AMD Radeon 890M vs AMD Radeon R9 M275X Comparison

The AMD Radeon R9 M275X and AMD Radeon 890M represent two very different eras of graphics hardware. The R9 M275X is an end-of-life mobile discrete GPU from early 2014, built on the GCN 1.0 architecture with a 28 nm process. The 890M is an active integrated graphics processor from mid-2024, using the much newer RDNA 3.5 architecture on a 4 nm node. The benchmark data in the database shows a decisive performance gap, but the details matter for understanding where each part fits.

Head-to-Head Benchmarks

The database includes only two direct head-to-head benchmark comparisons between these two GPUs, and the 890M wins both by a substantial margin. In the Geekbench OpenCL test, the R9 M275X scores 11,200 points, while the 890M scores 37,254 points. This is a delta of -69.9% from the perspective of the R9 M275X, meaning the 890M is roughly 3.3 times faster in raw compute throughput. The Vulkan results are even more lopsided. The R9 M275X records 9,964 points, while the 890M reaches 40,808 points, a delta of -75.6%. That translates to the 890M delivering about 4.1 times the performance in Vulkan workloads.

Looking at the broader benchmark averages, the R9 M275X has an average score of 10,582 across its recorded tests, while the 890M averages 9,210 across a much wider set of tests. This is an important nuance: the average is dragged down for the 890M because it includes several Passmark tests where the scores are quite low relative to its Geekbench results. For instance, the 890M scores only 33 in Passmark DirectX 10, 73 in DirectX 11, and 37 in DirectX 12, while its Passmark G3D score is 8,076 and its GPU compute score is 4,157. The R9 M275X has no such low-end outliers in its record, so its average is more consistent with its two strong Geekbench scores.

In terms of percentile ranking, the R9 M275X sits at the 49th percentile among all GPUs in the database, while the 890M sits at the 45th percentile. Despite the 890M winning every head-to-head test, its percentile is slightly lower because the database includes a broader set of workloads that expose the limitations of an integrated part. The R9 M275X has zero wins in the head-to-head comparison, while the 890M has two. This is a clear sweep, but the percentile data reminds us that raw compute wins do not always translate to a higher overall ranking when many other workloads are considered.

Where Each One Wins

The 890M is the clear winner for any workload that relies on modern graphics APIs or heavy compute. Its Geekbench Vulkan score of 40,808 is more than four times the R9 M275X's 9,964, which indicates a massive advantage in Vulkan-based games and applications. The OpenCL result of 37,254 versus 11,200 shows a similar advantage for compute tasks like GPU-accelerated rendering, machine learning inference, or physics simulations. The 890M also supports DirectX 12 Ultimate (12_2), while the R9 M275X only reaches DirectX 12 (11_1). This means the 890M can handle the latest DirectX 12 features, including ray tracing and mesh shaders, whereas the R9 M275X is limited to older feature levels.

The R9 M275X does not win any recorded benchmark, so there is no workload category in which the data shows it outperforming the 890M. However, the R9 M275X has a higher average benchmark score (10,582 versus 9,210) and a higher percentile rank (49 versus 45). This suggests that in the specific set of tests it was run through, it delivered more consistent performance across a variety of legacy workloads. Its 2 GB dedicated GDDR5 memory with a 128-bit bus and 72.00 GB/s bandwidth means it does not share system memory, which can be an advantage in scenarios where the integrated GPU must compete with the CPU for memory bandwidth. The 890M's memory bandwidth is listed as "System Dependent," so its actual throughput varies with the host system's RAM configuration.

For gaming at older DirectX 9 or 10 titles, the 890M's Passmark scores are low (97 in DirectX 9, 33 in DirectX 10), which might indicate that legacy API performance is not its strength. The R9 M275X has no recorded Passmark scores, so we cannot directly compare those specific tests, but its GCN 1.0 architecture was designed for the DirectX 11 era. If a user is running very old games that rely heavily on fixed-function pipelines or older shader models, the R9 M275X might be more consistent, though the data does not prove this.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The AMD Radeon R9 M275X has an average benchmark score of 10,582, while the AMD Radeon 890M averages 9,210. Despite this, the 890M wins every direct head-to-head test.

Q: What is the performance difference in Geekbench OpenCL?

A: The 890M scores 37,254 points versus the R9 M275X's 11,200 points. That is a delta of -69.9% relative to the R9 M275X, meaning the 890M is approximately 3.3 times faster.

Q: Does the 890M support ray tracing?

A: Yes, the 890M has 16 ray tracing cores and supports DirectX 12 Ultimate (12_2). The R9 M275X does not list any ray tracing cores and only supports DirectX 12 (11_1).

Q: What are the memory configurations of these two GPUs?

A: The R9 M275X has 2 GB of dedicated GDDR5 memory with a 128-bit bus and 72.00 GB/s bandwidth. The 890M uses system shared memory, with its type, bus width, and bandwidth all listed as "System Shared" or "System Dependent."

Q: Which GPU has a higher boost clock?

A: The 890M has a boost clock of 2900 MHz, while the R9 M275X boosts to 925 MHz. The 890M's base clock is 400 MHz, which is lower than the R9 M275X's base of 900 MHz, but the boost ceiling is much higher.

Q: Are these GPUs in the same performance percentile?

A: No. The R9 M275X is at the 49th percentile among all GPUs, while the 890M is at the 45th percentile. The 890M's percentile is lower despite its higher raw scores, due to its weaker results in some Passmark tests.

