the Tropo chip on AMD's GCN 1.0 architecture, also built at TSMC but on a 28 nm process. It contains 1,500 million transistors on a 123 mm² die, with a density of 12.2 million per mm².

Those numbers explain a lot. The GTX 1070 has 4.8x the transistor count and a 2.6x larger die, but it is also built on a more advanced node. The Pascal architecture is newer, with a release date in June 2016 versus May 2015 for the R9 M375X. The GTX 1070's predecessor is the GeForce 900 series, and its successor is the GeForce 20 series. The R9 M375X's predecessor is the Solar System series, and its successor is Polaris Mobile.

The compute resources differ dramatically. The GTX 1070 has 1920 shading units, 120 texture mapping units, and 64 raster operation pipelines. The R9 M375X has 640 shading units, 40 TMUs, and 16 ROPs. That is a 3x advantage in shading units, a 3x advantage in TMUs, and a 4x advantage in ROPs for the GTX 1070. Pixel rate is 107.7 GPixel/s versus 16.24 GPixel/s, a 6.6x gap. Texture rate is 202.0 GTexel/s versus 40.60 GTexel/s, a 5x gap. FP32 performance is 6.463 TFLOPS versus 1,299.2 GFLOPS, a 5x gap. The GTX 1070 also has FP16 capability at 101.0 GFLOPS, while the R9 M375X has no recorded FP16 figure.

Both cards support DirectX 12, but the GTX 1070 supports feature level 12_1 while the R9 M375X supports 11_1. Both support OpenGL 4.6. Vulkan support differs: the GTX 1070 lists Vulkan 1.4, while the R9 M375X lists Vulkan 1.2.170. The GTX 1070 is a dual-slot card with a 150 W TDP and a 1x 8-pin power connector, requiring a 450 W PSU. The R9 M375X has no recorded TDP, slot width, power connector, or PSU requirement in the database. The GTX 1070 is 267 mm long, 112 mm high, and 40 mm wide. The R9 M375X has no recorded dimensions.

Head-to-Head Benchmarks

The database records two head-to-head tests: Geekbench OpenCL and Geekbench Vulkan. Both are wins for the GTX 1070.

In Geekbench OpenCL, the GTX 1070 scores 44700 against the R9 M375X's 8273. That is a delta of 440.3%, meaning the GTX 1070 is roughly 5.4x faster in raw OpenCL compute. This is the largest gap in any recorded metric between the two cards. The R9 M375X's OpenCL score is closer to its own Vulkan score than to the GTX 1070's OpenCL score, which shows how far behind it is in this workload.

In Geekbench Vulkan, the GTX 1070 scores 22121 against 8377, a delta of 164.1%. The GTX 1070 is 2.6x faster here. The R9 M375X's Vulkan score is actually slightly higher than its OpenCL score (8377 versus 8273), suggesting it handles Vulkan relatively better than OpenCL, but the GTX 1070 still dominates.

The aggregate benchmark scores confirm the trend. The GTX 1070 averages 9780 across all recorded tests, while the R9 M375X averages 8325. The GTX 1070's individual test scores include 3DMark Steel Nomad DX12 at 1082, Geekbench Metal at 18801, Passmark DirectX 10 at 82, DirectX 11 at 100, DirectX 12 at 48, DirectX 9 at 197, G2D at 846, G3D at 13498, and GPU Compute at 6102. The R9 M375X has only two recorded benchmarks: Geekbench OpenCL at 8273 and Vulkan at 8377.

Specification Differences

The two cards differ in nearly every specification recorded. Here are the fields where they diverge:

• Process node: 16 nm (GTX 1070) versus 28 nm (R9 M375X)
• Transistors: 7,200 million versus 1,500 million
• Die size: 314 mm² versus 123 mm²
• Transistor density: 22.9M / mm² versus 12.2M / mm²
• Base clock: 1506 MHz versus 925 MHz
• Boost clock: 1683 MHz versus 1015 MHz
• Memory clock: 2002 MHz (8 Gbps effective) versus 1125 MHz (4.5 Gbps effective)
• Memory size: 8 GB versus 2 GB
• Memory type: GDDR5 on both, but bus width is 256 bit versus 128 bit
• Memory bandwidth: 256.3 GB/s versus 72.00 GB/s
• Shading units: 1920 versus 640
• TMUs: 120 versus 40
• ROPs: 64 versus 16
• Pixel rate: 107.7 GPixel/s versus 16.24 GPixel/s
• Texture rate: 202.0 GTexel/s versus 40.60 GTexel/s
• FP32: 6.463 TFLOPS versus 1,299.2 GFLOPS
• FP16: 101.0 GFLOPS (1:64) versus not recorded
• TDP: 150 W versus not recorded
• Slot width: Dual-slot versus not recorded
• Power connectors: 1x 8-pin versus not recorded
• Suggested PSU: 450 W versus not recorded
• Display outputs: 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.4a versus not recorded
• DirectX support: 12 (12_1) versus 12 (11_1)
• Vulkan support: 1.4 versus 1.2.170
• OpenGL support: 4.6 on both
• Release date: 2016-06-09 versus 2015-05-04
• Launch MSRP: 379 USD versus not recorded

