NVIDIA CMP 90HX vs NVIDIA RTX A1000 Mobile Comparison

The Verdict

The recorded data draws a clear line between these two Ampere-based NVIDIA parts. The NVIDIA CMP 90HX is the outright performance leader, delivering a Geekbench OpenCL score of 69000 against the RTX A1000 Mobile’s 48703, a 41.7% advantage. The CMP 90HX also sits at the 90th percentile of all GPUs in the database, while the RTX A1000 Mobile lands at the 85th percentile. For any workload that leans on raw compute throughput, the CMP 90HX is the only choice from this pairing.

The RTX A1000 Mobile, however, is not without its own rationale. It is an IGP-class part with a 60 W TDP, no power connectors, and a PCIe 4.0 x8 interface, making it suitable for portable or embedded systems where space and power are constrained. The CMP 90HX, by contrast, is a dual-slot card requiring two 8-pin connectors and a 700 W suggested PSU, with no display outputs at all. The data suggests two different worlds: one built for maximum throughput in a mining or compute rig, the other for mobile workstations where efficiency and integration matter more than peak scores.

Strictly from the benchmarks, the verdict is simple: pick the CMP 90HX if compute performance is the sole criterion, and pick the RTX A1000 Mobile if the platform demands a low-power, integrated solution. There is no middle ground in the measured results; the delta is too large and the form factors too divergent.

Architecture Differences

Both GPUs share the Ampere architecture and are fabricated on Samsung’s 8 nm process, but the silicon is fundamentally different in scale. The CMP 90HX uses the GA102 chip, a 628 mm² die with 28,300 million transistors, yielding a transistor density of 45.1M per mm². The RTX A1000 Mobile uses the GA107 chip, a 200 mm² die with 8,700 million transistors, at a density of 43.5M per mm². The density figures are nearly identical, but the absolute resources are not.

The CMP 90HX packs 6400 shading units, 200 texture mapping units, 80 ROPs, 50 RT cores, and 200 tensor cores. The RTX A1000 Mobile has 2048 shading units, 64 TMUs, 32 ROPs, 16 RT cores, and 64 tensor cores. That is a 3.1x difference in shading units and a 3.1x difference in tensor cores, which directly explains the large performance gap in the OpenCL benchmark.

Memory architecture further separates them. The CMP 90HX uses 10 GB of GDDR6X on a 320-bit bus, delivering 760.3 GB/s of bandwidth. The RTX A1000 Mobile uses 4 GB of GDDR6 on a 128-bit bus, with 176.0 GB/s. The bandwidth gap is 4.3x, far larger than the core resource gap. The CMP 90HX also runs at much higher clocks: 1500 MHz base and 1710 MHz boost, versus 630 MHz base and 1140 MHz boost for the mobile part. The memory clocks differ too, with the CMP 90HX at 1188 MHz (19 Gbps effective) and the RTX A1000 Mobile at 1375 MHz (11 Gbps effective).

Both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The RTX A1000 Mobile has display outputs (portable device dependent), while the CMP 90HX has none. The CMP 90HX’s bus interface is listed as PCIe 1.0 x4, which is a curious limitation for a high-end compute card, while the RTX A1000 Mobile uses PCIe 4.0 x8. This suggests the CMP 90HX is not designed for general-purpose data transfer, but rather for isolated compute tasks where the bus is less relevant.

Head-to-Head Benchmarks

The only head-to-head test in the database is Geekbench OpenCL, and the result is decisive. The CMP 90HX scores 69000, while the RTX A1000 Mobile scores 48703. The delta is 41.7% in favor of the CMP 90HX. That is not a marginal win; it is a dominant one. For context, the CMP 90HX’s nearest rivals in the database include the Intel Arc A770 at 68809 (0.3% behind), the AMD Radeon Instinct MI25 at 68562 (0.6% behind), the AMD Radeon Pro WX 8200 at 69870 (1.2% ahead), and the NVIDIA Quadro P6000 at 69986 (1.4% ahead). The CMP 90HX sits within 1.4% of these cards, indicating it is competitively positioned among high-end desktop and workstation GPUs.