Specification Differences

The key specification differences between the two parts are substantial. The R9 M275X uses a 28 nm process node, while the 890M uses a 4 nm node, both from TSMC. Transistor count jumps from 1,500 million in the R9 M275X to 34,000 million in the 890M, and die size grows from 123 mm² to 233 mm². Transistor density improves from 12.2 million per mm² to 145.9 million per mm².

Clock speeds differ significantly. The R9 M275X has a base clock of 900 MHz and a boost of 925 MHz, while the 890M has a base of 400 MHz and a boost of 2,900 MHz. The memory situation is entirely different: the R9 M275X has 2 GB of GDDR5 on a 128-bit bus with 72.00 GB/s bandwidth, while the 890M uses system shared memory with no fixed bandwidth figure.

Compute resources also diverge. The R9 M275X has 640 shading units, 40 texture mapping units (TMUs), and 16 raster operation units (ROPs). The 890M has 1,024 shading units, 64 TMUs, and 32 ROPs. The 890M also includes 16 ray tracing cores, which the R9 M275X lacks entirely. Pixel rate jumps from 14.80 GPixel/s on the R9 M275X to 92.80 GPixel/s on the 890M, and texture rate jumps from 37.00 GTexel/s to 185.6 GTexel/s. Floating point performance (FP32) goes from 1,184.0 GFLOPS on the R9 M275X to 5.939 TFLOPS on the 890M, and the 890M also lists FP16 at 5.939 TFLOPS, while the R9 M275X has no FP16 figure.

The bus interface differs as well: the R9 M275X uses PCIe 3.0 x16, while the 890M uses PCIe 4.0 x8. Power draw is not listed for the R9 M275X, but the 890M has a TDP of 15 W and is described as an IGP with no power connectors. The R9 M275X has no listed TDP, slot width, or power connector details. The 890M's display outputs are "Portable Device Dependent," while the R9 M275X has no display output information in the database.

Architecture Differences

The architectural gap is fundamental. The R9 M275X is built on GCN 1.0, AMD's Graphics Core Next architecture from the early 2010s. Its chip is codenamed Venus, and it belongs to the "Gem System (R9 M200)" generation. The 890M uses RDNA 3.5, AMD's much newer architecture, with a chip codenamed Strix Point, and belongs to the "Navi III IGP (Strix Point Mobile)" generation. The process node difference is massive: 28 nm versus 4 nm, which explains the transistor density jump from 12.2 million per mm² to 145.9 million per mm².

The 890M's architecture brings features that the R9 M275X simply cannot support. It has 16 dedicated ray tracing cores, enabling hardware-accelerated ray tracing in games and applications that support it. The R9 M275X has no ray tracing hardware. The 890M also supports DirectX 12 Ultimate (12_2), which includes features like variable rate shading and mesh shaders, while the R9 M275X is limited to DirectX 12 (11_1). Vulkan support differs as well: the R9 M275X supports Vulkan 1.2.170, while the 890M supports Vulkan 1.4. OpenGL support is identical at 4.6 for both.

The 890M's integrated nature means it shares system memory with the CPU, which can be a bottleneck depending on the host platform's memory speed and bandwidth. The R9 M275X, being a discrete GPU, has its own dedicated GDDR5 memory that does not compete with the CPU for bandwidth. The 890M also has a much lower TDP of 15 W, which makes it suitable for thin-and-light portable devices, whereas the R9 M275X has no listed TDP, suggesting it was designed for larger laptops with more cooling headroom. The 890M's predecessor is listed as "Navi II IGP," and it has no successor, while the R9 M275X's predecessor is "Solar System" and its successor is "Polaris Mobile." The production status confirms this: the R9 M275X is end-of-life, while the 890M is active.

The Verdict

The data is unequivocal: the AMD Radeon 890M outperforms the AMD Radeon R9 M275X in every direct benchmark comparison. Its Geekbench OpenCL and Vulkan scores are roughly 3.3 and 4.1 times higher, respectively. Anyone building or buying a system today should choose the 890M if they need modern graphics API support, ray tracing, or high compute throughput. The 890M's 5.939 TFLOPS of FP32 performance and 16 ray tracing cores make it viable for contemporary gaming at reasonable settings, especially in an integrated part that consumes only 15 W.

The R9 M275X is a relic from 2014. It has no ray tracing support, limited DirectX 12 feature levels, and a fraction of the compute capability. Its only advantages in the recorded data are a slightly higher average benchmark score (10,582 versus 9,210) and a higher percentile rank (49 versus 45). This is likely due to the 890M's poor showing in legacy Passmark tests, which pull down its average. If a user is running exclusively old DirectX 9 or 10 applications, the R9 M275X might offer more consistent frame pacing, but the 890M still has a higher peak performance in modern APIs.

For a new system, the 890M is the obvious choice. It is active, supports the latest standards, and offers dramatically higher raw performance. For a legacy system that already has an R9 M275X, there is no data-driven reason to upgrade to an 890M unless the user needs ray tracing or Vulkan 1.4 support, since the R9 M275X already handles OpenGL 4.6 and DirectX 12 (11_1) adequately. However, the 890M's much higher boost clock, doubled shading units, and quadrupled pixel rate mean that any workload that scales with these resources will see massive gains. The verdict is simple: the 890M is the better GPU by every measured benchmark, and the R9 M275X should only be considered if legacy API consistency matters more than raw performance.