L40G: specs and benchmarks

VS

Summary

NVIDIA started L40G sales 13 October 2022. This is an Ada Lovelace architecture desktop card based on 4 nm manufacturing process and primarily aimed at designers. 24 GB of GDDR6 memory clocked at 2.25 GHz are supplied, and together with 384 Bit memory interface this creates a bandwidth of 864.0 GB/s.

Compatibility-wise, this is a dual-slot graphics card attached via PCIe 4.0 x16 interface. Its manufacturer default version has a length of 267 mm. 1x 16-pin power connector is required, and power consumption is at 300 Watt.

Primary details

Some basic facts about L40G: architecture, market segment, release date etc.

Place in the rankingnot rated
Place by popularitynot in top-100
ArchitectureAda Lovelace (2022−2024)
GPU code nameAD102
Market segmentWorkstation
Release date13 October 2022 (2 years ago)

Detailed specifications

L40G's specs such as number of shaders, GPU base clock, manufacturing process, texturing and calculation speed. These parameters indirectly speak of L40G's performance, but for precise assessment you have to consider its benchmark and gaming test results.

Pipelines / CUDA cores18176of 21760 (GeForce RTX 5090)
Core clock speed1005 MHzof 2610 MHz (Radeon RX 6500 XT)
Boost clock speed2475 MHzof 3599 MHz (Radeon RX 7990 XTX)
Number of transistors76,300 millionof 208,000 million (B200 SXM 192 GB)
Manufacturing process technology4 nmof 3 nm (Arc Graphics 140V)
Power consumption (TDP)300 Wattof 2400 Watt (Data Center GPU Max Subsystem)
Texture fill rate1,406of 2,554 (Radeon Instinct MI300X)
Floating-point processing power89.97 TFLOPSof 109.7 (GeForce RTX 5090)
ROPs192
TMUs568of 1280 (Data Center GPU Max NEXT)
Tensor Cores568of 1216 (Radeon Instinct MI300X)
Ray Tracing Cores142of 170 (GeForce RTX 5090)

Form factor & compatibility

This section provides details about the physical dimensions of L40G and its compatibility with other computer components. This information is useful when selecting a computer configuration or upgrading an existing one. For desktop graphics cards, it includes details about the interface and bus (for motherboard compatibility) and additional power connectors (for power supply compatibility).

InterfacePCIe 4.0 x16
Length267 mm
Width2-slot
Supplementary power connectors1x 16-pin

VRAM capacity and type

Parameters of memory installed on L40G: its type, size, bus, clock and resulting bandwidth. Note that GPUs integrated into processors have no dedicated memory and use a shared part of system RAM instead.

Memory typeGDDR6
Maximum RAM amount24 GBof 294912 (Radeon Instinct MI325X)
Memory bus width384 Bitof 8192 Bit (Radeon Instinct MI250X)
Memory clock speed2250 MHzof 20000 (RTX 5000 Ada Generation Mobile)
Memory bandwidth864.0 GB/sof 5,171 GB/s (Radeon Instinct MI300X)

Connectivity and outputs

Types and number of video connectors present on L40G. As a rule, this section is relevant only for desktop reference graphics cards, since for notebook ones the availability of certain video outputs depends on the laptop model, while non-reference desktop models can (though not necessarily will) bear a different set of video ports.

Display Connectors1x HDMI 2.1, 3x DisplayPort 1.4a
HDMI+

API compatibility

APIs supported by L40G, sometimes including their particular versions.

DirectX12 Ultimate (12_2)
Shader Model6.7
OpenGL4.6
OpenCL3.0
Vulkan1.3
CUDA8.9

Benchmark performance

Synthetic benchmark performance of L40G. The combined score is measured on a 0-100 point scale.



We have no data on L40G benchmark results.


AMD equivalent

According to our data, the closest AMD alternative to L40G is Radeon PRO W7900.

Similar GPUs

Here is our recommendation of several graphics cards that are more or less close in performance to the one reviewed.

All comparisons with L40G

Community ratings

Here you can see the user rating of the graphics card, as well as rate it yourself.


5 5 votes

Rate L40G on a scale of 1 to 5:

  • 1
  • 2
  • 3
  • 4
  • 5

Questions & comments

Here you can ask a question about L40G, agree or disagree with our judgements, or report an error or mismatch.