AMD Radeon HD 8790M vs AMD Radeon R7 M340 Comparison

The AMD Radeon HD 8790M and AMD Radeon R7 M340 are two mobile graphics processors separated by roughly two years of development. Despite the newer name and generation tag, the benchmark data tells a story of regression in several key areas. The recorded measurements show the older chip posting higher scores in both available tests, while the newer part trails by a significant margin in one of them. This page examines the head-to-head results, the architectural differences that explain them, and what the data implies for each GPU.

Head-to-Head Benchmarks

The database contains two direct comparison results for these GPUs, both favoring the AMD Radeon HD 8790M. In the Geekbench OpenCL test, the HD 8790M scores 5017 against 4932 for the R7 M340. That is a 1.7% advantage for the older chip, a modest but consistent lead that suggests the two parts operate in a similar performance tier for compute workloads. The R7 M340’s result places it within 1.7% of its rival, and the margin is small enough that real-world differences could vary depending on driver state or thermal conditions.

The Vulkan test tells a very different story. Here the HD 8790M scores 6365, while the R7 M340 manages only 5193. That is a 22.6% advantage for the HD 8790M, a substantial gap that cannot be explained by minor clock variations. The R7 M340 loses nearly a quarter of its rival’s performance in this API, which is notable because both GPUs support the same Vulkan version (1.2.170). The delta suggests the newer chip’s architecture delivers considerably lower throughput under this workload, possibly due to its narrower memory interface or reduced shader count.

Looking at the average benchmark score across all recorded tests, the HD 8790M sits at 5691, while the R7 M340 averages 5063. That is an 11% overall difference in favor of the older part. The HD 8790M’s nearest rival, the Intel Iris Pro Graphics P6300, scores 5712, a 0.4% difference from the HD 8790M. The NVIDIA GeForce GTX 670MX and GTX 550 Ti sit within 0.7% of the HD 8790M, meaning the older AMD chip is effectively in a dead heat with those three GPUs. The NVIDIA Quadro M500M trails by 1.6%.

For the R7 M340, its nearest rival is the AMD Radeon R7 240, which matches its average score exactly at 5063, a 0% delta. The AMD FirePro W4170M is 0.6% behind, the AMD Radeon R5 M430 is 0.9% behind, and the integrated AMD Radeon R7 Graphics is 1.3% behind. This places the R7 M340 at the bottom of a tight cluster of low-end AMD parts, none of which come close to the HD 8790M’s position.

The data shows a clear hierarchy. The HD 8790M outscores the R7 M340 in both individual tests and on average. The Vulkan result is the standout metric, with a 22.6% gap that dwarfs the OpenCL difference. A user transitioning from the HD 8790M to the R7 M340 would likely notice a slowdown in Vulkan-based applications, while OpenCL workloads would feel nearly identical.

Architecture Differences

The two GPUs share a 28 nm TSMC manufacturing process and both use AMD’s Graphics Core Next architecture, but they belong to different generations. The HD 8790M is built on GCN 1.0, while the R7 M340 uses GCN 3.0. This generational jump does not translate into superior performance in the recorded data, which raises questions about how the newer architecture was configured.

The HD 8790M uses the Mars chip with 950 million transistors on a 77 mm² die. The R7 M340 uses the Meso chip with 1,550 million transistors on a 125 mm² die. The transistor density is nearly identical: 12.3 million per square millimeter for the HD 8790M versus 12.4 million for the R7 M340. The newer chip is physically larger and packs more transistors, yet it delivers lower benchmark scores, which suggests the additional silicon is not being used for raw compute throughput.

The compute configuration differs significantly. The HD 8790M has 384 shading units, 24 texture mapping units, and 8 ROPs. The R7 M340 has 320 shading units, 20 TMUs, and 8 ROPs. The older chip has 20% more shading units and 20% more TMUs, which explains its higher theoretical throughput in several metrics. The HD 8790M achieves 691.2 GFLOPS of FP32 compute, while the R7 M340 reaches 653.4 GFLOPS. The texture rate is 21.60 GTexel/s for the HD 8790M versus 20.42 GTexel/s for the R7 M340. The pixel rate, however, favors the newer chip: 8.168 GPixel/s versus 7.200 GPixel/s for the HD 8790M, due to the R7 M340’s higher clock speed.

