NVIDIA CMP 90HX vs NVIDIA GeForce RTX 5070 Ti Comparison

The Verdict

The data presents a stark generational contrast. The NVIDIA GeForce RTX 5070 Ti is the definitive winner in raw compute performance, posting a Geekbench OpenCL score of 212,363 against the NVIDIA CMP 90HX’s 69,000. This represents a 67.5% advantage for the newer card, a gap so large that it defines every practical use case. The CMP 90HX, designed for mining, has no display outputs, while the RTX 5070 Ti offers full display connectivity, making it the only viable choice for gaming or any visual workload. The CMP 90HX’s only claim to relevance is its 90th percentile ranking among all GPUs, which is respectable, but the RTX 5070 Ti’s 86th percentile is misleading given its vastly superior raw score. The RTX 5070 Ti is the clear pick for anyone needing compute, gaming, or content creation. The CMP 90HX is a niche product, and its end-of-life status, combined with a PCIe 1.0 x4 interface, makes it a poor choice for modern systems.

Architecture Differences

The two cards represent entirely different design philosophies and eras. The CMP 90HX is built on the Ampere architecture, using the GA102 chip manufactured on Samsung’s 8 nm process. The RTX 5070 Ti, in contrast, uses the Blackwell 2.0 architecture with the GB203 chip on TSMC’s 5 nm node. This node advantage is significant, as the RTX 5070 Ti packs 45,600 million transistors into a 378 mm² die, achieving a transistor density of 120.6M per mm². The CMP 90HX has 28,300 million transistors on a much larger 628 mm² die, resulting in a density of just 45.1M per mm². The newer chip is not only denser but also more efficient in its layout.

Core counts follow the same trend. The RTX 5070 Ti has 8,960 shading units, 280 TMUs, and 96 ROPs, while the CMP 90HX offers 6,400 shading units, 200 TMUs, and 80 ROPs. The RTX 5070 Ti also leads in ray tracing with 70 RT cores versus 50, and in tensor core count with 280 versus 200. The CMP 90HX’s memory configuration is a 10 GB GDDR6X setup on a 320-bit bus, yielding 760.3 GB/s of bandwidth. The RTX 5070 Ti uses 16 GB of GDDR7 on a 256-bit bus, which produces a higher 896.0 GB/s bandwidth despite the narrower interface. Clock speeds are also decisively in favor of the newer card: the RTX 5070 Ti has a base clock of 2295 MHz and a boost of 2452 MHz, compared to the CMP 90HX’s 1500 MHz base and 1710 MHz boost. The CMP 90HX’s memory runs at 1188 MHz (19 Gbps effective), while the RTX 5070 Ti’s memory runs at 1750 MHz (28 Gbps effective).

The bus interface is a critical architectural difference. The CMP 90HX uses PCIe 1.0 x4, a severely limited interface, while the RTX 5070 Ti uses PCIe 5.0 x16. This alone makes the older card unsuitable for general-purpose computing in modern motherboards. The CMP 90HX also lacks any display outputs, while the RTX 5070 Ti provides 1x HDMI 2.1b and 3x DisplayPort 2.1b. Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, so API compatibility is not a differentiator. The RTX 5070 Ti is a dual-slot card measuring 304 mm in length, 137 mm in height, and 48 mm in width, while the CMP 90HX is also dual-slot but shorter at 285 mm and slimmer at 112 mm in height. The CMP 90HX requires 2x 8-pin power connectors, while the RTX 5070 Ti uses a single 16-pin connector; both recommend a 700 W PSU.

Head-to-Head Benchmarks

The only direct benchmark recorded in the database is Geekbench OpenCL, and it is not close. The RTX 5070 Ti scores 212,363, while the CMP 90HX scores 69,000. The delta is 67.5% in favor of the RTX 5070 Ti. This is not a marginal improvement; it is a complete generational leap. The CMP 90HX’s nearest rivals in the database include the Intel Arc A770 at 68,809 (a 0.3% difference) and the AMD Radeon Instinct MI25 at 68,562 (0.6% higher). The AMD Radeon Pro WX 8200 and NVIDIA Quadro P6000 are slightly ahead at 69,870 and 69,986, respectively. This places the CMP 90HX in a performance tier that is competitive with older workstation and mid-range consumer cards, but far behind the modern flagship-class RTX 5070 Ti.

