AMD Radeon HD 8790M vs AMD Radeon R7 M465 Comparison

The AMD Radeon R7 M465 and AMD Radeon HD 8790M are both end-of-life mobile GPUs from AMD, separated by three generations of architecture and design philosophy. While they share the same manufacturer, core count, and memory type, the data shows they are distinctly different performers, with the newer R7 M465 holding a clear edge in raw compute while the older HD 8790M counters with superior memory bandwidth. This analysis breaks down the benchmark data, architectural shifts, and specification differences to determine which GPU is better suited for specific workloads.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test, and it decisively favors the newer part. The AMD Radeon R7 M465 scores 5,841 points, while the AMD Radeon HD 8790M scores 5,017 points. This represents a 16.4% delta in favor of the R7 M465, a substantial margin that underscores the generational leap in compute efficiency. In terms of average benchmark score across all recorded tests, the R7 M465 also leads with 5,841 versus the HD 8790M's 5,691, though the HD 8790M has an additional Vulkan score of 6,365 that is not present in the R7 M465's data.

Contextualizing these scores against their nearest rivals reveals how each GPU sits in the broader mobile landscape. The R7 M465’s OpenCL score of 5,841 places it within 0.3% of the AMD Radeon R5 M435 (5,859) and 1.5% behind the Intel UHD Graphics 730 (5,929). It is 1.9% ahead of the NVIDIA GeForce GTX 550 Ti (5,731) in this test, which is notable because that NVIDIA part is a desktop-class GPU from a previous era. The HD 8790M, meanwhile, finds itself in a tighter cluster: its average score of 5,691 is 0.4% behind the Intel Iris Pro Graphics P6300 (5,712), 0.5% behind the NVIDIA GeForce GTX 670MX (5,721), and 0.7% behind the same GTX 550 Ti (5,731). The HD 8790M is 1.6% ahead of the NVIDIA Quadro M500M (5,604).

The 16.4% OpenCL delta between the two AMD parts is the headline figure. It suggests that for compute-heavy tasks such as OpenCL acceleration, the R7 M465 is the clear choice. However, the HD 8790M’s Vulkan score of 6,365 indicates it is capable in that API, though no comparable Vulkan result exists for the R7 M465 to make a direct comparison. Both GPUs share the same 33rd percentile ranking among all GPUs, meaning they occupy a similar tier of overall performance, but the R7 M465 does so with a higher peak compute score.

FAQ

Q: Which GPU has a higher raw compute performance in OpenCL?

A: The AMD Radeon R7 M465 is 16.4% faster in Geekbench OpenCL, scoring 5,841 versus the HD 8790M's 5,017. This is the only head-to-head benchmark recorded, and the R7 M465 wins it outright.

Q: Does the HD 8790M have any performance advantage in any benchmark?

A: The HD 8790M has a recorded Geekbench Vulkan score of 6,365, which is higher than its OpenCL score. However, the R7 M465 has no Vulkan benchmark recorded, so a direct comparison in that API is not possible from the data.

Q: How do these GPUs compare to their nearest rivals?

A: The R7 M465 is 1.9% ahead of the NVIDIA GeForce GTX 550 Ti and 1.5% behind the Intel UHD Graphics 730 in average score. The HD 8790M is 0.7% behind the GTX 550 Ti and 0.5% behind the NVIDIA GeForce GTX 670MX, while being 1.6% ahead of the NVIDIA Quadro M500M.

Q: Which GPU has a higher memory bandwidth?

A: The HD 8790M has a 128-bit memory bus providing 64.00 GB/s bandwidth, which is nearly double the R7 M465's 64-bit bus and 36.00 GB/s bandwidth. This is a significant advantage for the older part in bandwidth-sensitive scenarios.

Q: Are these GPUs from the same architectural generation?

A: No. The R7 M465 uses GCN 3.0 architecture, while the HD 8790M uses the older GCN 1.0 architecture. This difference explains the R7 M465's higher compute efficiency despite lower clock speeds.

Q: What is the transistor density difference between the two chips?

A: The R7 M465's Topaz chip has a transistor density of 12.4M / mm², while the HD 8790M's Mars chip has a density of 12.3M / mm². They are nearly identical in this metric, despite the R7 M465 having 1,550 million transistors versus 950 million.

Architecture Differences

The fundamental architectural split is between GCN 3.0 on the R7 M465 and GCN 1.0 on the HD 8790M. This generational gap explains why the newer GPU delivers 16.4% higher OpenCL performance while operating at lower clock speeds. The R7 M465’s Topaz chip is built on 28 nm at TSMC, same as the HD 8790M’s Mars chip, but packs significantly more transistors: 1,550 million versus 950 million. This translates to a die size of 125 mm² for the Topaz versus 77 mm² for the Mars, meaning the R7 M465 has more than 60% more silicon area to work with.

