AMD Radeon 890M vs NVIDIA B200 SXM6 Comparison

Head-to-Head Benchmarks

The recorded data contains no direct head-to-head benchmark comparisons between the AMD Radeon 890M and the NVIDIA B200 SXM6. The head-to-head benchmark array is empty, and neither part has wins recorded against the other in shared tests. This is expected given their entirely different market positions: the Radeon 890M is an integrated graphics processor for mobile systems, while the B200 SXM6 is a server accelerator module.

The AMD Radeon 890M has an extensive benchmark record across ten tests. Its strongest results include a Geekbench Vulkan score of 40,808 and a Geekbench OpenCL score of 37,254. In PassMark testing, the 890M scores 8,076 in G3D, 979 in G2D, and 4,157 in GPU compute. DirectX-specific PassMark results are 97 for DirectX 9, 73 for DirectX 11, 37 for DirectX 12, and 33 for DirectX 10. The 3DMark Steel Nomad DX12 test yields a score of 590. The average benchmark score across all recorded tests is 9,210.

The NVIDIA B200 SXM6 has no benchmark scores recorded in the database. Its average benchmark score is listed as zero, and its nearest rivals array is empty. The percentile versus all GPUs field shows 50, which places it at the median position in the database ranking, but without actual score data this cannot be interpreted as a measured performance level. The 890M sits at the 45th percentile versus all GPUs, with its nearest rivals being the AMD Radeon Vega 8 at 9,221 average score (0.1% ahead), the NVIDIA GeForce GTX 960 at 9,273 (0.7% ahead), the NVIDIA GeForce GTX 465 at 9,294 (0.9% ahead), and the NVIDIA GeForce GTX 850M at 9,302 (1% ahead). These delta values indicate the 890M is statistically clustered with those four parts, all within one percent of each other.

FAQ

Q: Which processor has the higher FP32 compute throughput?

A: The NVIDIA B200 SXM6 delivers 69.34 TFLOPS of FP32 performance, compared to 5.939 TFLOPS for the AMD Radeon 890M. The B200 SXM6 is approximately 11.7 times higher in this metric.

Q: What memory configurations do the two processors use?

A: The AMD Radeon 890M uses system shared memory, with its bus width, type, and bandwidth all listed as system dependent. The NVIDIA B200 SXM6 has 180 GB of HBM3e memory on an 8192-bit bus, providing 8.19 TB/s of bandwidth and running at 2000 MHz with 8 Gbps effective speed.

Q: Which processor has more shading units and texture mapping units?

A: The NVIDIA B200 SXM6 has 18,944 shading units and 592 TMUs. The AMD Radeon 890M has 1,024 shading units and 64 TMUs. The B200 SXM6 has approximately 18.5 times more shading units and 9.25 times more TMUs.

Q: What are the power requirements for each processor?

A: The AMD Radeon 890M has a TDP of 15 W and requires no power connectors. The NVIDIA B200 SXM6 has a TDP of 1000 W and its suggested PSU rating is 1400 W.

Q: What API support does each processor offer?

A: The AMD Radeon 890M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The NVIDIA B200 SXM6 lists DirectX, OpenGL, and Vulkan all as N/A.

Q: How do the transistor counts and die sizes compare?

A: The NVIDIA B200 SXM6 contains 208,000 million transistors on a 1,628 mm² die, while the AMD Radeon 890M contains 34,000 million transistors on a 233 mm² die. The B200 SXM6 has roughly 6.1 times more transistors and a die about 7 times larger.

The Verdict

The data indicates two processors built for completely separate workloads, and the benchmark record reflects that separation. The AMD Radeon 890M has a full suite of measured scores across DirectX, OpenCL, Vulkan, and PassMark tests, placing it at the 45th percentile of all GPUs in the database. Its nearest rivals are older mainstream parts like the GeForce GTX 960 and GTX 850M, all within one percent in average score. This positions the 890M as a capable integrated solution for mobile systems where power draw is constrained to 15 W.

The NVIDIA B200 SXM6 has no measured benchmark scores in the database. Its percentile field shows 50, but with an average score of zero and no rival comparisons, the recorded data cannot confirm any performance level relative to other parts. What the specification data does show is a part with 69.34 TFLOPS FP32, 180 GB of HBM3e memory at 8.19 TB/s bandwidth, 18,944 shading units, and 592 tensor cores. The 1000 W TDP and 1400 W suggested PSU make it clear this is a server-class module with no display outputs.

