AMD Steam Machine GPU vs NVIDIA Rubin GPU Comparison

Head-to-Head Benchmarks

The recorded data shows no direct head-to-head benchmark results between the AMD Steam Machine GPU and the NVIDIA Rubin GPU. Both entries carry an average benchmark score of zero, and the head-to-head benchmark array is empty. The percentile versus all GPUs is identical for both parts at 50, indicating that without measured performance data, both occupy the same neutral position in the database. What can be compared are the raw compute specifications, and those differences are substantial.

The AMD Steam Machine GPU delivers 17.56 TFLOPS of FP32 compute, while the NVIDIA Rubin GPU delivers 130.0 TFLOPS. That places the NVIDIA part at roughly 7.4 times the FP32 throughput of the AMD part. In FP16 workloads, the AMD GPU offers 17.56 TFLOPS with a 1:1 ratio, meaning FP16 and FP32 throughput are equal. The NVIDIA Rubin GPU provides 260.0 TFLOPS of FP16 with a 2:1 ratio, doubling its FP32 rate. The NVIDIA part therefore leads in both precision formats by very wide margins.

Pixel throughput tells a different story. The AMD Steam Machine GPU achieves 156.8 GPixel/s, while the NVIDIA Rubin GPU achieves 54.41 GPixel/s. The AMD part is approximately 2.9 times faster in pixel fill rate. This is a notable result because it indicates that the AMD GPU is designed for rasterization-heavy workloads where pixel output matters, while the NVIDIA part prioritizes compute throughput over traditional pixel pushing.

Texture rate shows the NVIDIA Rubin GPU ahead at 2,031.2 GTexel/s versus 274.4 GTexel/s for the AMD Steam Machine GPU. That is a 7.4 times advantage for the NVIDIA part, roughly matching the FP32 ratio. The NVIDIA part also carries a much larger shading unit count at 28,716 versus 1,792 for the AMD part, which is approximately 16 times more shading units.

Memory bandwidth is another decisive split. The AMD GPU uses 8 GB of GDDR6 on a 128 bit bus, yielding 288.0 GB/s. The NVIDIA Rubin GPU uses 288 GB of HBM4 on a 16384 bit bus, yielding 22.1 TB/s. The NVIDIA bandwidth advantage is about 77 times greater. Memory capacity differs by a factor of 36, with the NVIDIA part at 288 GB versus 8 GB for the AMD part.

Where Each One Wins

The AMD Steam Machine GPU wins in pixel fill rate, delivering 156.8 GPixel/s against 54.41 GPixel/s for the NVIDIA Rubin GPU. This suggests the AMD part is better suited to workloads that are heavily rasterization-bound, where the number of pixels written per second directly impacts frame output. The ROP count supports this interpretation, with the AMD part carrying 64 ROPs versus 24 ROPs for the NVIDIA part.

The AMD part also wins on power efficiency in a practical sense. Its TDP is 110 W, while the NVIDIA Rubin GPU has a TDP of 2300 W. The AMD card draws no external power connectors, indicating that it operates entirely from slot power. The NVIDIA part requires a suggested PSU of 2700 W and mounts as an SXM module, which is a server form factor rather than a consumer graphics card.

The NVIDIA Rubin GPU wins decisively in compute throughput, memory capacity, memory bandwidth, and texture rate. Its FP32 figure of 130.0 TFLOPS versus 17.56 TFLOPS for the AMD part establishes a clear compute advantage. The FP16 figure of 260.0 TFLOPS versus 17.56 TFLOPS reinforces that lead. The NVIDIA part has 896 tensor cores, while the AMD part has no tensor core entry in the database. The AMD part does have 28 RT cores, while the NVIDIA part has no RT core count listed. The AMD GPU also supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, while the NVIDIA Rubin GPU lists N/A for all three APIs.

The AMD GPU includes display outputs with 1x HDMI 2.1a and 1x DisplayPort 2.1. The NVIDIA Rubin GPU lists no outputs. That makes the AMD part the only one of the two that can drive a display directly.

The Verdict

The data indicates that these two GPUs are not competing in the same product category. The AMD Steam Machine GPU is a console-oriented GPU from the Valve generation, built on the Navi 33 chip with RDNA 3.0 architecture. The NVIDIA Rubin GPU is a server part from the Rubin generation, built on the GR100 chip. The process nodes differ: the AMD part is on 6 nm, the NVIDIA part is on 3 nm. Transistor counts are 13,300 million for the AMD part versus 336,000 million for the NVIDIA part, a 25-fold difference. Die size is 204 mm² for the AMD part versus 1456 mm² for the NVIDIA part, which is about 7.1 times larger. Transistor density is 65.2M per mm² for the AMD part versus 230.8M per mm² for the NVIDIA part.

A user selecting a GPU for a display-connected, low-power, rasterization-focused system would be directed by the data toward the AMD Steam Machine GPU. It offers display outputs, a 110 W TDP, and a higher pixel fill rate. A user selecting a GPU for compute-heavy server workloads with massive memory capacity would be directed toward the NVIDIA Rubin GPU. Its 288 GB of HBM4 memory and 22.1 TB/s bandwidth are in a different class entirely.

