Wiwynn states the accelerator, the count and the physical specification for this machine but publishes no compute throughput and no accelerator memory total, so neither is shown. The power figure Wiwynn gives is a power-supply rating rather than a measured draw, so it is not recorded as system power.
Wiwynn logo

Wiwynn ES100G2

rack
2× L4
Power
—
kW Total
Memory
—
GB Total

Wiwynn's short-depth 1U edge server, sold for edge AI inference and vRAN or NFV. A single 5th Gen Intel Xeon socket taking processors up to 225 W, two NVIDIA L4 accelerators, eight DDR5-4400/4800 DIMM slots of up to 128 GB each, two M.2 NVMe bays, and an OCP 3.0 NIC alongside two half-height half-length PCIe Gen5 x16 slots and one full-height half-length. The chassis is 43.5 by 438 by 420 mm and 10 kg, on an 800 W AC-DC supply. Wiwynn publishes no compute throughput and no accelerator memory total for it. Its description refers to two NVIDIA A4 accelerators, which is not a part NVIDIA sells; the specification table on the same page says L4, and that is what is recorded here.

We will point you at suppliers who have it. Free, and no signup.

System Details

GPU Configuration

GPU Model: NVIDIA L4 24GB
GPU Count: 2 GPUs
Architecture: Ada Lovelace
Interconnect: —

System Specifications

Form Factor: rack
Total Power: —
Total Memory: —
Memory Bandwidth: 600 GB/s

Powered by NVIDIA L4

This system utilizes 2 × NVIDIA L4 24GB GPUs, each delivering exceptional performance for AI and HPC workloads.

Per GPU TDP

72W

Per GPU Memory

24 GB

Process Node

4nm

Architecture

Ada Lovelace

Documentation & Resources

Flopper Spec Sheet

Wiwynn ES100G2 specifications, generated from our database. Printable.

Vendor product page

Wiwynn ES100G2 documentation

View ↗

Typical Use Cases

AI/ML Training
High-Performance Computing
Data Analytics

The Wiwynn ES100G2 runs 2× L4 GPUs.

© 2026 Flopper.io - Compare the hardware powering AI