Recommended Free Tools
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Intel’s June 2024 guidance put an eight-accelerator Gaudi 2 kit with a universal baseboard (UBB) at $65,000 and an eight-accelerator Gaudi 3 kit with a UBB at $125,000. These were modeling figures for system providers—not current retail prices or confirmed prices for complete servers. Gaudi 3 does have Ethernet built into each accelerator, but Intel’s announcement does not say that the kit price includes the host server, external switches, cabling, or a deployable cluster.
What the $65,000 and $125,000 figures refer to
Intel announced both figures at Computex on June 4, 2024. The configurations were eight Gaudi accelerators plus a universal baseboard, offered as kit guidance to system providers:
| Configuration | Intel’s June 2024 figure | What the figure describes |
|---|---|---|
| Gaudi 2, 8× OAM with UBB | $65,000 | Eight Gaudi 2 accelerators and a universal baseboard kit for system providers |
| Gaudi 3, 8× OAM with UBB | $125,000 | Eight Gaudi 3 accelerators and a universal baseboard kit |
OAM means Open Accelerator Module, a module form factor used for data-center accelerators. In this context, “8× OAM UBB” means eight OAM accelerators mounted on a universal baseboard; it does not by itself describe a complete server. Intel’s press-kit footnote says the prices are modeling guidance, tells buyers to consult an OEM for final pricing, and notes that results can vary with volumes and lead times. Intel’s Computex 2024 press kit
Does the kit price include networking?
Networking built into Gaudi 3
Yes, in the sense that Ethernet connectivity is integrated into the Gaudi 3 accelerator modules. Intel’s 2024 product announcement specifies 24 Ethernet ports rated at 200 gigabits per second per accelerator. Intel’s networking material describes how the design supports communication among accelerators and connections beyond the baseboard. Intel’s Gaudi 3 announcement
#1 Best Overall
- Graphics Card Interface: Pci E
What that does not establish
Integrated accelerator ports are not the same as a complete data-center network. The published kit-price announcement does not itemize external switches, cables, rack integration, host CPUs, system memory, storage, cooling, or deployment. Get the OEM’s bill of materials and quote to confirm what is included in a particular system.
What the Gaudi 3 UBB contributes
Intel identifies the Gaudi 3 HLB-325 as an eight-OAM baseboard. Its product brief describes direct all-to-all links among the eight accelerators, with 4.2 TB/s of bidirectional bandwidth for that on-board connectivity. That is a baseboard specification, not evidence that the announced kit price buys a complete host server or cluster. Intel Gaudi 3 HLB-325 product brief
Rank #2
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
How to evaluate a real Gaudi quote
The 2024 figures are useful historical reference points, but procurement decisions need a current, configuration-specific quote. Ask the OEM to state the exact hardware and services included, and compare systems on equivalent terms:
- Accelerator generation and count, and whether the product is an OAM/UBB system or a PCIe card.
- Host server components, including processors, system memory, storage, chassis, power, and cooling.
- Networking: accelerator ports, on-board topology, external switches and cables, and any required network integration.
- Software and model support for the intended training or inference workload.
- OEM support, delivery timing, region-specific availability, and final delivered price.
Intel’s 2024 press materials named Dell, HPE, Lenovo, Supermicro, ASUS, Foxconn, Gigabyte, Inventec, Quanta, and Wistron as companies planning Gaudi 3 systems. That announcement records plans at the time; it does not confirm current stock, configurations, or prices. Contact OEMs directly for current offerings. Intel’s Computex 2024 press kit
Rank #3
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
Do not confuse the UBB kit with the PCIe card
Intel also lists a Gaudi 3 PCIe Gen5 accelerator card for integration into compatible servers. That is a separate form factor and should not be treated as the eight-accelerator OAM/UBB configuration or priced as though the two products were interchangeable. Intel Gaudi product overview
Quick Recap
Rank #4
- Robust Design:Constructed to withstand high temperatures, the V100 16GB SXM2 card operates efficiently up to 105℃.
- Advanced Connectivity:Features a SXM2 connector for seamless integration with a wide range of systems, ensuring compatibility.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