For the mobile integrated graphics segment, the 890M is the only one of the two that has recorded benchmark data, and that data shows it competing within one percent of the GeForce GTX 960 and GTX 850M. For server compute workloads, the B200 SXM6 offers memory capacity and bandwidth that the 890M cannot approach, since the 890M relies entirely on system shared memory. The 890M does offer DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4 support, while the B200 SXM6 lists no graphics API support. Any selection between these two parts depends entirely on whether the target system is a portable device with an integrated GPU or a server needing a high-bandwidth accelerator module. The benchmark database does not contain overlapping tests to rank them against each other.

Specification Differences

The two processors differ in nearly every recorded specification field. The AMD Radeon 890M uses a 4 nm process node at TSMC, while the NVIDIA B200 SXM6 uses a 5 nm node, also at TSMC. Transistor density is higher on the 890M at 145.9 million transistors per square millimeter versus 127.8 million for the B200 SXM6. Clock behavior differs substantially: the 890M has a base clock of 400 MHz and a boost clock of 2900 MHz, while the B200 SXM6 has a base clock of 120 MHz and a boost clock of 1830 MHz. Memory configurations are completely different, with the 890M using system shared memory and the B200 SXM6 using 180 GB of HBM3e on an 8192-bit bus with 8.19 TB/s bandwidth.

The 890M has 1,024 shading units, 64 TMUs, 32 ROPs, and 16 ray tracing cores. The B200 SXM6 has 18,944 shading units, 592 TMUs, 24 ROPs, and 592 tensor cores, with no ray tracing core count listed. Pixel rate is higher on the 890M at 92.80 GPixel/s versus 43.92 GPixel/s for the B200 SXM6, but texture rate is much higher on the B200 SXM6 at 1,083.4 GTexel/s versus 185.6 GTexel/s. FP32 and FP16 throughput are both 5.939 TFLOPS on the 890M and 69.34 TFLOPS on the B200 SXM6. Power draw ranges from 15 W on the 890M to 1000 W on the B200 SXM6. The 890M is an IGP with no power connectors, while the B200 SXM6 is an SXM Module with a 1400 W suggested PSU. Bus interfaces also differ: PCIe 4.0 x8 for the 890M and PCIe 6.0 x16 for the B200 SXM6. Display outputs are portable device dependent on the 890M and absent on the B200 SXM6. The 890M supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, while the B200 SXM6 lists all APIs as N/A. The 890M was released on 2024-07-14, and the B200 SXM6 on 2024-10-31. The B200 SXM6 has a launch MSRP of 34,999 USD.

Architecture Differences

The AMD Radeon 890M is built on the RDNA 3.5 architecture with the Strix Point chip, belonging to the Navi III IGP generation for Strix Point Mobile. The NVIDIA B200 SXM6 uses the Blackwell architecture with the GB100 chip, part of the Server Blackwell generation. The 890M's predecessor is listed as Navi II IGP, while the B200 SXM6's predecessor is Server Hopper and its successor is Server Rubin. The 890M is based on a 4 nm TSMC process with 34,000 million transistors on a 233 mm² die. The B200 SXM6 uses a 5 nm TSMC process with 208,000 million transistors on a 1,628 mm² die. Despite the larger node, the B200 SXM6 has roughly six times the transistor count of the 890M.

Ray tracing support is present on the 890M with 16 RT cores, while the B200 SXM6 has no RT core count recorded. Tensor cores are exclusive to the B200 SXM6, with 592 units, and no tensor core count is listed for the 890M. The 890M has 32 ROPs, which is higher than the B200 SXM6's 24 ROPs, a notable inversion given the B200 SXM6's far larger shading unit count. Both parts list FP16 throughput at a 1:1 ratio with FP32, meaning neither uses a half-rate FP16 path. The 890M has no dedicated memory of its own, sharing system memory with the host processor, whereas the B200 SXM6 carries 180 GB of HBM3e on a wide 8192-bit interface. The 890M is fabricated with a higher transistor density, 145.9 million transistors per square millimeter versus 127.8 million for the B200 SXM6, which reflects the different design goals of a low-power integrated GPU versus a high-performance server accelerator. The B200 SXM6 lists no display outputs, consistent with a compute-focused server module, while the 890M's display outputs depend on the portable device in which it is installed.