AMD Radeon R5 M335 vs AMD Radeon R7 M365X Comparison

The AMD Radeon R7 M365X and AMD Radeon R5 M335 are two end-of-life mobile GPUs from the same GCN 1.0 era, and the recorded data makes the comparison straightforward: the R7 M365X wins both head-to-head benchmarks, posting a 25.2 percent lead in Geekbench OpenCL and a narrower 2.8 percent edge in Geekbench Vulkan. Both chips were built on TSMC's 28 nm process and connect over PCIe 3.0 x8, but they differ sharply in memory subsystem, shading resources, and die scale, and those differences explain the results.

FAQ

Q: Which GPU is faster overall?

A: The AMD Radeon R7 M365X. It holds an average benchmark score of 5416 against the R5 M335's 4752, and it wins both head-to-head tests: Geekbench OpenCL by 25.2 percent (5939 versus 4745) and Geekbench Vulkan by 2.8 percent (4893 versus 4758).

Q: How do the two compare against the wider GPU landscape?

A: The R7 M365X sits in the 32nd percentile of all GPUs in the database, while the R5 M335 lands in the 28th percentile. Both are entry-level parts positioned near the bottom third of the market.

Q: Which GPU has the better memory setup?

A: The R7 M365X, by a wide margin. It pairs 1024 MB of GDDR5 on a 128-bit bus for 64.00 GB/s of bandwidth. The R5 M335 uses 2 GB of DDR3 on a 64-bit bus for only 14.40 GB/s, less than a quarter of the bandwidth despite the larger capacity.

Q: Do the two GPUs support the same APIs?

A: Yes. Both report DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170.

Q: Are these GPUs still in production?

A: No. Both carry an end-of-life production status. The R7 M365X was released on 4 May 2015, and the R5 M335 followed on 20 October 2015.

Q: How do their theoretical rates compare?

A: The R5 M335 actually posts slightly higher theoretical rates, at 8.240 GPixel/s pixel fill, 20.60 GTexel/s texture fill, and 659.2 GFLOPS FP32, versus 6.600 GPixel/s, 19.80 GTexel/s, and 633.6 GFLOPS for the R7 M365X. Measured performance nevertheless favors the R7 M365X.

Architecture Differences

On paper the two chips share most of their architectural DNA. Both are AMD GCN 1.0 designs manufactured by TSMC on a 28 nm node, both sit in the Gem System generation of mobile parts (R7 M300 and R5 M300 lines respectively), both trace their lineage back to the Solar System predecessor family, and both were eventually succeeded by Polaris Mobile. Neither GPU includes ray tracing cores or tensor cores, consistent with their era and market position. Both expose identical modern API support for their class: DirectX 12 with feature level 11_1, OpenGL 4.6, and Vulkan 1.2.170.

The divergence is in scale and memory. The R7 M365X, internally designated Litho, is built from 950 million transistors on a 77 mm² die. The R5 M335, designated Exo, uses 690 million transistors on a 56 mm² die. Curiously, both chips hit the same recorded transistor density of 12.3M per mm², which is exactly what you would expect from two dies cut from the same 28 nm GCN family, differing mainly in how much of the design is enabled.

That scale difference shows up in the shader count. The R7 M365X carries 384 shading units and 24 texture mapping units, against 320 shading units and 20 TMUs for the R5 M335. Raster output units are tied at 8 on both parts. The memory story is far less balanced: the R7 M365X runs GDDR5 at 1000 MHz (4 Gbps effective) over a 128-bit bus, delivering 64.00 GB/s. The R5 M335 runs DDR3 at 900 MHz (1800 Mbps effective) over a 64-bit bus, delivering just 14.40 GB/s. The R5 M335 counters with double the memory capacity, 2 GB against 1024 MB, which is the only memory metric where it leads.

Head-to-Head Benchmarks

Only two head-to-head tests exist in the database, and the R7 M365X swept them both.

The OpenCL result is the headline number. In geekbench_opencl the R7 M365X scored 5939 against 4745 for the R5 M335, a 25.2 percent victory. That is a decisive gap in a segment where rivals are often separated by single-digit percentages, and it aligns with the one subsystem difference that dominates compute workloads: bandwidth. A 64.00 GB/s GDDR5 interface feeding 384 shaders gives the R7 M365X a comfortable advantage in data-parallel OpenCL tasks, while the R5 M335's 14.40 GB/s DDR3 channel leaves its 320 shaders starved.

The Vulkan result is much closer. In geekbench_vulkan the R7 M365X scored 4893 against 4758, a 2.8 percent win. Graphics-oriented workloads spread the load across the geometry, rasterization, and shading pipeline, and here the R5 M335's slightly higher theoretical rates (659.2 GFLOPS FP32 and 8.240 GPixel/s against 633.6 GFLOPS and 6.600 GPixel/s) keep it nearly level despite its weaker memory subsystem. A 2.8 percent gap is within the range where the two chips feel broadly similar in lighter graphics tasks.

