NVIDIA CMP 70HX vs NVIDIA GeForce RTX 3090 Comparison

Head-to-Head Benchmarks

The recorded data shows a decisive overall victory for the NVIDIA GeForce RTX 3090, which takes both of the two head-to-head benchmark tests. The CMP 70HX does not secure a single win in the direct comparison, with the scoreboard reading 2 wins for the RTX 3090 and 0 for the CMP 70HX.

The most significant gap appears in the Geekbench OpenCL test, which measures general-purpose compute performance. The RTX 3090 scores 172,758, while the CMP 70HX manages 25,135. That translates to a delta of -85.5%, meaning the RTX 3090 is roughly 5.9 times faster in this workload. This is an enormous margin, and it reflects the fundamental difference in scale between the two GPUs. The RTX 3090's raw compute resources simply overwhelm the CMP 70HX in this type of workload. For any task that leans heavily on OpenCL compute, the RTX 3090 is in a different league entirely.

The second test, Geekbench Vulkan, shows a narrower but still clear gap. The RTX 3090 posts 53,927, while the CMP 70HX reaches 35,817. The delta here is -33.6%, indicating that the RTX 3090 is about half again as fast as the CMP 70HX in Vulkan graphics workloads. While the CMP 70HX is more competitive in this test, it still falls well short of the RTX 3090. The Vulkan result is a better reflection of real-world gaming and graphics API performance, and it shows that the CMP 70HX can hold its own to a degree, but it cannot match the top-tier performance of the RTX 3090.

Looking at the broader benchmark averages, the two cards sit at different positions in the database. The CMP 70HX has an average benchmark score of 30,476 across its two recorded tests, while the RTX 3090 has an average of 27,565 across its ten recorded tests. This average is skewed by the different test sets, but the head-to-head numbers are the authoritative comparison here. The RTX 3090's nearest rivals in the database include the NVIDIA GeForce RTX 4070 Mobile with a delta of 0.5 percent, the AMD Radeon RX 6700 XT with a delta of 0.5 percent, the AMD Radeon Pro Vega 20 with a delta of -1 percent, and the AMD Radeon RX 7800M with a delta of -1.1 percent. The CMP 70HX's nearest rivals are the NVIDIA Tesla M60 with a delta of 0 percent, the AMD Radeon RX 6700 with a delta of 0.1 percent, the AMD Radeon RX 6800 with a delta of 1.3 percent, and the NVIDIA GeForce RTX 3070 Ti with a delta of 1.8 percent.

The percentile rankings tell a similar story. The CMP 70HX sits at the 75th percentile among all GPUs in the database, while the RTX 3090 sits at the 73rd percentile. These are close percentile positions, but the head-to-head results clearly favor the RTX 3090. The CMP 70HX's percentile is boosted by its limited test set, while the RTX 3090's percentile is pulled down by the inclusion of older DirectX and PassMark tests where it scores relatively lower. In the two tests where both cards were measured side by side, the RTX 3090 wins decisively in both.

The Verdict

The data is unambiguous: the NVIDIA GeForce RTX 3090 is the faster card in every recorded head-to-head benchmark. If the choice is purely about performance, the RTX 3090 wins outright. The OpenCL gap of -85.5 percent is a massive chasm, and even the Vulkan gap of -33.6 percent is substantial. Anyone who needs maximum compute throughput or strong graphics API performance should take the RTX 3090 without hesitation.

However, the CMP 70HX is not without its place. Its average benchmark score of 30,476 is actually higher than the RTX 3090's average of 27,565, though this is due to the different test suites. The CMP 70HX also holds the higher percentile ranking at 75 versus 73. For buyers who are strictly looking at the specific tests recorded for the CMP 70HX, it may appear competitive in aggregate. But the direct comparison shows that this is a misleading picture.

The CMP 70HX is a mining-focused product with no display outputs, making it unsuitable for any normal desktop use that requires a monitor connection. The RTX 3090, by contrast, offers full display connectivity with one HDMI 2.1 port and three DisplayPort 1.4a outputs. If you need to see the output of your GPU, the CMP 70HX is not an option at all. The RTX 3090 is the only one of the two that can function as a conventional graphics card.

Architecture Differences

Both GPUs are built on the Ampere architecture and use Samsung's 8 nm process node. The transistor density is nearly identical: 44.4 million transistors per square millimeter for the CMP 70HX and 45.1 million for the RTX 3090. The similarity ends there.

The CMP 70HX uses the GA104 chip, which packs 17,400 million transistors onto a die size of 392 square millimeters. The RTX 3090 uses the much larger GA102 chip, with 28,300 million transistors on a 628 square millimeter die. That is a significant difference in raw silicon area and transistor count, and it directly explains the performance gap.

