AMD Radeon 890M vs NVIDIA GeForce GTX 650 Ti Comparison

The Verdict

The database comparison between the AMD Radeon 890M and the NVIDIA GeForce GTX 650 Ti is decisively one-sided in raw compute and modern API performance. The 890M leads in both recorded head-to-head benchmarks, with a 372.9% advantage in Geekbench OpenCL and a 395.9% advantage in Geekbench Vulkan. The 890M also holds a higher average benchmark score of 9210 against 8053 for the GTX 650 Ti, and it sits at the 45th percentile among all GPUs compared to the 42nd percentile for the older NVIDIA part.

The GTX 650 Ti is an end-of-life product from 2012, built on the Kepler architecture, and its only wins in this comparison are structural: it has a dedicated 1024 MB GDDR5 frame buffer with 86.40 GB/s bandwidth, whereas the 890M relies on system shared memory with bandwidth that is system dependent. For any user considering a modern integrated solution versus a legacy discrete card, the data points clearly toward the 890M for compute-heavy tasks, modern API support, and overall benchmark standing. The GTX 650 Ti retains relevance only for scenarios where a fixed local memory pool and a specific driver environment from that era are required, but the recorded measurements show no performance scenario where it outpaces the 890M.

Architecture Differences

The two GPUs come from radically different design eras. The AMD Radeon 890M is built on the RDNA 3.5 architecture, manufactured on a 4 nm process at TSMC, with 34,000 million transistors on a 233 mm² die. The NVIDIA GeForce GTX 650 Ti uses the Kepler architecture, also manufactured by TSMC but on a 28 nm process, with 2,540 million transistors on a 221 mm² die. The transistor density difference is stark: the 890M packs 145.9 million transistors per mm², while the GTX 650 Ti manages only 11.5 million per mm². That density gap explains how the 890M can deliver far higher throughput from a similar die area.

The 890M is an integrated graphics processor (IGP) within the Strix Point mobile chip, using PCIe 4.0 x8, while the GTX 650 Ti is a discrete single-slot card with a 145 mm length, PCIe 3.0 x16, and a 1x 6-pin power connector. The 890M has no power connectors and a 15 W TDP, while the GTX 650 Ti consumes 110 W and requires a 300 W suggested PSU. The 890M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, whereas the GTX 650 Ti supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. That difference in DirectX feature level is significant: the 890M can execute 12_2 workloads, while the GTX 650 Ti is capped at 11_0.

The 890M features 1024 shading units, 64 texture mapping units, 32 ROPs, and 16 ray tracing cores. The GTX 650 Ti has 768 shading units, 64 TMUs, and 16 ROPs, with no ray tracing cores. The 890M also has a boost clock of 2900 MHz and a base clock of 400 MHz, while the GTX 650 Ti has no recorded base or boost clocks in the database. Memory architecture differs fundamentally: the 890M uses system shared memory with no dedicated bus width, while the GTX 650 Ti uses 1024 MB of GDDR5 on a 128 bit bus with 86.40 GB/s bandwidth. The 890M's memory clock is listed as system shared, and its bandwidth is system dependent.

Where Each One Wins

Based on the recorded data, the AMD Radeon 890M wins every benchmark where both parts have scores. In Geekbench OpenCL, the 890M scores 37254 against 7877 for the GTX 650 Ti, a 372.9% lead. In Geekbench Vulkan, the 890M scores 40808 against 8229, a 395.9% lead. The 890M also has a higher average benchmark score of 9210 versus 8053, and it ranks in the 45th percentile of all GPUs, three percentile points above the GTX 650 Ti at 42nd.

The 890M's wins extend across its full benchmark suite, which includes 3DMark Steel Nomad DX12 at 590, Passmark DirectX 10 at 33, DirectX 11 at 73, DirectX 12 at 37, DirectX 9 at 97, G2D at 979, G3D at 8076, and GPU compute at 4157. The GTX 650 Ti has no recorded scores in any of those tests, so no direct comparison is possible for those workloads, but the two available cross-benchmarks show a dominant pattern. The 890M also supports ray tracing cores, which the GTX 650 Ti lacks, and it supports DirectX 12 Ultimate, while the GTX 650 Ti only reaches DirectX 12 (11_0). For any modern workload using Vulkan, OpenCL, or ray tracing, the 890M is the only sensible choice from this pair.

The GTX 650 Ti's only structural advantage is its dedicated memory. It has 1024 MB of GDDR5 with a fixed 128 bit bus and 86.40 GB/s bandwidth. The 890M's memory performance is system dependent, meaning its effective bandwidth varies with the host platform's shared memory configuration. In a scenario where a user needs a fixed local frame buffer with predictable bandwidth and is constrained to the GTX 650 Ti's driver era, that part remains usable. But the recorded data shows no benchmark win for the GTX 650 Ti, so its advantage is purely situational and not performance-based in the measured tests.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The AMD Radeon 890M has an average benchmark score of 9210, while the NVIDIA GeForce GTX 650 Ti has an average benchmark score of 8053.