Context from the rivals list reinforces where each card sits. The R7 M365X's average score of 5416 places it essentially level with the AMD Radeon 610M (5444, a 0.5 percent gap), the Intel UHD Graphics P630 (5370, 0.9 percent ahead of it), the NVIDIA Quadro M4000 (5467, 0.9 percent ahead), and the AMD Radeon R7 M445 (5358, 1.1 percent behind it). The R5 M335's 4752 average is similarly clustered with the AMD Radeon R8 M445DX (4727, 0.5 percent behind it), the AMD Radeon R5 M255 (4788, 0.7 percent ahead), the NVIDIA GeForce RTX 3080 12 GB (4791, 0.8 percent ahead), and the NVIDIA Quadro P400 (4684, 1.5 percent behind). Both cards compete in tightly packed score bands, which makes the R7 M365X's higher cluster placement the meaningful difference between them.

Specification Differences

The fields where the two GPUs actually differ:

• Chip designation: R7 M365X is Litho, R5 M335 is Exo.
• Product line: R7 M300 tier versus R5 M300 tier, both within the Gem System generation.
• Release date: 4 May 2015 for the R7 M365X, 20 October 2015 for the R5 M335.
• Transistors: 950 million versus 690 million.
• Die size: 77 mm² versus 56 mm².
• Shading units: 384 versus 320.
• TMUs: 24 versus 20.
• Memory size: 1024 MB versus 2 GB.
• Memory type: GDDR5 versus DDR3.
• Bus width: 128 bit versus 64 bit.
• Memory bandwidth: 64.00 GB/s versus 14.40 GB/s.
• Memory clock: 1000 MHz (4 Gbps effective) versus 900 MHz (1800 Mbps effective).
• Pixel rate: 6.600 GPixel/s versus 8.240 GPixel/s.
• Texture rate: 19.80 GTexel/s versus 20.60 GTexel/s.
• FP32 compute: 633.6 GFLOPS versus 659.2 GFLOPS.
• Power connectors: not specified for the R7 M365X, listed as none for the R5 M335.
• Display outputs: not specified for the R7 M365X, listed as portable-device dependent for the R5 M335.

Everything else lines up identically: GCN 1.0 architecture, TSMC 28 nm fabrication, 12.3M transistors per mm² density, 8 ROPs, PCIe 3.0 x8 interface, identical API support, end-of-life status, Solar System predecessor, and Polaris Mobile successor.

Where Each One Wins

AMD Radeon R7 M365X wins on: measured compute performance (25.2 percent OpenCL lead), overall average benchmark score (5416 versus 4752), percentile placement (32nd versus 28th), memory bandwidth (64.00 GB/s versus 14.40 GB/s), shader count (384 versus 320), texture mapping units (24 versus 20), and bus width (128 bit versus 64 bit). For any workload that streams data through the GPU, the GDDR5 subsystem is the decisive advantage, and the OpenCL result demonstrates it directly.

AMD Radeon R5 M335 wins on: memory capacity (2 GB versus 1024 MB), theoretical pixel fill (8.240 versus 6.600 GPixel/s), theoretical texture fill (20.60 versus 19.80 GTexel/s), theoretical FP32 compute (659.2 versus 633.6 GFLOPS), and near-parity in Vulkan graphics workloads (4758 versus 4893, only 2.8 percent behind). The extra gigabyte of memory is the only practical advantage, relevant when a workload's dataset simply does not fit in 1024 MB.

The pattern is clear: the R5 M335's paper rates are marginally higher, but the R7 M365X converts its superior memory subsystem and larger shader array into real benchmark wins. Theoretical throughput figures did not predict the outcomes here, bandwidth did.

The Verdict

The data supports one conclusion: choose the R7 M365X unless the R5 M335's larger memory capacity is specifically required. The R7 M365X won every head-to-head test in the database, leads the average score by a comfortable margin, sits four percentile points higher against the full GPU population, and competes in a meaningfully stronger rival cluster. Anyone running OpenCL compute workloads should strongly prefer it, given the 25.2 percent advantage on that test.

The case for the R5 M335 is narrow. It matches the R7 M365X within 2.8 percent on Vulkan graphics tests, offers double the video memory at 2 GB, and posts marginally higher theoretical fill-rate and FP32 numbers. If a specific workload needs more than 1024 MB of GPU memory and is graphics-bound rather than bandwidth-bound, the R5 M335 holds up reasonably well. For everything else in the recorded data, the R7 M365X is the stronger GPU. Both are end-of-life parts at the low end of the historical market, so this comparison matters mainly for assessing older hardware, and on that basis the R7 M365X is the clear pick.