AMD Ryzen Z2 Go GPU vs NVIDIA GeForce RTX 5090 D V2 Comparison

Head-to-Head Benchmarks

The recorded data for these two products is extremely lopsided. The NVIDIA GeForce RTX 5090 D V2 carries a single benchmark entry, 3DMark Steel Nomad DX12, with a score of 16504. The AMD Ryzen Z2 Go GPU has no benchmark entries in the database, and its average benchmark score is recorded as zero. The head-to-head benchmark list is empty, and the win tally shows zero wins for the AMD part and zero wins for the NVIDIA part, because no direct comparative test was recorded.

That said, the available data still permits a meaningful comparison. The RTX 5090 D V2 achieves a percentile rank of 59 among all GPUs in the database, meaning it sits above the median performer. Its nearest rivals in the database are NVIDIA T400 with an average score of 16508 and a delta of 0 percent, AMD Radeon PRO W7500 with 16415 and a delta of 0.5 percent, NVIDIA RTX PRO 6000 Blackwell with 16408 and a delta of 0.6 percent, and AMD Radeon RX 5700 XT with 16361 and a delta of 0.9 percent. In every case, the RTX 5090 D V2 is within one percentage point of these scores, which indicates the recorded Steel Nomad result places it in a tightly clustered band of mid-range performers in this particular workload, despite the massive hardware specifications listed for the card.

The AMD Ryzen Z2 Go GPU, by contrast, has a percentile rank of 50 and an average benchmark score of zero. With no recorded scores, the database cannot place it against any rivals, and no nearest rival data exists for it. The interpretation is straightforward: the AMD part has no measured performance data in this database, so any head-to-head numerical comparison is impossible. What can be compared are the raw specifications, which show enormous gaps in every compute metric. The RTX 5090 D V2 lists 21760 shading units against 768 for the AMD part, 680 texture mapping units against 48, 176 raster output units against 32, 170 ray tracing cores against 12, and 680 tensor cores against none listed for the AMD part. The FP32 throughput is 104.8 TFLOPS versus 4.147 TFLOPS, a factor of roughly 25. The FP16 figure is 104.8 TFLOPS for NVIDIA versus 8.294 TFLOPS for AMD. Pixel rate is 423.6 GPixel/s versus 86.40 GPixel/s, and texture rate is 1,636.8 GTexel/s versus 129.6 GTexel/s.

The only metric where the AMD part shows any advantage is power consumption. The TDP is 28 W versus 575 W for the NVIDIA card. That is a 20.5x difference in rated power draw. The AMD part also requires no power connectors, while the NVIDIA card lists a single 16-pin connector and a suggested PSU of 950 W. These numbers frame the two products as entirely different classes of hardware, one designed for minimal power envelopes and the other for maximum compute throughput.

Where Each One Wins

Based strictly on the recorded data, the NVIDIA GeForce RTX 5090 D V2 wins every measurable performance category. The benchmark score of 16504 in Steel Nomad DX12 is the only actual performance result in the database, and it places the card at the 59th percentile. The AMD Ryzen Z2 Go GPU has no scores, so it cannot claim any benchmark win.

For use-case analysis, the AMD part wins on power efficiency. At 28 W TDP, it operates in a completely different thermal envelope than the 575 W NVIDIA card. The AMD part uses LPDDR5 memory with 16 GB capacity, a 128-bit bus, and 102.4 GB/s bandwidth. That memory configuration suits low-power, integrated-style workloads where battery life and thermal limits matter more than raw throughput. The absence of power connectors reinforces that this is a self-contained solution, likely for handheld or compact devices.

The NVIDIA card wins on memory capacity and bandwidth. It lists 24 GB of GDDR7 memory, a 384-bit bus, and 1.34 TB/s bandwidth. That is over 13 times the bandwidth of the AMD part. For workloads that depend on large datasets moving quickly, such as high-resolution rendering or large language model inference, the NVIDIA card has a decisive advantage. The 170 ray tracing cores and 680 tensor cores give it dedicated hardware for ray-traced graphics and AI acceleration, features the AMD part lacks entirely in terms of tensor cores and has far fewer of in ray tracing.

