AMD Radeon 820M vs AMD Radeon 890M Comparison

The Verdict

The recorded data splits these two integrated GPUs cleanly by capability tier. The AMD Radeon 890M is the substantially stronger part, built around an 8x larger shader array and delivering compute and pixel throughput that the AMD Radeon 820M cannot approach. The 820M, with its 128 shading units, 8 texture units, and 4 ROPs, is positioned for lightweight graphics and basic display output. The 890M, with 1024 shading units, 64 texture units, and 32 ROPs, is the part to select when the workload involves modern 3D rendering, higher resolution output, or compute acceleration.

Benchmark evidence places the 890M at the 45th percentile among all GPUs in the database, with an average score of 9210. Its nearest rivals include the AMD Radeon Vega 8, which scores 9221 and trails by only 0.1%, and the NVIDIA GeForce GTX 960, which scores 9273 and leads by 0.7%. The 890M sits effectively at parity with those discrete and older integrated parts. The 820M has no recorded benchmark scores and no nearest rivals in the database, so its relative standing cannot be quantified from the measurements available. For any buyer choosing between these two, the data points unambiguously to the 890M as the higher-performance option, while the 820M serves as the lower-cost, lower-throughput fallback for machines where graphics demands are minimal.

Where Each One Wins

The 890M wins across every measurable category where both parts have data. Its FP32 throughput of 5.939 TFLOPS versus 716.8 GFLOPS for the 820M represents an 8.3x advantage in raw shader compute. Texture fill rate follows the same pattern: 185.6 GTexel/s versus 22.40 GTexel/s, an 8.3x gap. Pixel throughput favors the 890M by a similar margin, 92.80 GPixel/s versus 11.20 GPixel/s, again an 8.3x difference. These ratios track the shading unit count exactly, since both parts run at similar clock speeds, with the 820M boosting to 2800 MHz and the 890M to 2900 MHz.

For ray tracing workloads, the 890M includes 16 ray tracing cores versus 2 on the 820M. That gives the larger part an 8x advantage in dedicated RT hardware, which translates to substantially better performance in DirectX 12 Ultimate titles that use ray-traced effects. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, so feature compatibility is identical; only the execution resources differ.

The 820M wins in one narrow respect: power draw. Both parts carry a 15 W TDP, so the 820M delivers its lower throughput within the same thermal envelope, meaning it is more efficient per watt for basic tasks. The 820M also has a later release date, 2025-02-28 versus 2024-07-14 for the 890M, which may indicate a newer platform integration. Neither part has a launch MSRP in the database, so no price comparison is possible.

Architecture Differences

Both GPUs use the RDNA 3.5 architecture and belong to the Navi III IGP generation for Strix Point Mobile. The manufacturing process is identical: 4 nm at TSMC. The 890M uses the Strix Point chip, while the 820M uses the Krackan Point 2 chip. Both are integrated graphics processors with a PCIe 4.0 x8 bus interface, no power connectors, and display outputs that depend on the portable device they are installed in.

The transistor counts differ significantly. The 890M packs 34,000 million transistors on a 233 mm² die, yielding a transistor density of 145.9M per mm². The 820M's transistor count and die size are listed as unknown in the database, so no direct comparison is possible for those figures. The 890M's larger die accommodates its 8x expansion in execution resources: 1024 shading units, 64 TMUs, and 32 ROPs, versus 128 shading units, 8 TMUs, and 4 ROPs on the 820M. Ray tracing cores scale identically, 16 versus 2.

Clock behavior is similar but not identical. Both have a 400 MHz base clock. The 890M boosts to 2900 MHz, 100 MHz higher than the 820M's 2800 MHz boost. This 3.6% clock advantage slightly amplifies the 890M's resource advantage in real workloads. Memory is system-shared on both parts, with bus width and bandwidth dependent on the host system's memory configuration. FP16 throughput matches FP32 on both parts at a 1:1 ratio, indicating no dedicated FP16 acceleration path.

