AMD Radeon R5 M240 vs AMD Radeon R7 M350 Comparison

# AMD Radeon R5 M240 vs AMD Radeon R7 M350

The AMD Radeon R5 M240 and AMD Radeon R7 M350 are both end-of-life mobile GPUs from AMD's Gem System generation, but they represent different points in that lineup. The R5 M240 belongs to the R5 M200 family, while the R7 M350 is part of the R7 M300 family. Despite sharing a 28 nm TSMC process node, the two chips diverge significantly in architecture, compute resources, and memory configuration. Benchmark data shows the R7 M350 edges out the R5 M240 in OpenCL performance, though the margin is remarkably thin.

Head-to-Head Benchmarks

The only direct head-to-head benchmark available is Geekbench OpenCL, where the AMD Radeon R7 M350 scores 6991 against the AMD Radeon R5 M240's 6975. This gives the R7 M350 a win by a mere 0.2 percent — a statistical tie in practical terms. The R7 M350 also has a Geekbench Vulkan score of 5662, a test the R5 M240 does not have a recorded result for, indicating broader API coverage in the newer part.

Placing these scores in context, the R5 M240's 6975 OpenCL result puts it in the 39th percentile of all GPUs tracked. Its nearest rivals are a tight cluster: the NVIDIA GeForce GTX 675M sits 0.4 percent behind at 6946, the NVIDIA GeForce GTX 680M is 0.7 percent ahead at 7023, and the NVIDIA T600 is 0.8 percent ahead at 7035. The AMD FirePro M5100 trails by 2.1 percent with a score of 6830. This places the R5 M240 in a performance band where the GTX 675M and GTX 680M — both older high-end mobile parts — are essentially indistinguishable from it in OpenCL compute.

The R7 M350's average benchmark score across its two recorded tests is 6327, dragged down substantially by its Vulkan result of 5662. Its 36th percentile ranking reflects this blended average. In terms of nearest rivals for the average score, the AMD Radeon Pro WX 4100 matches it exactly at 6330 with a 0.0 percent delta, while the NVIDIA Quadro K620 is 0.7 percent behind at 6282. Interestingly, the NVIDIA GeForce RTX 5070 Ti SUPER and RTX 4070 Ti SUPER AD102 both sit at 6270, 0.9 percent behind — a reminder that these modern desktop cards score in the same range as this old mobile part in this specific compute workload.

The data indicates the R7 M350's OpenCL advantage over the R5 M240 is negligible for real-world compute tasks. However, the R7 M350's Vulkan capability, which the R5 M240 entirely lacks, represents a functional superiority that raw OpenCL scores do not capture.

Architecture Differences

The architectural gap between these two GPUs is substantial. The R5 M240 is built on GCN 1.0 architecture using the Jet chip, while the R7 M350 uses the newer GCN 3.0 architecture with the Meso chip. Both are fabricated by TSMC on a 28 nm process, but the transistor counts differ enormously: the R5 M240 packs 690 million transistors on a 56 mm² die, yielding a transistor density of 12.3 million per mm². The R7 M350 more than doubles this with 1,550 million transistors on a 125 mm² die, at a nearly identical density of 12.4 million per mm². The die size difference of 69 mm² is the most striking physical divergence.

Compute resources scale accordingly. The R7 M350 has 384 shading units, 24 texture mapping units, and 8 render output units. The R5 M240 has 320 shading units, 20 TMUs, and the same 8 ROPs. This gives the R7 M350 a 20 percent advantage in shading units and TMUs, while ROP count stays equal. The clock speeds are nearly identical: both have a 1000 MHz base clock, with the R5 M240 boosting to 1030 MHz and the R7 M350 to 1015 MHz. The R5 M240 actually has a 15 MHz higher boost clock, but the R7 M350's extra hardware more than compensates.

Memory configurations differ in capacity and speed, though both use DDR3 on a 64-bit bus. The R5 M240 has 1024 MB of memory running at 900 MHz (1800 Mbps effective), producing 14.40 GB/s of bandwidth. The R7 M350 has 4 GB of memory at 1000 MHz (2 Gbps effective), yielding 16.00 GB/s. The R7 M350 offers 4 times the memory capacity and 11 percent more bandwidth, while both are constrained by the narrow 64-bit bus.

Compute throughput numbers reflect the resource differences. The R7 M350 delivers 779.5 GFLOPS of FP32 performance, compared to 659.2 GFLOPS for the R5 M240 — an 18 percent advantage. The R7 M350 also supports FP16 at a 1:1 ratio (779.5 GFLOPS), while the R5 M240 has no recorded FP16 capability. Pixel rates are nearly identical at 8.120 GPixel/s for the R7 M350 versus 8.240 GPixel/s for the R5 M240, while texture rates favor the R7 M350 at 24.36 GTexel/s versus 20.60 GTexel/s.

API support also diverges. The R7 M350 supports DirectX 12 (12_0), while the R5 M240 is limited to DirectX 12 (11_1). Both support OpenGL 4.6 and Vulkan 1.2.170, though the R5 M240 has no Vulkan benchmark score recorded. Both use a PCIe 3.0 x8 bus interface, and neither has a recorded TDP, slot width, or power connector specification.