The GTX 1070 also has recorded dimensions (267 mm x 112 mm x 40 mm), while the R9 M375X has none. Both use PCIe 3.0 x16. Both are end-of-life products.

FAQ

Q: Which GPU wins in Geekbench OpenCL?

A: The NVIDIA GeForce GTX 1070 wins with a score of 44700 against the AMD Radeon R9 M375X's 8273, a delta of 440.3%.

Q: Which GPU wins in Geekbench Vulkan?

A: The GTX 1070 wins with 22121 against 8377, a delta of 164.1%.

Q: How do their average benchmark scores compare?

A: The GTX 1070 has an average benchmark score of 9780, while the R9 M375X averages 8325. The GTX 1070 is higher by roughly 17.5%.

Q: What are the nearest rivals to each card?

A: The GTX 1070's nearest rivals are the AMD FirePro W5000 (-0.2%), NVIDIA Quadro M2000M (-0.5%), NVIDIA Tesla M10 (+0.6%), and NVIDIA Tesla C2070 (+0.7%). The R9 M375X's nearest rivals are the NVIDIA Quadro K1200 (+0.7%), NVIDIA GeForce GTX 675MX (-1.2%), AMD Radeon 880M (-1.3%), and NVIDIA GeForce MX330 (-1.6%).

Q: What is the memory bandwidth difference?

A: The GTX 1070 has 256.3 GB/s of bandwidth, while the R9 M375X has 72.00 GB/s. The GTX 1070 also has 8 GB of GDDR5 memory on a 256-bit bus, versus 2 GB on a 128-bit bus.

Q: Which card has more shading units?

A: The GTX 1070 has 1920 shading units, while the R9 M375X has 640. The GTX 1070 also has 120 TMUs and 64 ROPs, versus 40 TMUs and 16 ROPs on the R9 M375X.

Where Each One Wins

The GTX 1070 wins every recorded benchmark where both cards have data. In OpenCL, it is 440.3% ahead. In Vulkan, it is 164.1% ahead. The GTX 1070 also holds the overall average score advantage at 9780 versus 8325. Its wins come from every major hardware category: more shaders, more texture units, more ROPs, higher clocks, more memory, wider bus, and higher bandwidth. The GTX 1070 is also the only one of the two with a recorded TDP, PSU requirement, display outputs, and physical dimensions, making it a fully specified desktop part.

The R9 M375X has no recorded wins in the head-to-head tests. Its only advantage is in percentile ranking relative to its own rivals: it sits at the 43rd percentile of all GPUs, while the GTX 1070 sits at the 47th percentile. That is a small edge in overall standing, but it does not translate into any benchmark victory. The R9 M375X's nearest rivals are lower-tier mobile or older parts like the GTX 675MX and MX330, while the GTX 1070's rivals are workstation cards like the Quadro M2000M and Tesla M10. The R9 M375X is also smaller in die size and transistor count, and it has no recorded FP16 capability, no recorded TDP, and no recorded launch MSRP.

For use cases, the GTX 1070 is the choice for any OpenCL or Vulkan compute workload, given its massive lead in both. It is also the card with the higher memory capacity (8 GB versus 2 GB), which matters for workloads that need to hold larger datasets. The R9 M375X is not competitive in any recorded metric against the GTX 1070, but it may still be relevant for systems that require a lower-power, smaller footprint solution, though the database does not record its TDP or dimensions to confirm that. The GTX 1070's 150 W TDP and dual-slot design suggest it is a full-size desktop card, while the R9 M375X's lack of recorded dimensions and power data suggests it may be a mobile or OEM part.

In summary, the GTX 1070 wins every head-to-head test, every compute metric, and every memory specification. The R9 M375X's only recorded edge is its percentile standing relative to its own peers, which is a statistical artifact of being compared to weaker hardware. The data shows no scenario where the R9 M375X outperforms the GTX 1070.