AMD Radeon 890M vs AMD Radeon HD 8870M Comparison

The AMD Radeon 890M and AMD Radeon HD 8870M represent two distant generations of mobile graphics, and the benchmark data confirms a decisive shift in performance. The 890M, an integrated processor from 2024, utterly outclasses the 2013 discrete HD 8870M in the single available head-to-head metric, while also offering a far richer feature set and modern architecture. The data shows a generational leap that makes the older card largely obsolete for contemporary workloads.

Head-to-Head Benchmarks

The sole direct comparison available is the Geekbench OpenCL test, and the result is a landslide. The AMD Radeon 890M scores 37,254 points, while the AMD Radeon HD 8870M manages only 8,462 points. This translates to a 340.3% advantage for the 890M, meaning it delivers more than four times the raw compute performance of the older GPU. This is not a marginal improvement; it is a complete overhaul in processing capability, driven by a decade of architectural evolution and a vastly more efficient manufacturing process.

Looking at the broader benchmark landscape, the 890M’s average benchmark score is 9,210, placing it in the 45th percentile of all GPUs. Its closest rivals, according to the data, are the AMD Radeon Vega 8 (average score 9,221, a -0.1% difference) and the NVIDIA GeForce GTX 960 (average score 9,273, a -0.7% difference). This positions the 890M as a competent entry-level performer, competitive with dedicated graphics cards from a few generations ago. In contrast, the HD 8870M’s average score is 8,462, placing it in the 43rd percentile. Its nearest rival is the NVIDIA GeForce MX330, with an average score of 8,458 (a 0.1% difference), indicating that even in its own time, the HD 8870M was a modest performer, and its capabilities have not aged well.

The 890M’s individual benchmark results further illustrate its strengths. In Geekbench Vulkan, it scores 40,808 points, demonstrating strong modern API performance. Its Passmark G3D score is 8,076, and its Passmark G2D score is 979. The Passmark GPU Compute test yields a score of 4,157. While the HD 8870M lacks these specific benchmark entries, its single OpenCL score of 8,462 is dwarfed by the 890M’s 37,254 in the same test, reinforcing the massive performance gulf.

The 890M also shows proficiency in newer DirectX 12 workloads, with a 3DMark Steel Nomad DX12 score of 590. The HD 8870M has no equivalent modern benchmark result, but its API support for only DirectX 12 (11_1) suggests it would struggle with the advanced features required by such tests. The performance delta in the head-to-head test is so large that it effectively defines the entire comparison: the 890M is in a different performance class entirely.

FAQ

Q: How much faster is the AMD Radeon 890M than the HD 8870M in OpenCL?

A: The 890M is 340.3% faster, scoring 37,254 versus the HD 8870M’s 8,462 in the Geekbench OpenCL benchmark.

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

A: The AMD Radeon 890M has an average benchmark score of 9,210, while the AMD Radeon HD 8870M has an average score of 8,462.

Q: Which GPU has a higher percentile ranking among all GPUs?

A: The 890M ranks in the 45th percentile, while the HD 8870M ranks in the 43rd percentile, showing the 890M is slightly better positioned in the overall performance distribution.

Q: What are the closest rivals to the AMD Radeon 890M?

A: According to the data, its nearest rivals include the AMD Radeon Vega 8 (average score 9,221, -0.1% delta) and the NVIDIA GeForce GTX 960 (average score 9,273, -0.7% delta).

Q: Does the HD 8870M support modern APIs like the 890M?

A: No. The 890M supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the HD 8870M only supports DirectX 12 (11_1) and Vulkan 1.2.170.

Q: What is the production status of each card?

A: The AMD Radeon 890M is listed as "Active," whereas the AMD Radeon HD 8870M is marked as "End-of-life."

Architecture Differences

The architectural chasm between these two GPUs is the primary driver of their performance disparity. The 890M is built on the RDNA 3.5 architecture, part of the Navi III IGP (Strix Point Mobile) generation, and is fabricated on a 4 nm process by TSMC. This is a massive leap from the HD 8870M, which uses the ancient GCN 1.0 architecture and a 28 nm process. The newer node allows for significantly higher transistor density: the 890M packs 34,000 million transistors on a 233 mm² die (a density of 145.9M / mm²), while the HD 8870M has only 1,500 million transistors on a 123 mm² die (a density of 12.2M / mm²). This is a 12x difference in transistor density, explaining the 890M’s superior efficiency and capability.