Where Each One Wins

The R7 M350 wins in virtually every measurable performance category. Its OpenCL score of 6991 beats the R5 M240's 6975, and it alone among the two has a Vulkan score of 5662. It offers higher FP32 throughput (779.5 GFLOPS versus 659.2 GFLOPS), higher texture fill rate (24.36 GTexel/s versus 20.60 GTexel/s), and more memory capacity (4 GB versus 1024 MB). For any compute workload that scales with shading units or memory capacity, the R7 M350 is the clear choice.

The R5 M240's advantages are narrow but real. Its boost clock of 1030 MHz is 15 MHz higher than the R7 M350's 1015 MHz, and its pixel rate of 8.240 GPixel/s slightly exceeds the R7 M350's 8.120 GPixel/s. For pixel-bound operations that are ROP-limited, the R5 M240 has a marginal edge. Additionally, its smaller die size (56 mm² versus 125 mm²) and lower transistor count (690 million versus 1,550 million) suggest lower power consumption and thermal load, though no TDP figures are recorded to confirm this.

In terms of benchmark percentile, the R5 M240 actually ranks higher at the 39th percentile compared to the R7 M350's 36th percentile. This is a consequence of the R7 M350's Vulkan score being averaged into its overall ranking, pulling it down. For OpenCL-only users, the two GPUs are effectively equivalent, with the R7 M350's 0.2 percent lead being within run-to-run variance.

For gaming, neither part is well-suited by modern standards, but the R7 M350's 4 GB memory buffer is a significant practical advantage for titles that exceed 1 GB of video memory. The R5 M240's 1024 MB limit will cause texture thrashing or reduced detail settings in such games. For compute applications that use FP16, the R7 M350's 1:1 FP16 support is a distinct advantage, though its relevance depends on the software's ability to use half-precision math.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The AMD Radeon R7 M350 scores 6991 in Geekbench OpenCL, compared to 6975 for the AMD Radeon R5 M240. The R7 M350 wins by 0.2 percent, a margin within typical run-to-run variation.

Q: Do both GPUs support the same DirectX version?

A: No. The R7 M350 supports DirectX 12 (12_0), while the R5 M240 is limited to DirectX 12 (11_1). Both support OpenGL 4.6 and Vulkan 1.2.170.

Q: How much memory does each GPU have?

A: The R5 M240 has 1024 MB of DDR3 memory, while the R7 M350 has 4 GB of DDR3 memory. Both use a 64-bit memory bus, with bandwidth of 14.40 GB/s for the R5 M240 and 16.00 GB/s for the R7 M350.

Q: What are the transistor and die size differences?

A: The R5 M240 uses the Jet chip with 690 million transistors on a 56 mm² die. The R7 M350 uses the Meso chip with 1,550 million transistors on a 125 mm² die. Both are fabricated by TSMC on a 28 nm process with nearly identical transistor densities around 12.3–12.4 million per mm².

Q: Which GPU has more shading units?

A: The R7 M350 has 384 shading units, compared to 320 for the R5 M240. The R7 M350 also has 24 texture mapping units versus 20, while both have 8 render output units.

Q: Does the R5 M240 have any Vulkan benchmark result?

A: No. Only the R7 M350 has a recorded Geekbench Vulkan score of 5662. The R5 M240 has no Vulkan benchmark data in the record, despite listing Vulkan 1.2.170 API support.

Specification Differences

The two GPUs differ across nearly every hardware specification except for process node, foundry, ROP count, bus interface, and API support for OpenGL and Vulkan. Both use a 28 nm TSMC process, have 8 render output units, and connect via PCIe 3.0 x8. The R5 M240 is built on GCN 1.0 architecture with the Jet chip, while the R7 M350 uses GCN 3.0 with the Meso chip. The R5 M240 comes from the R5 M200 generation, the R7 M350 from the R7 M300 generation, and their release dates differ by roughly eight months: the R5 M240 launched on September 17, 2014, and the R7 M350 on May 4, 2015.

Transistor count scales from 690 million to 1,550 million, and die size from 56 mm² to 125 mm². The R7 M350 has 384 shading units and 24 TMUs to the R5 M240's 320 and 20, respectively. Clock speeds are close but not identical: both have a 1000 MHz base, but the R5 M240 boosts to 1030 MHz versus 1015 MHz for the R7 M350. Memory differs substantially: 1024 MB versus 4 GB, 900 MHz versus 1000 MHz, and 14.40 GB/s versus 16.00 GB/s bandwidth. Compute rates favor the R7 M350 at 779.5 GFLOPS FP32 versus 659.2 GFLOPS, and it adds FP16 at 779.5 GFLOPS while the R5 M240 has none. Pixel rates slightly favor the R5 M240 at 8.240 GPixel/s versus 8.120 GPixel/s, while texture rates favor the R7 M350 at 24.36 GTexel/s versus 20.60 GTexel/s. DirectX support differs as noted, with the R7 M350 reaching 12_0 and the R5 M240 capped at 11_1. Benchmark results give the R7 M350 one recorded Vulkan score of 5662, bringing its average to 6327, while the R5 M240 has only the OpenCL score of 6975 and a higher 39th percentile versus the R7 M350's 36th.