그래픽 카드

NVIDIA L4 24GB GDDR6 Tensor Core GPU (Refurbished)

정가 ₩5,705,000
제품 정보로 건너뛰기
1 1
그래픽 카드

NVIDIA L4 24GB GDDR6 Tensor Core GPU (Refurbished)

NVIDIA L4 24GB GDDR6 Tensor Core GPU (Refurbished)

정가 ₩5,705,000
재고 1개
세금이 포함됩니다. 배송료는 결제 시 계산됩니다.
재고 1개
특급 배송 가능.
귀하의 국가에서는 배송 시 수입 VAT 또는 수수료를 부과할 수 있습니다.
1년 보증 포함됨.
전체 세부 정보 보기

개요

Condition: Refurbished, fully tested and working perfectly. Restored to like-new condition. Original box and accessories not included.

Warranty: 1 year return-to-base warranty included. 14-day returns accepted.

The NVIDIA L4 Tensor Core GPU delivers a versatile platform for AI, graphics and video processing in the cloud and at the edge. Built on the NVIDIA Ada Lovelace architecture with 24GB of GDDR6 memory, it provides breakthrough multi-precision performance for deep learning training and inference, video transcoding, AI audio and video effects, rendering, data analytics, virtual workstations and virtual desktops.

Remarkably efficient, the L4 packs all of that capability into a passively cooled, single-slot, low-profile card with a maximum power envelope of just 72W, drawing everything it needs from the PCIe slot with no auxiliary power connector required. It slots into nearly any server, from mainstream enterprise racks to space- and power-constrained edge deployments, and handles challenging ambient environments with ease.

Dedicated hardware video engines, with two encoders and four decoders on board, make the L4 exceptional for video-heavy AI pipelines, delivering dramatic gains in AI video performance and energy efficiency over CPU-based infrastructure. As part of NVIDIA AI, it supports all major AI frameworks and neural network models.

구성품

Refurbished NVIDIA L4 24GB Tensor Core GPU



Original box and accessories are not included.

사양

Manufacturer

NVIDIA

Part Number

900-2G193-0000-000

GPU Architecture

NVIDIA Ada Lovelace

GPU Memory

24GB GDDR6

Memory Bus

192-bit

Memory Bandwidth

300 GB/s

Bus Interface

PCI Express 4.0 x16 (Gen4 x16, x8; Gen3 x16 supported)

GPU Clock

795 MHz base / 2040 MHz boost

FP32 Performance

30.3 TFLOPS

Video Engines

2x NVENC, 4x NVDEC, 4x JPEG decoders

Display Outputs

None (compute and virtualization accelerator)

Max Power Consumption

72W

Power Connector

None (powered by PCIe slot)

Thermal Solution

Passive

Form Factor

Half-height, half-length, single-slot (HHHL), low-profile and full-height brackets

vGPU Software Support

NVIDIA vGPU 15.2 or later

Board Weight

270g