Q: How large is the 890M's lead in Geekbench Vulkan?

A: The 890M scores 40808 in Geekbench Vulkan, compared to 8229 for the GTX 650 Ti, a delta of 395.9%.

Q: Does the GTX 650 Ti support ray tracing?

A: No. The GTX 650 Ti has no ray tracing cores, while the 890M has 16 ray tracing cores.

Q: What is the memory configuration difference?

A: The 890M uses system shared memory with system dependent bandwidth, while the GTX 650 Ti has 1024 MB of GDDR5 on a 128 bit bus with 86.40 GB/s bandwidth.

Q: Which GPU supports a newer DirectX version?

A: The 890M supports DirectX 12 Ultimate (12_2), while the GTX 650 Ti supports DirectX 12 (11_0).

Q: What are the production statuses of these GPUs?

A: The 890M is listed as Active, with a release date of 2024-07-14, while the GTX 650 Ti is End-of-life, with a release date of 2012-10-08.

Head-to-Head Benchmarks

The database contains two direct comparisons between these GPUs, and both are landslide wins for the AMD Radeon 890M. In Geekbench OpenCL, the 890M posts a score of 37254 against 7877 for the GTX 650 Ti. That is a 372.9% delta, meaning the 890M delivers nearly five times the OpenCL compute performance. The gap is even larger in Geekbench Vulkan, where the 890M scores 40808 against 8229, a 395.9% delta. The Vulkan result is particularly telling because it reflects modern API efficiency, where the 890M's RDNA 3.5 architecture and 16 ray tracing cores can fully engage, while the GTX 650 Ti's Kepler design predates Vulkan's widespread adoption and lacks the hardware features to compete.

The 890M's broader benchmark suite provides additional context. Its Passmark G3D score is 8076, and its GPU compute score is 4157. Its 3DMark Steel Nomad DX12 score is 590, which is a modern ray-traced workload that the GTX 650 Ti cannot run due to its DirectX 12 (11_0) feature level. The 890M also achieves a pixel rate of 92.80 GPixel/s and a texture rate of 185.6 GTexel/s, compared to 14.85 GPixel/s and 59.39 GTexel/s for the GTX 650 Ti. Those rate figures align with the benchmark deltas: the 890M is roughly six times faster in pixel throughput and over three times faster in texture throughput.

The GTX 650 Ti's nearest rivals in the database are the GTX 880M (0.2% higher average score), GTX 650 Ti Boost (0.2% lower), Quadro P5000 (0.2% higher), and GRID K2 (0.3% lower). This tight cluster around 8053 suggests the GTX 650 Ti sits at the expected performance level for its era. The 890M's nearest rivals are the AMD Radeon Vega 8 (0.1% lower), GTX 960 (0.7% lower), GTX 465 (0.9% lower), and GTX 850M (1% lower), all clustered around 9210. Interesting, the 890M's closest rival is another integrated GPU, the Vega 8, which is only 0.1% behind in average score. This indicates the 890M is at the top of the integrated GPU class but still below the mid-range discrete cards of the past generation, though it towers over the GTX 650 Ti.

Specification Differences

The two GPUs differ in nearly every specification category. The 890M uses a 4 nm process with 34,000 million transistors, while the GTX 650 Ti uses a 28 nm process with 2,540 million transistors. Die size is similar at 233 mm² versus 221 mm², but transistor density is 145.9M per mm² versus 11.5M per mm². The 890M has 1024 shading units, 64 TMUs, 32 ROPs, and 16 ray tracing cores. The GTX 650 Ti has 768 shading units, 64 TMUs, 16 ROPs, and no ray tracing cores. The 890M's boost clock is 2900 MHz with a base of 400 MHz, while the GTX 650 Ti has no recorded base or boost clocks. The 890M's memory is system shared, while the GTX 650 Ti has 1024 MB of GDDR5 with a 128 bit bus and 86.40 GB/s bandwidth.

Power requirements differ sharply: the 890M is a 15 W IGP with no power connectors, while the GTX 650 Ti is a 110 W single-slot card with a 1x 6-pin connector and a 300 W suggested PSU. The 890M uses PCIe 4.0 x8, while the GTX 650 Ti uses PCIe 3.0 x16. Display outputs are portable device dependent for the 890M, while the GTX 650 Ti has 2x DVI and 1x mini-HDMI 1.4a. The 890M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The GTX 650 Ti supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The 890M has a release date of 2024-07-14 and is Active, while the GTX 650 Ti has a release date of 2012-10-08 and is End-of-life, with a launch MSRP of 149 USD. The 890M's predecessor is Navi II IGP, while the GTX 650 Ti's predecessor is GeForce 500 and its successor is GeForce 700.