The transistor density is nearly identical at 12.4M / mm² for the R7 M465 and 12.3M / mm² for the HD 8790M, indicating that the density gains are minimal between the two generations. Instead, the performance uplift comes from architectural improvements in GCN 3.0 that allow better instruction-level parallelism and compute unit utilization. The R7 M465 also supports FP16 at a 1:1 ratio with FP32, both rated at 786.4 GFLOPS, while the HD 8790M has no listed FP16 capability. This makes the R7 M465 more suitable for workloads leveraging half-precision arithmetic.

The R7 M465 supports DirectX 12 (12_0), whereas the HD 8790M only supports DirectX 12 (11_1). Both support OpenGL 4.6 and Vulkan 1.2.170, so API compatibility is otherwise equal. The R7 M465 uses a PCIe 3.0 x8 interface, while the HD 8790M uses an MXM-A (3.0) connector, reflecting their different intended deployment scenarios: the former is a more integrated solution, while the latter is a modular MXM card.

Specification Differences

The most glaring specification gap is memory bandwidth. The HD 8790M has a 128-bit bus width yielding 64.00 GB/s, while the R7 M465 has a 64-bit bus yielding only 36.00 GB/s. This is a 78% advantage for the HD 8790M in raw bandwidth, which could matter in texture-heavy or framebuffer-intensive workloads. Both have 2 GB of GDDR5 memory, but the R7 M465 runs at 1125 MHz (4.5 Gbps effective) while the HD 8790M runs at 1000 MHz (4 Gbps effective) per pin, so the R7 M465’s higher clock partially compensates for its narrower bus.

Core configuration is identical: both have 384 shading units, 24 texture mapping units, and 8 ROPs. However, clock speeds differ. The R7 M465 has a base clock of 730 MHz and a boost clock of 1024 MHz, while the HD 8790M has a base clock of 850 MHz and a boost clock of 900 MHz. The R7 M465’s boost clock is 124 MHz higher, which contributes to its higher pixel rate of 8.192 GPixel/s versus 7.200 GPixel/s, and texture rate of 24.58 GTexel/s versus 21.60 GTexel/s.

The R7 M465 reaches 786.4 GFLOPS in FP32, while the HD 8790M reaches 691.2 GFLOPS, a difference of 95.2 GFLOPS. The HD 8790M is specified as an MXM Module with no power connectors, while the R7 M465 has no slot width or power connector data listed. The HD 8790M’s display outputs are marked as "Portable Device Dependent," while the R7 M465’s are not listed. The release dates are also distinct: the HD 8790M launched in 2013, while the R7 M465 launched in 2016, and the R7 M465 is the successor to the HD 8790M’s generation.

The Verdict

The AMD Radeon R7 M465 is the superior compute performer, with a 16.4% lead in OpenCL and higher peak clock speeds. For users running OpenCL-accelerated applications, scientific workloads, or any task that leverages FP32 or FP16 compute, the R7 M465 is the clear choice. Its GCN 3.0 architecture and higher boost clock of 1024 MHz deliver 786.4 GFLOPS, outpacing the HD 8790M’s 691.2 GFLOPS. The R7 M465 also offers better future-proofing with DirectX 12 (12_0) support and explicit FP16 throughput.

The AMD Radeon HD 8790M, despite being older, holds one significant advantage: memory bandwidth. Its 128-bit bus and 64.00 GB/s bandwidth are nearly double the R7 M465’s 36.00 GB/s. In scenarios where texture streaming or large framebuffer reads dominate, such as high-resolution gaming with heavy textures, the HD 8790M could potentially offset its compute deficit. However, no benchmark data confirms this advantage in practice, and its lower pixel rate of 7.200 GPixel/s and texture rate of 21.60 GTexel/s suggest it is slower at filling pixels.

The data indicates that the R7 M465 is the more balanced and capable GPU for modern workloads. Its 33rd percentile ranking matches the HD 8790M, but it achieves that rank with a higher average score of 5,841 versus 5,691. The HD 8790M’s Vulkan score of 6,365 is intriguing, but without a comparable R7 M465 Vulkan result, it cannot be used to justify choosing the older part. For compute-driven tasks, the R7 M465 wins unequivocally. For bandwidth-sensitive tasks, the HD 8790M has a theoretical edge, but the lack of corroborating benchmark data makes it a speculative choice. Ultimately, the R7 M465’s architectural maturity and higher compute throughput make it the recommended GPU for users prioritizing raw performance in OpenCL and modern API support.