Memory is where the two diverge most sharply. The HD 8790M uses 2 GB of GDDR5 on a 128 bit bus, delivering 64.00 GB/s of bandwidth. The R7 M340 also has 2 GB, but it is DDR3 on a 64 bit bus, yielding only 16.00 GB/s. That is a 75% reduction in memory bandwidth. The R7 M340’s memory clocks at 1000 MHz with 2 Gbps effective, while the HD 8790M runs at 1000 MHz with 4 Gbps effective. The narrower bus and slower memory type are likely the primary reasons the R7 M340 falls so far behind in the Vulkan test, since memory bandwidth is often a bottleneck for graphics workloads.

Clock speeds favor the newer chip. The R7 M340 runs at 943 MHz base and 1021 MHz boost, while the HD 8790M runs at 850 MHz base and 900 MHz boost. The higher clocks help the R7 M340 close some of the gap in pixel rate and texture rate, but they cannot compensate for the loss of shading units and memory bandwidth.

The API support differs slightly. The HD 8790M supports DirectX 12 (11_1), while the R7 M340 supports DirectX 12 (12_0). Both support OpenGL 4.6 and Vulkan 1.2.170. The newer DirectX feature level on the R7 M340 is a point in its favor for modern titles that use those features, though the benchmark data does not include any DirectX tests to quantify the impact.

The generation naming reinforces the intended positioning. The HD 8790M belongs to the Solar System generation (HD 8700M) and lists London as its predecessor and Gem System as its successor. The R7 M340 belongs to the Gem System generation (R7 M300), with Solar System as its predecessor and Polaris Mobile as its successor. The R7 M340 is therefore the direct successor to the HD 8700M line, yet it loses in every recorded benchmark. The data implies a regression, not an improvement.

The Verdict

The recorded data clearly favors the AMD Radeon HD 8790M. It wins both head-to-head tests, holds a higher average score, and occupies a higher percentile rank. The HD 8790M sits at the 33rd percentile among all GPUs, while the R7 M340 sits at the 30th. The older part’s 22.6% lead in the Vulkan test is the defining result of this comparison, indicating that the HD 8790M is meaningfully faster in at least one modern API.

The R7 M340 does have advantages in the specification sheet. It runs at higher clocks, supports DirectX 12 (12_0) instead of 11_1, and has a larger die with more transistors. It also offers FP16 compute at 653.4 GFLOPS with a 1:1 ratio to FP32, a feature the HD 8790M does not list. For workloads that leverage FP16, the R7 M340 could potentially perform better than its FP32 numbers suggest, but the database contains no FP16 benchmark results to confirm this.

The R7 M340’s memory configuration is its biggest liability. A 64 bit DDR3 bus with 16.00 GB/s bandwidth is a severe constraint for a discrete GPU, and the Vulkan result likely reflects this bottleneck. The HD 8790M’s 128 bit GDDR5 bus provides four times the bandwidth, which is a decisive advantage in graphics-heavy tasks.

For users choosing between these two parts based on the database alone, the HD 8790M is the stronger option. It wins in OpenCL by a small margin and in Vulkan by a large one. The R7 M340 should only be considered if its DirectX 12 (12_0) support or FP16 capability is specifically needed, since those are the only areas where it holds a clear spec advantage.