The RTX A1000 Mobile, on the other hand, is closest to the AMD Radeon RX 6800 XT at 48477 (1.5% ahead), the AMD Radeon RX 6550M at 46702 (2.2% behind), the Intel Arc A530M at 46614 (2.4% behind), and the AMD Radeon RX 5600M at 46601 (2.5% behind). It also has a Vulkan score of 46782, which is not directly compared to the CMP 90HX in the head-to-head data, but it is notably lower than its own OpenCL result. The CMP 90HX has no Vulkan score recorded, so the comparison is limited to OpenCL only.

The 41.7% delta is consistent with the hardware differences. The CMP 90HX has more than three times the shading units and more than four times the memory bandwidth, so the performance gap is expected. What is striking is that the gap is not even larger, given the resource disparity. This suggests the RTX A1000 Mobile’s compute efficiency is relatively high, or that the OpenCL workload is not fully saturating the CMP 90HX’s resources.

Specification Differences

The two cards differ in nearly every measurable specification. The CMP 90HX uses the GA102 chip, while the RTX A1000 Mobile uses GA107. The CMP 90HX has 28,300 million transistors on a 628 mm² die; the RTX A1000 Mobile has 8,700 million on 200 mm². Transistor density is similar (45.1M versus 43.5M per mm²), but the physical scale is vastly different.

Clock speeds: the CMP 90HX runs at 1500 MHz base and 1710 MHz boost, while the RTX A1000 Mobile runs at 630 MHz base and 1140 MHz boost. Memory clocks are 1188 MHz (19 Gbps effective) versus 1375 MHz (11 Gbps effective). Memory size is 10 GB GDDR6X versus 4 GB GDDR6. Bus width is 320-bit versus 128-bit. Bandwidth is 760.9 GB/s versus 176.0 GB/s.

Core counts: 6400 shading units versus 2048, 200 TMUs versus 64, 80 ROPs versus 32, 50 RT cores versus 16, 200 tensor cores versus 64. Pixel rate is 136.8 GPixel/s versus 36.48 GPixel/s. Texture rate is 342.0 GTexel/s versus 72.96 GTexel/s. FP32 throughput is 21.89 TFLOPS versus 4.669 TFLOPS. FP16 is identical at 4.669 TFLOPS for both, but the FP32 figures show a 4.7x gap.

Power and physical dimensions: TDP is 320 W versus 60 W. The CMP 90HX is dual-slot, requires 2x 8-pin connectors and a 700 W suggested PSU. The RTX A1000 Mobile is IGP, has no power connectors, and no suggested PSU. The CMP 90HX is 285 mm long (11.2 inches) and 112 mm high (4.4 inches). The RTX A1000 Mobile has no recorded dimensions. The CMP 90HX has no display outputs, while the RTX A1000 Mobile outputs are portable device dependent. Bus interface is PCIe 1.0 x4 for the CMP 90HX versus PCIe 4.0 x8 for the RTX A1000 Mobile.

Release timing differs as well: the CMP 90HX was released on 2021-04-27, while the RTX A1000 Mobile on 2022-04-29. The CMP 90HX is end-of-life with no predecessor or successor; the RTX A1000 Mobile has a predecessor (Quadro Turing-M) and a successor (Ada-MW). Neither has a launch MSRP recorded.

FAQ

Q: Which GPU has a higher transistor density, and what does that indicate?

A: The CMP 90HX has a density of 45.1M transistors per mm², while the RTX A1000 Mobile has 45.0M per mm², a negligible difference. The CMP 90HX’s larger die size (628 mm²) and higher transistor count (28,300 million) mean it packs far more absolute resources, which is why its OpenCL score is 41.7% higher.

