Why choose Intel® Server D50DNP Family for your AI solutions?
High-performance processing
- Built-in AI acceleration and In-memory analytics acceleration:
Support 5th/4th Gen Intel® Xeon® Scalable processors or Intel® Xeon® CPU Max Series Integrated High Bandwidth Memory (HBM) delivers up to 1TB bandwidth - Advanced GPU Technology:
Support for the Intel Data Center GPU Max Series (formerly Ponte Vecchio)
Ultra-fast memory and I/O to accelerate data movement
- Large capacity DDR5 memory (1DPC – 16 DIMMs @ 5600 MT/s for 5th Gen)
- Faster PCIe 5.0 support for high performance add-in cards
- High-performance PCI 5.0
- Compute Express Link (CXL) 1.1 for ultra-high-speed processor-to-device and processor-to-memory connectivity
Multiple Module Options and Efficient Cooling Solutions
- Compute Module:
The 1U, half-width form factor is available in air-cooled or liquid-cooled options. The liquid-cooled option maximizes compute density to enhance cooling efficiency. - Management Module:
This 2U, half-width configuration offers additional slots and storage. It is used for lightweight management of a full rack or functions as a 2-socket, air-cooled, high-TDP compute module. - Intel® Data Center GPU Max Series Accelerator Module:
This liquid-cooled 1U, full-width form factor module allows integration of up to 4 accelerators (or 8 accelerators per 2U chassis) to ensure incredible performance density. - PCIe 5.0 Accelerator Module:
In a 2U, full-width form factor, it supports up to 4 full-height, full-length, double-width PCIe 5.0 accelerators per air-cooled unit to deliver ultra-dense performance for HPC and AI workloads.
Built-In Security Enhancements
- Intel® Crypto Acceleration:
Speeds up encryption with built-in accelerators for AES, SHA, and RSA/DH protocols. - Intel® Software Guard Extensions (Intel® SGX):
Enhances security for sensitive applications with secure enclaves. - Intel® Total Memory Encryption (Intel® TME):
Protects data and VMs with full memory encryption. - Intel® Platform Firmware Resilience (Intel® PFR):
Detects and restores firmware to a known-good state for improved security.
Explore the Intel® Server D50DNP Family
The Intel® Server D50DNP Family offers a variety of modules designed for tailored HPC and AI infrastructure optimization.
Features of D50DNP
- Processor:
- 2 5th/4th Gen Intel Xeon Scalable processors or Intel Xeon CPU Max Series
- Up to 250W or 385W
(Support for CPU with 385W TDP in liquid-cooled configuration is under validation.)
- Max Memory:
- (8)+(8) DDR5 DIMMs slots @ 5600(5th Gen)/4,800(4th Gen) MT/s
- Networking:
- (1) 10GBase-T LAN + (1) 1000Base-T dedicated IPMI ports
- Up to 400Gbps network card
Standard air-cooled
1U Half-Width Compute module
D50DNP1MHCPAC
- Processor: Up to 250W
- Fixed Storage: Max 2 NVMe/SATA M.2 slots per module
- PCIe slot: 2 low profile PCIe Gen 5
- Chassis configuration: 2U 4 nodes, standard air-cooled
EVAC air-cooled
1U Half-Width Compute module
D50DNP1MHEVAC
- Processor: Up to 270W
- Fixed Storage: Max 2 NVMe/SATA M.2 slots per module
- PCIe slot: 2 low profile PCIe Gen 5
- Chassis configuration: 2U 4 nodes, EVAC air-cooled
Direct liquid-cooled
1U Half-Width Compute module
D50DNP1MHCPLC
- Processor: Up to 385W*
- Fixed Storage: Max 2 NVMe/SATA M.2 slots per module
- PCIe slot: 2 low profile PCIe Gen 5
- Chassis configuration: 2U 4 nodes, Direct-to-chip liquid-cooled
Standard air-cooled
2U Half-Width Management module
D50DNP2MHSVAC
- Processor: Up to 350W
- Fixed Storage: Max 2 NVMe/SATA M.2 slots per module
- Hot-swap Storage: 2 2.5" U.2 NVMe SSDs
- PCIe slot: 4 low profile PCIe Gen 5
- Chassis configuration: 2U 2 nodes, standard air-cooled
Direct liquid-cooled / Ponte Vecchio Support
1U Full-Width Intel® Data Center GPU Max Series accelerator module
D50DNP1MFALLC (EOL)
- Processor: Up to 385W*
- Fixed Storage: Max 2 NVMe/SATA M.2 slots per module
- PCIe slot: 2 low profile PCIe Gen 5
- Chassis configuration: 2U 2 nodes, Direct-to-chip liquid-cooled
Standard air-cooled / Ponte Vecchio Support
2U Full-Width PCIe 5.0 accelerator module
D50DNP2MFALAC
- Processor: Up to 350W
- Fixed Storage: Max 2 NVMe/SATA M.2 slots per module
- Hot-swap Storage: 2 2.5" U.2 NVMe SSDs
- PCIe slot: 4 low profile PCIe Gen 5, 4 full-height full-length double-width cards
- Chassis configuration: 2U one node, standard air-cooled
Even higher performance with Intel® Data Center GPU Max Series
Intel® Data Center GPUs are designed to deliver high performance and efficiency across a variety of data center workloads. The Intel® Data Center GPU Max Series (formerly Ponte Vecchio) offers the highest performance and density. With support from the Intel® Data Center GPU Max Series, the MiTAC Intel Server D50DNP1MFALLC (EOL) and D50DNP2MFALAC acceleration modules deliver enhanced performance for demanding modern AI workloads.
The Intel Server System D50DNP1MFALLC (EOL) Acceleration Module supports up to 4 units of Intel Data Center GPU Max 1550 within a 1U sled, delivers robust and outstanding parallel computing performance compared to competition, the Intel Server System D50DNP2MFALAC Acceleration Module supports up to 4 units of Intel Data Center GPU Max 1100, spot lighting an exemplary high-density design with multi-computing nodes within a full-width 2U sled.
- D50DNP1MFALLC supports 4 Intel Data Center GPU Max 1550
- D50DNP2MFALAC supports 4 Intel Data Center GPU Max 1100
Intel® Server Chassis FC2000 Family (Optimus Beach v2) for Intel® Server D50DNP Family
Scalability with front I/O, shared power and cooling designs
1U or 2U chassis with air cooled or liquid cooled solution depending on your selected modules
Start Your AI Journey Today!
SUBSCRIBE NOW to get the latest news.