Specification Differences

The two GPUs differ in the following recorded fields:

• Chip: Mars (HD 8790M) versus Meso (R7 M340)
• Architecture: GCN 1.0 versus GCN 3.0
• Generation: Solar System (HD 8700M) versus Gem System (R7 M300)
• Transistors: 950 million versus 1,550 million
• Die size: 77 mm² versus 125 mm²
• Transistor density: 12.3M / mm² versus 12.4M / mm²
• Base clock: 850 MHz versus 943 MHz
• Boost clock: 900 MHz versus 1021 MHz
• Memory clock: 4 Gbps effective versus 2 Gbps effective
• Memory type: GDDR5 versus DDR3
• Bus width: 128 bit versus 64 bit
• Bandwidth: 64.00 GB/s versus 16.00 GB/s
• Shading units: 384 versus 320
• TMUs: 24 versus 20
• Pixel rate: 7.200 GPixel/s versus 8.168 GPixel/s
• Texture rate: 21.60 GTexel/s versus 20.42 GTexel/s
• FP32: 691.2 GFLOPS versus 653.4 GFLOPS
• FP16: not listed versus 653.4 GFLOPS (1:1)
• Bus interface: MXM-A (3.0) versus PCIe 3.0 x8
• Slot width: MXM Module versus not listed
• DirectX: 12 (11_1) versus 12 (12_0)
• Release date: 2013-03-31 versus 2015-05-04
• Predecessor: London versus Solar System
• Successor: Gem System versus Polaris Mobile

Fields that are identical include the manufacturer (AMD), process node (28 nm), foundry (TSMC), memory size (2 GB), ROP count (8), OpenGL version (4.6), Vulkan version (1.2.170), production status (End-of-life), and the absence of a launch MSRP.

FAQ

Q: Which GPU wins the Geekbench OpenCL test?

A: The AMD Radeon HD 8790M wins with a score of 5017 against 4932 for the AMD Radeon R7 M340, a 1.7% difference.

Q: How much faster is the HD 8790M in the Vulkan test?

A: The HD 8790M scores 6365 in Geekbench Vulkan, while the R7 M340 scores 5193. The HD 8790M leads by 22.6%.

Q: Why might the R7 M340 perform worse despite being newer?

A: The R7 M340 has 320 shading units versus 384, and its memory subsystem is a 64 bit DDR3 bus with 16.00 GB/s bandwidth, compared to a 128 bit GDDR5 bus with 64.00 GB/s on the HD 8790M.

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

A: The HD 8790M has an average benchmark score of 5691, while the R7 M340 averages 5063.

Q: Does the R7 M340 have any specification advantages over the HD 8790M?

A: Yes. The R7 M340 runs at higher clocks (943 MHz base, 1021 MHz boost), supports DirectX 12 (12_0) instead of 12 (11_1), and offers FP16 compute at 653.4 GFLOPS with a 1:1 ratio.

Q: How do the two GPUs rank among all GPUs?

A: The HD 8790M ranks at the 33rd percentile, while the R7 M340 ranks at the 30th percentile.

Where Each One Wins

The AMD Radeon HD 8790M wins in both available benchmark tests and in the overall average score. Its 22.6% lead in the Vulkan test makes it the clear choice for applications that use that API, which includes many modern games and compute frameworks. The 1.7% lead in OpenCL is smaller but still favors the older chip. The HD 8790M’s 64.00 GB/s of memory bandwidth and 384 shading units give it a structural advantage that shows up in compute-heavy workloads.

The R7 M340 does not win any benchmark tests, but it holds advantages in areas the database does not directly measure. Its DirectX 12 (12_0) feature level is higher than the HD 8790M’s 12 (11_1), which matters for titles that require the newer feature set. Its FP16 support at 653.4 GFLOPS could provide a benefit in workloads that use half-precision math, though no benchmark in the database verifies this. The higher clock speeds (1021 MHz boost versus 900 MHz) also give it a higher pixel rate of 8.168 GPixel/s, which could help in fill-rate-limited scenarios.

The use-case split comes down to this: for general compute and Vulkan-based graphics, the HD 8790M is the stronger performer. For users who specifically need DirectX 12 (12_0) features or FP16 compute, the R7 M340 offers capabilities the older part lacks. The recorded data contains two wins for the HD 8790M and zero for the R7 M340, so the default recommendation from the database is clear. The R7 M340 is a newer chip with some modern features, but the measurements show it cannot match its predecessor in raw performance.