| Quantity | Unit price |
|---|---|
| 20–49 | $9,800.00 |
| 50–74 | $9,450.00 |
| 75+ | $9,100.00 |
Need more than the quantities above? Request a bulk quote
We source this item through our supplier network and do not hold inventory. Price and availability are confirmed with you before any order is accepted.
Questions about this part? Talk to us.
Phone +1 872 293 4196 · Email supply@allchipsupply.com
Location: USA · Shipping: Worldwide Shipping
The Intel Gaudi 3 HL-325L is Intel's third-generation Gaudi AI accelerator in the OAM form factor, aimed at training and inference of large models. It carries 128 GB of HBM2e memory with 3.7 TB/s of bandwidth, and its compute is built from 64 tensor processor cores and 8 matrix multiplication engines.
A distinctive feature is networking: 24 integrated 200 GbE RoCE ports allow scale-out over standard Ethernet instead of a proprietary interconnect. The HL-325L is the air-cooled version with a maximum power of 900 W, and it installs in Gaudi 3 eight-accelerator server platforms. Confirm the server platform and module count when you request a quote.
| Memory | 128 GB HBM2e |
|---|---|
| Memory bandwidth | 3.7 TB/s |
| Networking | 24x 200 GbE RoCE ports (integrated) |
| Form factor | OAM module (HL-325L) |
| Matrix engines | 64 tensor processor cores, 8 MMEs |
| Max power | 900 W (air cooled) |
Specifications are summarized from manufacturer and listing information; confirm against the manufacturer datasheet before purchase.
Manufacturer datasheet / product page (opens the manufacturer's website)
Hopper, 80 GB HBM3, 3.35 TB/s, SXM5 for HGX H100 baseboards
AMD CDNA 3 · 192 GB HBM3 · OAM module
NVIDIA Hopper · 80 GB HBM2e · 2.0 TB/s
NVIDIA Hopper · 141 GB HBM3e · 4.8 TB/s
