Skip to main content

NVIDIA L2 (Entry-Level Ada Inference)

Overview

NVIDIA L2 is the entry-level inference card in the Ada Lovelace lineup, positioned below the L4. With 24GB GDDR6 memory and a TDP of just 50-75W, it fits in a single PCIe slot.

Targeted at edge servers, telecom operators, retail, and other scenarios requiring low-power AI inference.

Core Specifications

ItemSpec
ArchitectureAda Lovelace (AD102 cut-down)
ProcessTSMC 4N
CUDA Cores4,608
Tensor Cores144 (4th Gen)
RT Cores36 (3rd Gen)
Memory24 GB GDDR6
Memory Bandwidth384 GB/s (16 Gbps × 192-bit)
FP8 Tensor96 TFLOPS (sparse) / 48 TFLOPS (dense)
INT8 Tensor96 TOPS (dense) / 192 TOPS (sparse)
TDP50-75 W
Form FactorPCIe Gen4 ×16 single-slot / half-height half-length
Launch2024-Q4
Price$1,500-$2,000

L2 vs L4 vs T4 Comparison

MetricL2L4T4
ArchitectureAdaAdaTuring
CUDA Cores4,6087,6802,560
Memory24GB GDDR624GB GDDR616GB GDDR6
Bandwidth384 GB/s300 GB/s320 GB/s
FP8 Tensor (sparse)96 TFLOPS485 TFLOPSN/A
TDP50-75W72W70W
Form FactorSingle-slotSingle-slotSingle-slot

The L2's FP8 performance is about 20% of the L4 (96/485 sparse), but with a similar TDP → the L4 offers better performance-per-watt.

Use Cases

  • Edge server AI inference (5G MEC)
  • ✅ Telecom vRAN + AI convergence
  • ✅ Retail edge AI (video analytics)
  • ✅ Embedded data centers
  • ❌ Large model inference (use L40S/H100)
  • ❌ Training (lacks FP8 compute advantage)

Vendor Information

ItemDetail
VendorNVIDIA
Target MarketEdge servers, telecom, retail
Price$1,500-$2,000