The 890M's predecessor is listed as Navi II IGP, and the 820M shares the same predecessor. The 890M's production status is Active, and the 820M is also Active. Both support the same API set, so software compatibility is identical. The only architectural difference that would affect a user is the sheer resource multiplier on the 890M, plus its larger physical die and transistor budget.

FAQ

Q: Which GPU has more shading units?

A: The AMD Radeon 890M has 1024 shading units, while the AMD Radeon 820M has 128. That is an 8x difference in shader count.

Q: Do both GPUs support the same graphics APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The API feature set is identical.

Q: What is the boost clock difference between the two?

A: The 890M boosts to 2900 MHz, while the 820M boosts to 2800 MHz. The 890M runs 100 MHz faster at peak.

Q: How many ray tracing cores does each GPU have?

A: The 890M has 16 ray tracing cores, and the 820M has 2. The 890M has 8x the dedicated RT hardware.

Q: Which GPU has a higher average benchmark score?

A: The 890M has an average benchmark score of 9210 and sits at the 45th percentile among all GPUs. The 820M has no recorded benchmark scores in the database.

Q: Are both GPUs built on the same manufacturing process?

A: Yes, both use a 4 nm process at TSMC and both are based on the RDNA 3.5 architecture.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries comparing these two parts, so the analysis relies on the 890M's recorded scores and the 820M's absence of measurements. The 890M's benchmark results are substantial across multiple test suites. In 3DMark Steel Nomad DX12, it scores 590. Geekbench OpenCL returns 37254, and Geekbench Vulkan returns 40808. Passmark results include 8076 for G3D, 4157 for GPU compute, 979 for G2D, 97 for DirectX 9, 73 for DirectX 11, 37 for DirectX 12, and 33 for DirectX 10.

These numbers place the 890M in a competitive position relative to its nearest rivals. The AMD Radeon Vega 8 scores 9221, a 0.1% lead over the 890M's 9210. The NVIDIA GeForce GTX 960 scores 9273, a 0.7% lead. The NVIDIA GeForce GTX 465 scores 9294, a 0.9% lead. The NVIDIA GeForce GTX 850M scores 9302, a 1% lead. The 890M trails all four nearest rivals by less than one full percentage point on average, which puts it squarely in the same performance class as those older discrete GPUs.

The 820M has no benchmark scores, no percentile ranking beyond the 50th percentile default, and no nearest rivals. Its theoretical peak rates provide the only quantitative basis for comparison. The FP32 throughput of 716.8 GFLOPS is roughly one-eighth of the 890M's 5.939 TFLOPS. Texture rate of 22.40 GTexel/s is one-eighth of 185.6 GTexel/s. Pixel rate of 11.20 GPixel/s is one-eighth of 92.80 GPixel/s. These ratios confirm that the 820M scales exactly with its 8x smaller shader array, and the 890M's slightly higher boost clock adds a small additional edge.

In real-world terms, the 890M's Passmark G3D score of 8076 indicates it can handle modern 3D applications at playable settings, while its Geekbench Vulkan score of 40808 shows strong compute throughput for GPU-accelerated workloads. The 820M, lacking any recorded scores, cannot be positioned on the same curve. Its 128 shading units and 4 ROPs suggest it is suited to 2D desktop composition, video decode, and very light 3D tasks, but the database provides no empirical confirmation of that expectation.

The wins tally in the head-to-head section is 0 for the 820M and 0 for the 890M, because no direct comparative benchmarks were recorded. All performance conclusions derive from the 890M's absolute scores and the 820M's theoretical throughput figures. The 890M's average score of 9210 against a 45th percentile ranking means it beats roughly half of all GPUs in the database, while the 820M's 50th percentile is a default value with no supporting measurements. For a user prioritizing graphics capability, the 890M is the only defensible choice from the recorded data. For a user prioritizing minimal power draw, both parts share the same 15 W TDP, so the 820M offers no power advantage, only lower performance.