The compute resources differ dramatically. The CMP 70HX has 3,840 shading units, 120 texture mapping units, and 64 render output units. The RTX 3090 has 10,496 shading units, 328 texture mapping units, and 112 render output units. The RTX 3090 more than doubles the shading units and nearly triples the texture units. This is why the FP32 compute throughput is 35.58 TFLOPS for the RTX 3090 versus 10.71 TFLOPS for the CMP 70HX. The FP16 numbers are identical to the FP32 numbers for both cards, indicating a 1:1 ratio.

Ray tracing and tensor cores follow the same pattern. The CMP 70HX has 30 RT cores and 120 tensor cores, while the RTX 3090 has 82 RT cores and 328 tensor cores. The RTX 3090 is built for the full suite of modern graphics features, including ray tracing and AI acceleration, while the CMP 70HX has scaled-down versions of these blocks.

The CMP 70HX has a base clock of 1365 MHz and a boost clock of 1395 MHz. The RTX 3090 runs higher, with a base clock of 1395 MHz and a boost clock of 1695 MHz. The RTX 3090 both starts higher and boosts further, which compounds its architectural advantage.

Specification Differences

The memory subsystems are another major point of divergence. The CMP 70HX has 8 GB of GDDR6X memory on a 256-bit bus, delivering 608.3 GB/s of bandwidth. The RTX 3090 has 24 GB of GDDR6X memory on a 384-bit bus, delivering 936.2 GB/s of bandwidth. That is three times the capacity and roughly 54 percent more bandwidth. The effective memory speed is 19 Gbps for the CMP 70HX and 19.5 Gbps for the RTX 3090, a small difference that is dwarfed by the bus width advantage.

Pixel and texture rates reflect the larger silicon. The CMP 70HX outputs 89.28 GPixel/s and 167.4 GTexel/s. The RTX 3090 outputs 189.8 GPixel/s and 556.0 GTexel/s. The RTX 3090 more than doubles the pixel rate and more than triples the texture rate.

The power and physical characteristics are also distinct. The CMP 70HX has a suggested PSU of 200 W, while the RTX 3090 has a suggested PSU of 750 W and a TDP of 350 W. The CMP 70HX is a dual-slot card measuring 267 mm in length and 112 mm in height. The RTX 3090 is a triple-slot card measuring 336 mm in length, 140 mm in height, and 61 mm in width. Both use a single 12-pin power connector.

The bus interface differs as well. The CMP 70HX uses PCIe 1.0 x4, which is an unusual and severely limited interface. The RTX 3090 uses PCIe 4.0 x16, the full-bandwidth standard interface. The CMP 70HX has no display outputs, while the RTX 3090 offers one HDMI 2.1 port and three DisplayPort 1.4a outputs. The RTX 3090 was released on 2020-08-31, with a launch MSRP of 1,499 USD. The CMP 70HX has no recorded release date or launch MSRP in the database.

FAQ

Q: Which GPU is faster in OpenCL compute workloads?

A: The RTX 3090 is far faster, scoring 172,758 versus 25,135 for the CMP 70HX, a delta of -85.5 percent.

Q: How do the two cards compare in Vulkan performance?

A: The RTX 3090 wins again, scoring 53,927 versus 35,817 for the CMP 70HX, a delta of -33.6 percent.

Q: Does the CMP 70HX have any display outputs?

A: No, the CMP 70HX has no display outputs. The RTX 3090 has one HDMI 2.1 port and three DisplayPort 1.4a outputs.

Q: How much memory does each card have?

A: The CMP 70HX has 8 GB of GDDR6X memory, while the RTX 3090 has 24 GB of GDDR6X memory.

Q: What is the difference in shading units?

A: The CMP 70HX has 3,840 shading units, while the RTX 3090 has 10,496 shading units.

Q: Which card has a higher boost clock?

A: The RTX 3090 boosts to 1695 MHz, while the CMP 70HX boosts to 1395 MHz.

Where Each One Wins

The RTX 3090 wins in every scenario where raw performance matters. OpenCL compute workloads are its strongest area, with a massive 85.5 percent lead over the CMP 70HX. Vulkan graphics workloads also favor the RTX 3090, which holds a 33.6 percent advantage. Any task involving ray tracing, tensor core acceleration, or large memory buffers will favor the RTX 3090 due to its 82 RT cores, 328 tensor cores, and 24 GB of memory. The RTX 3090 also wins in any scenario requiring a display output, as the CMP 70HX cannot output video at all.

The CMP 70HX finds its niche in very specific conditions. Its average benchmark score of 30,476 is higher than the RTX 3090's 27,565, and it sits at the 75th percentile versus the RTX 3090's 73rd. The CMP 70HX also has a much lower suggested PSU requirement of 200 W versus 750 W for the RTX 3090, making it easier to integrate into systems with smaller power supplies. Its dual-slot design and shorter length of 267 mm make it more physically compact than the triple-slot, 336 mm RTX 3090. For a system that has no need for display outputs, requires a smaller physical footprint, and operates within a lower power budget, the CMP 70HX could be the more practical choice. But for anyone who needs the best possible performance in the recorded benchmarks, the RTX 3090 is the clear winner.