The RTX 5070 Ti’s nearest rivals tell a different story. Its average benchmark score is 49,957, but this figure is pulled down by multiple Passmark tests, including some low scores like Passmark DirectX 9 at 351 and Passmark DirectX 12 at 127. Its nearest rivals include the AMD Radeon RX Vega 64 at 50,001 (a 0.1% difference), the Intel Arc A550M at 49,737 (0.4% lower), and the AMD Radeon RX 6900 XT at 50,951 (2% higher). The AMD Radeon RX 6800 XT is 3.1% behind at 48,477. These comparisons show that, despite its massive OpenCL lead over the CMP 90HX, the RTX 5070 Ti sits in a crowded field of high-end GPUs when averaged across all tests.

The RTX 5070 Ti also has a broader benchmark portfolio, with scores in 3DMark Steel Nomad DX12 (6,604), Geekbench Vulkan (225,122), and Passmark GPU Compute (20,203). The CMP 90HX has only the single Geekbench OpenCL result. This lack of data for the CMP 90HX is itself a signal: it was not subjected to the same suite of tests, likely because it was not intended for standard graphics workloads. The RTX 5070 Ti’s Vulkan score of 225,122 is higher than its OpenCL score, indicating strong performance across different compute APIs.

The CMP 90HX does win in one narrow sense: its power draw is 320 W versus the RTX 5070 Ti’s 300 W, meaning the newer card delivers far more performance while consuming less power. The RTX 5070 Ti’s pixel rate is 235.4 GPixel/s and its texture rate is 686.6 GTexel/s, both far exceeding the CMP 90HX’s 136.8 GPixel/s and 342.0 GTexel/s. FP32 performance is 43.94 TFLOPS for the RTX 5070 Ti versus 21.89 TFLOPS for the CMP 90HX, and FP16 performance is identical at 43.94 TFLOPS versus 21.89 TFLOPS, with both cards offering a 1:1 ratio.

Specification Differences

The table below highlights only the fields where the two cards differ. The CMP 90HX is an end-of-life mining product from 2021, while the RTX 5070 Ti is an active GeForce 50-series card released in 2025. The RTX 5070 Ti has a launch MSRP of 749 USD. The chip, architecture, process node, foundry, transistor count, die size, and density all differ, as do the clock speeds, memory configuration, core counts, and power requirements. The CMP 90HX has no display outputs, while the RTX 5070 Ti has full connectivity. The bus interface is PCIe 1.0 x4 for the CMP 90HX and PCIe 5.0 x16 for the RTX 5070 Ti. The CMP 90HX measures 285 mm by 112 mm, while the RTX 5070 Ti measures 304 mm by 137 mm by 48 mm. The power connectors are 2x 8-pin versus 1x 16-pin, and the TDP is 320 W versus 300 W.

FAQ

Q: Which card has higher raw compute performance?

A: The NVIDIA GeForce RTX 5070 Ti, with a Geekbench OpenCL score of 212,363 compared to the NVIDIA CMP 90HX’s 69,000. The delta is 67.5% in favor of the RTX 5070 Ti.

Q: Does the CMP 90HX support display outputs?

A: No. The CMP 90HX has no display outputs. The RTX 5070 Ti provides 1x HDMI 2.1b and 3x DisplayPort 2.1b.

Q: How do their memory configurations compare?

A: The CMP 90HX has 10 GB of GDDR6X on a 320-bit bus with 760.3 GB/s bandwidth. The RTX 5070 Ti has 16 GB of GDDR7 on a 256-bit bus with 896.0 GB/s bandwidth.

Q: Which card has more shading units?

A: The RTX 5070 Ti has 8,960 shading units, while the CMP 90HX has 6,400. The RTX 5070 Ti also leads in TMUs (280 vs 200) and ROPs (96 vs 80).

Q: What are the power requirements?

A: The CMP 90HX has a TDP of 320 W and requires 2x 8-pin power connectors. The RTX 5070 Ti has a TDP of 300 W and uses a single 16-pin connector. Both recommend a 700 W PSU.

Q: Are there any advantages to the CMP 90HX?

A: The CMP 90HX has a higher percentile ranking (90th vs 86th) and a larger die size (628 mm² vs 378 mm²), but its only benchmark score is 69,000, which is far below the RTX 5070 Ti’s 212,363 in the same test.