Intel · Gaudi · Layer 2
Intel Gaudi 3
Buyers who want standard Ethernet networking instead of proprietary interconnect, and a genuinely lower acquisition price.
buy You can purchase this outright.
Specifications
- Memory
- 128 GB HBM2E
- Bandwidth
- 3.7 TB/s
- Compute
- Gaudi-generation BF16/FP8
- Power
- 600 W
- Interconnect
- Integrated RoCE over Ethernet
- Software
- SynapseAI, PyTorch
- Form factor
- accelerator
- Workload
- both
Verified 2026-09-09. Fields reading “not published” are exactly that — we do not estimate a figure a vendor withholds.
The catch
The smallest software ecosystem of the major options. Intel's roadmap for this line has also been the least predictable, which is a procurement risk independent of the silicon.
The economics
Priced to undercut. The saving is real; the engineering time to make your stack work on it is the cost nobody quotes.
Cost per million tokens is the metric that decides most real purchases, and there is no neutral benchmark for it — every published figure comes from a company selling one side of the comparison. Read all of them, ours included, as claims rather than measurements.
The power question underneath this
A rack of frontier accelerators draws 120–200 kW against a 2026 average of about 27 kW, and the US grid interconnection queue exceeds 2,600 GW with waits approaching five years. Whether you can energise this part is now a harder question than whether you can buy it.
Layer 1 — Energy · The interconnection queue · Rack power density · Tokens per watt
Where you actually rent this
Choosing the part is the easy half. The same accelerator rents for wildly different money depending only on who you rent it from — the cheapest and dearest published on-demand price for an H100 sit 5.6× apart — and the provider, not the chip, is what decides whether you can leave later.
Compared against
- Intel Gaudi 3 vs NVIDIA H100 — Standard Ethernet and a lower price against the deepest software ecosystem in computing.Compare with NVIDIA H100
Related silicon
- NVIDIA RTX PRO 6000 Blackwell — 96 GB GDDR7, buy
- Tenstorrent Blackhole — GDDR6 rather than HBM, buy
- AMD Helios (Instinct MI455X) — 31 TB of HBM4 per rack across 72 MI455X GPUs; 432 GB of HBM4 per GPU., rent
- AMD Instinct MI300X — 192 GB HBM3, rent
- AMD Instinct MI350X — 288 GB HBM3E, rent
- AMD Instinct MI355X — 288 GB HBM3E, rent
Sources
Want this whole layer as one table?
The accelerators on these pages as a single CSV — memory, bandwidth, compute, power, interconnect, software stack, availability, who each part is for and the catch — ready to sort, filter or load into a notebook.
Get the silicon table — free for research and evaluation
Four short lines and the download starts at once (CC-BY-NC-4.0, built from these pages the moment you ask). What you type is used only to answer the request and is never sold. Prefer Hugging Face? Request it there.
Using it inside a product, a client report or anything you sell? That needs a commercial licence.