Arc G3 vs H200 NVL

Primary details

GPU architecture, market segment, value for money and other general parameters compared.

Place in the ranking	not rated	not rated
Place by popularity	not in top-100	not in top-100
Architecture	Hopper (2022−2024)	Xe3-LPG (2026)
GPU code name	GH100	Panther Lake
Market segment	Workstation	Laptop
Release date	18 November 2024 (1 year ago)	1 June 2026 (less than a year ago)

Detailed specifications

General parameters such as number of shaders, GPU core base clock and boost clock speeds, manufacturing process, texturing and calculation speed. Real power consumption of some graphics cards can well exceed their nominal TDP, especially if overclocked.

Pipelines / CUDA cores	16896	1280
Core clock speed	1365 MHz	300 MHz
Boost clock speed	1785 MHz	2400 MHz
Number of transistors	80,000 million	no data
Manufacturing process technology	5 nm	3 nm
Power consumption (TDP)	600 Watt	25 Watt
Texture fill rate	942.5	96.00
Floating-point processing power	60.32 TFLOPS	6.144 TFLOPS
ROPs	24	20
TMUs	528	40
Tensor Cores	528	no data
Ray Tracing Cores	no data	10
L1 Cache	33 MB	640 KB
L2 Cache	50 MB	16 MB

Form factor & compatibility

Information on compatibility with other computer components. Useful when choosing a future computer configuration or upgrading an existing one. For desktop graphics cards it's interface and bus (motherboard compatibility), additional power connectors (power supply compatibility).

Interface	PCIe 5.0 x16	IGP
Length	267 mm	no data
Width	2-slot	no data
Supplementary power connectors	8-pin EPS	None

VRAM capacity and type

Parameters of VRAM installed: its type, size, bus, clock and resulting bandwidth. Integrated GPUs have no dedicated video RAM and use a shared part of system RAM.

Memory type	HBM3e	System Shared
Maximum RAM amount	141 GB	System Shared
Memory bus width	6144 Bit	System Shared
Memory clock speed	1593 MHz	System Shared
Memory bandwidth	4.89 TB/s	no data
Shared memory	-	+
Resizable BAR	+	-

Connectivity and outputs

This section shows the types and number of video connectors on each GPU. The data applies specifically to desktop reference models (for example, NVIDIA’s Founders Edition). OEM partners often modify both the number and types of ports. On notebook GPUs, video‐output options are determined by the laptop’s design rather than the graphics chip itself.

Display Connectors

No outputs

Portable Device Dependent

API and SDK support

List of supported 3D and general-purpose computing APIs, including their specific versions.

DirectX	N/A	12 Ultimate (12_2)
Shader Model	N/A	6.9
OpenGL	N/A	4.6
OpenCL	3.0	3.0
Vulkan	N/A	1.4
CUDA	9.0	-
DLSS	+	-

Pros & cons summary

---

Recency	18 November 2024	1 June 2026
Chip lithography	5 nm	3 nm
Power consumption (TDP)	600 Watt	25 Watt