The AMD part supports the same API levels as the NVIDIA card. Both list DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. That means software compatibility is identical at the API level. The difference lies in execution resources. The AMD part's 12 ray tracing cores can technically process DirectX 12 Ultimate workloads, but at 4.147 TFLOPS FP32 and 86.40 GPixel/s, the throughput ceiling is far lower.

In practical terms, the AMD part wins for scenarios where power draw is the limiting factor, such as fanless designs or battery-powered systems. The NVIDIA card wins for every scenario where performance is the goal and power delivery can be managed. The database shows no overlap in their intended operating points.

Architecture Differences

The two products come from different architectural generations and process nodes. The AMD Ryzen Z2 Go GPU uses the Rembrandt+ chip, built on RDNA 2.0 architecture, fabricated on a 6 nm process at TSMC. The transistor count is 13,100 million on a die size of 208 mm², giving a transistor density of 63.0 million transistors per square millimeter. The NVIDIA GeForce RTX 5090 D V2 uses the GB202 chip, built on Blackwell 2.0 architecture, fabricated on a 5 nm process at TSMC. The transistor count is 92,200 million on a die size of 750 mm², giving a density of 122.9 million transistors per square millimeter. The NVIDIA chip has roughly 7 times the transistor count and 3.6 times the die area, with nearly double the transistor density.

Clock behavior differs significantly. The AMD part has a base clock of 800 MHz and a boost clock of 2700 MHz. The NVIDIA card has a base clock of 2017 MHz and a boost clock of 2407 MHz. The AMD part's boost clock is higher than the NVIDIA card's boost clock, but the NVIDIA card's base clock is much higher. The AMD memory clock is 800 MHz with 6.4 Gbps effective, while the NVIDIA memory clock is 1750 MHz with 28 Gbps effective. Memory type differs: LPDDR5 for AMD versus GDDR7 for NVIDIA. Bus width is 128-bit versus 384-bit.

The compute architecture shows the scale of difference. The AMD part has 768 shading units, 48 TMUs, 32 ROPs, and 12 ray tracing cores. The NVIDIA part has 21760 shading units, 680 TMUs, 176 ROPs, 170 ray tracing cores, and 680 tensor cores. The FP32 and FP16 rates are identical for the NVIDIA card at 104.8 TFLOPS, indicating 1:1 FP16 to FP32 ratio. The AMD part lists 4.147 TFLOPS FP32 and 8.294 TFLOPS FP16 at a 2:1 ratio.

The physical form factors diverge completely. The AMD part has no listed slot width, no dimensions, and no bus interface. It lists one USB Type-C display output. The NVIDIA card is dual-slot, measures 304 mm in length, 137 mm in height, and 48 mm in width. It uses PCIe 5.0 x16 and has one HDMI 2.1b and three DisplayPort 2.1b outputs.

Release timing also differs. The AMD part has a release date of 2024-12-31, while the NVIDIA card has a release date of 2025-08-14. The NVIDIA card lists a predecessor as GeForce 40 and a successor as GeForce 60. The AMD part lists no predecessor or successor. Both are marked as Active in production status.

The NVIDIA card carries a launch MSRP of 2,299 USD. The AMD part has no launch MSRP listed.

The Verdict

The data in the database paints a clear picture. The NVIDIA GeForce RTX 5090 D V2 is a high-power, high-throughput desktop graphics card with a recorded benchmark score of 16504 in Steel Nomad DX12, placing it at the 59th percentile. Its nearest rivals all sit within one percentage point of that score, which suggests this particular benchmark may not fully stress the card's capabilities, or the database's comparison pool is limited to similar-scoring parts. The card's specifications, however, are massive: 104.8 TFLOPS FP32, 24 GB GDDR7, 1.34 TB/s bandwidth, and 170 ray tracing cores. This is hardware built for maximum compute density, with a 575 W TDP and a 950 W suggested PSU.

