AMD Radeon RX 9060 XT LP vs NVIDIA CMP 90HX Comparison

The Verdict

The data presents a clear split decision driven by workload and physical constraints. The AMD Radeon RX 9060 XT LP is the decisive winner in raw compute performance, posting an 88,183 Geekbench OpenCL score versus 69,000 for the NVIDIA CMP 90HX, a 21.8% advantage. For any task leveraging OpenCL compute, the AMD card is categorically faster.

However, the NVIDIA CMP 90HX is a fundamentally different product with a fatal flaw for general use: it has no display outputs. Its purpose is mining, not rendering to a screen. The AMD Radeon RX 9060 XT LP offers 16 GB of GDDR6 memory, a 140 W TDP, and a PCIe 5.0 x16 interface, making it a functional, modern graphics card for gaming and workstation tasks. The NVIDIA card's 10 GB of GDDR6X and 320-bit bus are irrelevant if you cannot connect a monitor.

The verdict is therefore contextual. If your sole criterion is raw OpenCL compute throughput and you have a headless mining or server rig, the AMD card wins on performance per the benchmark. If you need a usable graphics card with display outputs, the NVIDIA CMP 90HX is disqualified entirely, regardless of its 90th percentile ranking. The AMD card, with an 89th percentile ranking, is the only viable choice for a standard desktop. The NVIDIA card is a specialized tool; the AMD card is a general-purpose component. The data shows that the AMD card is the superior purchase for anyone who needs a functional GPU.

Architecture Differences

The two cards represent opposing architectural philosophies and generations. The NVIDIA CMP 90HX is built on the Ampere architecture using the GA102 chip, fabricated on Samsung's 8 nm process. The AMD Radeon RX 9060 XT LP uses the RDNA 4.0 architecture with the Navi 44 chip, built on TSMC's 4 nm node. This process difference is stark: the AMD chip packs 29,700 million transistors into a 199 mm² die, yielding a transistor density of 149.2M per mm². The NVIDIA chip has 28,300 million transistors on a much larger 628 mm² die, resulting in a density of just 45.1M per mm². The AMD design is dramatically more efficient in terms of transistor packing.

Core configurations differ fundamentally. The NVIDIA card has 6,400 shading units, 200 texture mapping units, and 80 ROPs. It also includes 50 ray tracing cores and 200 tensor cores, features tailored for its Ampere architecture. The AMD card has 2,048 shading units, 128 TMUs, and 64 ROPs, with 32 ray tracing cores. Despite having fewer shading units, the AMD card achieves higher raw throughput thanks to its much higher boost clock of 3050 MHz versus 1710 MHz for the NVIDIA card. The AMD card also lacks dedicated tensor cores, a notable omission for AI workloads that rely on those units.

Memory subsystems are entirely different. The NVIDIA card uses 10 GB of GDDR6X on a 320-bit bus, delivering 760.3 GB/s of bandwidth. The AMD card uses 16 GB of GDDR6 on a 128-bit bus, with 322.3 GB/s of bandwidth. The NVIDIA card has more than double the memory bandwidth, but the AMD card has 60% more capacity. The clocks also diverge: the NVIDIA card's memory runs at 1188 MHz (19 Gbps effective), while the AMD card's memory runs at 2518 MHz (20.1 Gbps effective). The AMD card's memory is faster per pin, but the narrower bus limits total bandwidth.

Power and interface differences are equally pronounced. The NVIDIA card is rated at 320 W TDP, requires two 8-pin power connectors, and suggests a 700 W power supply. The AMD card is rated at 140 W TDP, uses a single 8-pin connector, and suggests a 300 W PSU. The AMD card also supports PCIe 5.0 x16, while the NVIDIA card is limited to PCIe 1.0 x4, a legacy interface that severely bottlenecks data transfer. The NVIDIA card has no display outputs; the AMD card has 1x HDMI 2.1b and 2x DisplayPort 2.1a. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

FAQ

Q: Which card is faster in OpenCL compute benchmarks?

A: The AMD Radeon RX 9060 XT LP scores 88,183 in Geekbench OpenCL, which is 21.8% higher than the NVIDIA CMP 90HX's 69,000 score.

Q: Can the NVIDIA CMP 90HX be used as a standard desktop graphics card?

A: No. The NVIDIA card has no display outputs, meaning it cannot connect to any monitor. The AMD card has 1x HDMI 2.1b and 2x DisplayPort 2.1a outputs.

Q: How do their memory capacities and bandwidths compare?

A: The NVIDIA card has 10 GB of GDDR6X with 760.3 GB/s bandwidth on a 320-bit bus. The AMD card has 16 GB of GDDR6 with 322.3 GB/s bandwidth on a 128-bit bus. The NVIDIA card has higher bandwidth, but the AMD card has more capacity.

Q: What are the power requirements for each card?