The 890M features a modern unified shader design with 1024 shading units, 64 texture mapping units (TMUs), and 32 raster operations pipelines (ROPs). It also includes 16 dedicated ray tracing cores, a feature entirely absent from the HD 8870M. The older GPU has 640 shading units, 40 TMUs, and 16 ROPs. This difference in raw execution resources is substantial. The 890M’s clock speeds are also much higher, with a base of 400 MHz and a boost of 2900 MHz, compared to the HD 8870M’s base of 725 MHz and boost of 775 MHz.

These architectural and clock differences translate directly into compute throughput. The 890M delivers 5.939 TFLOPS of FP32 performance, while the HD 8870M manages only 992.0 GFLOPS. The 890M also supports FP16 compute at a 1:1 ratio, a feature the HD 8870M does not have. The pixel and texture rates tell a similar story: the 890M achieves 92.80 GPixel/s and 185.6 GTexel/s, versus the HD 8870M’s 12.40 GPixel/s and 31.00 GTexel/s. The 890M is a next-generation part designed for modern workloads, while the HD 8870M is a product of an older era with far more limited capabilities.

Specification Differences

The two GPUs differ across nearly every core specification. The 890M is an integrated graphics processor (IGP) with a 15 W TDP, while the HD 8870M is a discrete mobile chip. The 890M uses System Shared memory, meaning its size, type, and bus width are all dependent on the host system. Its bandwidth is also defined as System Dependent. In contrast, the HD 8870M has a dedicated 2 GB of GDDR5 memory on a 128-bit bus, providing a fixed bandwidth of 72.00 GB/s. The 890M’s memory clock is listed as "System Shared," whereas the HD 8870M operates at 1125 MHz (or 4.5 Gbps effective).

The interface also differs: the 890M uses PCIe 4.0 x8, while the HD 8870M uses the older PCIe 3.0 x16. The 890M has a 4 nm process node, versus the HD 8870M’s 28 nm. The 890M’s core clocks are significantly higher, with a boost of 2900 MHz versus 775 MHz for the HD 8870M. The 890M supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the HD 8870M is limited to DirectX 12 (11_1) and Vulkan 1.2.170. The 890M has 16 ray tracing cores, which the HD 8870M completely lacks. Finally, the 890M is an IGP with no power connectors, while the HD 8870M’s power connector specifications are not listed. The 890M is an active product, while the HD 8870M is end-of-life.

Where Each One Wins

The AMD Radeon 890M wins in essentially every measurable category. Its only head-to-head benchmark victory is in Geekbench OpenCL, where it is 340.3% ahead. This makes it the clear choice for any modern computing task, from general productivity to light gaming and content creation. Its support for DirectX 12 Ultimate and Vulkan 1.4 means it can handle contemporary game engines and graphics APIs, while its ray tracing cores offer a glimpse into future-proofing. The 890M’s high FP32 and FP16 throughput make it suitable for compute-heavy applications like video encoding, photo editing, and some machine learning tasks. Its high percentile ranking (45th) and competitive average score relative to dedicated GPUs like the GTX 960 confirm it is a viable option for entry-level gaming on the go.

The AMD Radeon HD 8870M has no benchmark wins in this comparison. Its only advantage lies in having a dedicated 2 GB frame buffer with a fixed 72.00 GB/s bandwidth, which can be beneficial for certain older games that require a set amount of VRAM, but this is a minor point given its overall performance deficit. Its PCIe 3.0 x16 interface is also a legacy feature. In practice, the HD 8870M would only be relevant for running very old or non-demanding software, and even then, its 992.0 GFLOPS of compute power severely limits its usefulness. The data is unambiguous: the 890M is the superior product in every meaningful performance metric, and the HD 8870M is a relic that cannot compete with modern integrated graphics.