Q: Does the RTX A1000 Mobile have any benchmark advantage over the CMP 90HX?

A: In the recorded data, no. The CMP 90HX wins the only head-to-head test (Geekbench OpenCL) by 41.7%. The RTX A1000 Mobile has an additional Vulkan score of 46782, but no comparable CMP 90HX Vulkan result exists, so no direct comparison can be made.

Q: Why does the CMP 90HX have such a high bandwidth figure?

A: The CMP 90HX uses 10 GB of GDDR6X memory on a 320-bit bus, producing 760.3 GB/s of bandwidth. The RTX A1000 Mobile uses 4 GB of GDDR6 on a 128-bit bus, yielding 176.0 GB/s. The higher bus width and faster memory type drive the 4.3x bandwidth advantage.

Q: Are these GPUs in the same performance tier per the database?

A: No. The CMP 90HX is at the 90th percentile of all GPUs, while the RTX A1000 Mobile is at the 85th percentile. The CMP 90HX’s nearest rivals are desktop and workstation cards like the Intel Arc A770 and NVIDIA Quadro P6000, while the RTX A1000 Mobile’s rivals are mobile and mid-range parts like the AMD Radeon RX 6550M and Intel Arc A530M.

Q: Which GPU is more suitable for a system with limited power delivery?

A: The RTX A1000 Mobile, with a 60 W TDP, no power connectors, and an IGP form factor. The CMP 90HX requires 320 W, dual 8-pin connectors, and a 700 W suggested PSU, making it incompatible with constrained power environments.

Q: What is the difference in FP32 compute performance?

A: The CMP 90HX delivers 21.89 TFLOPS of FP32, while the RTX A1000 Mobile delivers 4.669 TFLOPS. The CMP 90HX is 4.7x faster in raw single-precision compute, which aligns with its 41.7% advantage in the OpenCL benchmark.

Where Each One Wins

The CMP 90HX wins in every measured compute category. Its OpenCL score of 69000 is 41.7% higher than the RTX A1000 Mobile’s 48703. It also leads in pixel rate (136.8 GPixel/s versus 36.48 GPixel/s), texture rate (342.0 GTexel/s versus 72.96 GTexel/s), and FP32 throughput (21.89 TFLOPS versus 4.669 TFLOPS). For workloads that are bandwidth-hungry, the CMP 90HX’s 760.3 GB/s versus 176.0 GB/s is a decisive advantage. This makes it the clear winner for compute-heavy tasks like large-scale rendering, scientific simulation, or any OpenCL-accelerated workload where the card’s lack of display outputs is irrelevant.

The RTX A1000 Mobile wins in practicality and integration, not in raw performance. Its 60 W TDP is 5.3x lower than the CMP 90HX’s 320 W. It requires no power connectors, fits as an IGP, and uses a PCIe 4.0 x8 interface, which is more modern and flexible than the CMP 90HX’s PCIe 1.0 x4. It also has display outputs (portable device dependent), whereas the CMP 90HX has none. For a mobile workstation or a compact embedded system, the RTX A1000 Mobile is the only viable option. The CMP 90HX’s 285 mm length and dual-slot design simply will not fit in such platforms.

The data also shows the RTX A1000 Mobile has a Vulkan score of 46782, which is lower than its OpenCL score of 48703. This suggests it may be less efficient in Vulkan workloads, but without a CMP 90HX Vulkan score, no cross-card conclusion can be drawn. The CMP 90HX is a specialized mining GPU with no display outputs and a restricted PCIe interface, so its use case is narrow: maximum compute throughput in a dedicated rig. The RTX A1000 Mobile is a general-purpose mobile GPU with broader compatibility, but its performance ceiling is far lower.

In summary, the CMP 90HX wins on every benchmark metric and every core specification. The RTX A1000 Mobile wins on power efficiency, form factor, and system integration. The choice depends entirely on whether the priority is raw compute or platform compatibility. The recorded data does not support any other conclusion.