A: The NVIDIA CMP 90HX has a 320 W TDP, uses two 8-pin power connectors, and suggests a 700 W power supply. The AMD Radeon RX 9060 XT LP has a 140 W TDP, uses one 8-pin connector, and suggests a 300 W power supply.

Q: Which card has newer architecture and process technology?

A: The AMD card uses RDNA 4.0 on TSMC's 4 nm process. The NVIDIA card uses Ampere on Samsung's 8 nm process. The AMD card is built on a newer, denser process node.

Q: Do both cards support the same graphics APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. However, the NVIDIA card has 200 tensor cores, while the AMD card has none.

Specification Differences

The two cards differ on nearly every measurable specification. The process node is a major divergence: NVIDIA uses 8 nm from Samsung, while AMD uses 4 nm from TSMC. The transistor counts are close in absolute terms (28,300 million vs 29,700 million), but the die sizes are vastly different: 628 mm² for NVIDIA versus 199 mm² for AMD. The transistor density reflects this, with AMD at 149.2M per mm² versus NVIDIA's 45.1M per mm².

Clock speeds differ substantially. The NVIDIA base clock is 1500 MHz with a boost of 1710 MHz. The AMD base clock is 1380 MHz, but its boost clock is 3050 MHz, and it has a game clock of 2450 MHz. Memory clocks are also different: NVIDIA runs at 1188 MHz (19 Gbps effective), while AMD runs at 2518 MHz (20.1 Gbps effective). Memory configuration is a key split: NVIDIA has 10 GB of GDDR6X on a 320-bit bus with 760.3 GB/s bandwidth; AMD has 16 GB of GDDR6 on a 128-bit bus with 322.3 GB/s bandwidth.

The compute units are configured differently. NVIDIA has 6,400 shading units, 200 TMUs, 80 ROPs, 50 RT cores, and 200 tensor cores. AMD has 2,048 shading units, 128 TMUs, 64 ROPs, and 32 RT cores, with no tensor cores. Pixel and texture rates favor AMD: 195.2 GPixel/s and 390.4 GTexel/s versus 136.8 GPixel/s and 342.0 GTexel/s for NVIDIA. FP32 performance also favors AMD at 24.99 TFLOPS versus 21.89 TFLOPS for NVIDIA, with both offering 1:1 FP16.

Power and connectivity are dramatically different. NVIDIA has a 320 W TDP, two 8-pin connectors, and a 700 W PSU recommendation. AMD has a 140 W TDP, one 8-pin connector, and a 300 W PSU recommendation. The bus interface is a major gap: NVIDIA uses PCIe 1.0 x4, while AMD uses PCIe 5.0 x16. Display outputs are absent on NVIDIA but present on AMD (1x HDMI 2.1b, 2x DisplayPort 2.1a). Physical dimensions are only provided for NVIDIA: 285 mm in length and 112 mm in height. Production status differs: NVIDIA is end-of-life, released in 2021, while AMD is active, released in 2025.

Head-to-Head Benchmarks

The only head-to-head benchmark available is Geekbench OpenCL, and it is a decisive victory for the AMD Radeon RX 9060 XT LP. The AMD card scores 88,183 points, while the NVIDIA CMP 90HX scores 69,000 points. This represents a 21.8% performance gap in favor of AMD. The AMD card wins the single head-to-head comparison, giving it a 1-0 win tally.

Contextualizing this score against their respective peer groups is instructive. The NVIDIA card's 69,000 score places it at the 90th percentile of all GPUs. Its nearest rival is the Intel Arc A770, which scores 68,809, a difference of just 0.3%. The NVIDIA card also edges out the AMD Radeon Instinct MI25 (68,562, 0.6% slower) but trails the AMD Radeon Pro WX 8200 (69,870, which is 1.2% faster) and the NVIDIA Quadro P6000 (69,986, 1.4% faster). This shows the CMP 90HX is tightly clustered with these high-end workstation cards, but none of them reach the AMD RX 9060 XT LP's level.

The AMD card's 88,183 OpenCL score places it at the 89th percentile. Its nearest rival is the NVIDIA CMP 30HX at 63,842, which is essentially identical in average score (0% delta). Other close rivals include the AMD Radeon RX 7600M (63,775, 0.1% faster) and the AMD Radeon Pro Vega 56 (63,693, 0.2% faster). These rivals are all significantly below the RX 9060 XT LP's OpenCL score, indicating that this specific benchmark is a standout result for the AMD card. The 21.8% delta between the two cards in this comparison is far larger than any delta between the NVIDIA card and its nearest rivals, underscoring that the AMD card is in a higher performance tier for OpenCL compute despite its lower overall percentile rank. The AMD card's Vulkan score of 39,476 is not directly comparable to the NVIDIA card, as no Vulkan score is provided for the latter.