The AMD Ryzen Z2 Go GPU is a fundamentally different product. It has no recorded benchmark scores, no nearest rivals, and a percentile rank of 50. Its specifications are those of a low-power part: 28 W TDP, 16 GB LPDDR5, 102.4 GB/s bandwidth, and 4.147 TFLOPS FP32. The 800 MHz base clock and 2700 MHz boost clock show a wide boost range, typical of a power-constrained design that can ramp up when thermal headroom allows. The single USB Type-C display output and lack of power connectors confirm an integrated or low-profile usage model.

No direct performance comparison exists in the database. The win tally is zero for both parts. The only benchmark score belongs to the NVIDIA card, and it is not compared against the AMD part. Therefore, any verdict must come from the specification data.

For users who need maximum compute performance, the RTX 5090 D V2 is the only part with measured performance data, and its specifications dwarf the AMD part in every compute metric. For users who need minimal power consumption in a compact form factor, the AMD part is the only option that fits that profile. The 28 W TDP versus 575 W TDP difference is the defining specification. The AMD part can operate without power connectors; the NVIDIA card requires a 16-pin connector and a 950 W PSU.

The database records the NVIDIA card's launch MSRP as 2,299 USD. No pricing data exists for the AMD part. Both products are Active in production status. The NVIDIA card has a clear predecessor and successor, indicating it sits in an established product line. The AMD part has neither, suggesting it may be a one-off or embedded solution.

The verdict from the data: the NVIDIA GeForce RTX 5090 D V2 is the performance leader by every measurable specification, and it has the only recorded benchmark score. The AMD Ryzen Z2 Go GPU is the power efficiency leader, with a TDP that is 28 W versus 575 W. The choice between them depends entirely on whether the primary constraint is compute throughput or power draw. The database provides no evidence that the AMD part can approach the NVIDIA card's performance, nor any evidence that the NVIDIA card can operate within the AMD part's power envelope.

FAQ

Q: What is the only recorded benchmark score in the database for these two products?

A: The NVIDIA GeForce RTX 5090 D V2 has a single 3DMark Steel Nomad DX12 score of 16504. The AMD Ryzen Z2 Go GPU has no recorded benchmark scores, with an average benchmark score of zero.

Q: How does the RTX 5090 D V2 compare to its nearest rivals in the database?

A: The nearest rivals are NVIDIA T400 with an average score of 16508 and a 0 percent delta, AMD Radeon PRO W7500 with 16415 and a 0.5 percent delta, NVIDIA RTX PRO 6000 Blackwell with 16408 and a 0.6 percent delta, and AMD Radeon RX 5700 XT with 16361 and a 0.9 percent delta. The RTX 5090 D V2 is within one percentage point of all of them.

Q: What are the FP32 compute rates for each product?

A: The AMD Ryzen Z2 Go GPU has 4.147 TFLOPS FP32. The NVIDIA GeForce RTX 5090 D V2 has 104.8 TFLOPS FP32.

Q: What memory configurations do the two products use?

A: The AMD part uses 16 GB of LPDDR5 on a 128-bit bus with 102.4 GB/s bandwidth and 6.4 Gbps effective memory speed. The NVIDIA card uses 24 GB of GDDR7 on a 384-bit bus with 1.34 TB/s bandwidth and 28 Gbps effective memory speed.

Q: What is the TDP difference between the two products?

A: The AMD Ryzen Z2 Go GPU has a TDP of 28 W and requires no power connectors. The NVIDIA GeForce RTX 5090 D V2 has a TDP of 575 W, uses a single 16-pin power connector, and lists a suggested PSU of 950 W.

Q: Which product has tensor cores and what is the count?

A: The NVIDIA GeForce RTX 5090 D V2 has 680 tensor cores. The AMD Ryzen Z2 Go GPU lists no tensor cores in the database.

Q: What are the process nodes and die sizes for each chip?

A: The AMD chip, Rembrandt+, uses a 6 nm process at TSMC with a die size of 208 mm² and 13,100 million transistors. The NVIDIA chip, GB202, uses a 5 nm process at TSMC with a die size of 750 mm² and 92,200 million transistors.