Neither part has benchmark scores recorded, so the database does not currently support a performance ranking between them. The specification comparison is the only available basis for differentiation.

FAQ

Q: Which GPU has more FP32 compute power?

A: The NVIDIA Rubin GPU delivers 130.0 TFLOPS of FP32 compute, while the AMD Steam Machine GPU delivers 17.56 TFLOPS.

Q: Which GPU has higher pixel fill rate?

A: The AMD Steam Machine GPU achieves 156.8 GPixel/s, which is higher than the 54.41 GPixel/s of the NVIDIA Rubin GPU.

Q: What memory configurations do the two GPUs use?

A: The AMD Steam Machine GPU uses 8 GB of GDDR6 on a 128 bit bus with 288.0 GB/s bandwidth. The NVIDIA Rubin GPU uses 288 GB of HBM4 on a 16384 bit bus with 22.1 TB/s bandwidth.

Q: Does the NVIDIA Rubin GPU support DirectX, OpenGL, or Vulkan?

A: No. The database lists N/A for DirectX, OpenGL, and Vulkan on the NVIDIA Rubin GPU. The AMD Steam Machine GPU supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: Which GPU has display outputs?

A: The AMD Steam Machine GPU has 1x HDMI 2.1a and 1x DisplayPort 2.1. The NVIDIA Rubin GPU lists no display outputs.

Q: What are the TDP figures for each GPU?

A: The AMD Steam Machine GPU has a TDP of 110 W and uses no external power connectors. The NVIDIA Rubin GPU has a TDP of 2300 W and a suggested PSU of 2700 W.

Architecture Differences

The AMD Steam Machine GPU uses the RDNA 3.0 architecture with the Navi 33 chip and the codename Hotpink Bonefish. It belongs to the Console GPU (Valve) generation. The process node is 6 nm from TSMC. The transistor count is 13,300 million on a die size of 204 mm², giving a transistor density of 65.2M per mm².

The NVIDIA Rubin GPU uses the Rubin architecture with the GR100 chip. It belongs to the Server Rubin (Rxx) generation. The process node is 3 nm from TSMC. The transistor count is 336,000 million on a die size of 1456 mm², giving a transistor density of 230.8M per mm². The NVIDIA part has a predecessor listed as Server Blackwell.

The AMD GPU has 1792 shading units, 112 TMUs, 64 ROPs, 28 RT cores, and no tensor core entry. The NVIDIA GPU has 28672 shading units, 896 TMUs, 24 ROPs, no RT core entry, and 896 tensor cores. The NVIDIA part has nearly 16 times the shading units and 8 times the TMUs, but fewer than half the ROPs.

The AMD GPU is fabricated with a smaller die but also a much smaller transistor budget. The NVIDIA die is about 7.1 times larger by area and carries about 25 times more transistors. The node advantage for the NVIDIA part (3 nm versus 6 nm) contributes to a transistor density that is about 3.5 times higher.

Clock behavior differs as well. The AMD GPU has a base clock of 1720 MHz, a game clock of 2250 MHz, and a boost clock of 2450 MHz. The NVIDIA GPU has a base clock of 700 MHz and a boost clock of 2267 MHz, with no game clock listed. The memory clock for the AMD part is 2250 MHz at 18 Gbps effective, while the NVIDIA part runs at 2695 MHz at 10.8 Gbps effective.

Specification Differences

The AMD Steam Machine GPU measures 156 mm in length, 152 mm in height, and 162 mm in width. The NVIDIA Rubin GPU has no dimensions listed in the database. The AMD part is a slot-powered card with no external power connectors, while the NVIDIA part is an SXM module with no power connector details and a suggested PSU of 2700 W.

The bus interface differs. The AMD GPU has no bus interface listed, while the NVIDIA Rubin GPU uses PCIe 6.0 x16.

Memory type, bus width, and bandwidth all differ. The AMD part uses GDDR6 with a 128 bit bus and 288.0 GB/s bandwidth. The NVIDIA part uses HBM4 with a 16384 bit bus and 22.1 TB/s bandwidth. Memory size is 8 GB for the AMD part and 288 GB for the NVIDIA part.

FP16 behavior differs. The AMD GPU runs FP16 at 17.56 TFLOPS with a 1:1 ratio to FP32. The NVIDIA GPU runs FP16 at 260.0 TFLOPS with a 2:1 ratio.

API support differs completely. The AMD GPU lists DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The NVIDIA GPU lists N/A for all three.

Display outputs differ. The AMD GPU has 1x HDMI 2.1a and 1x DisplayPort 2.1. The NVIDIA GPU has no outputs.

Release dates differ. The AMD Steam Machine GPU has a release date of 2026-06-28, while the NVIDIA Rubin GPU has a release date of 2025-12-31. Both are marked as Active production status. Neither part has a launch MSRP in the database.