Supported Operating Systems
Technical Documentation: Supported Operating Systems for the 'Aetheria X1' Server Platform
This document serves as the definitive technical reference for the operating system compatibility and configuration guidelines for the **Aetheria X1** High-Density Compute Node. This platform is designed for enterprise virtualization, large-scale database operations, and high-performance computing (HPC) workloads.
1. Hardware Specifications
The Aetheria X1 platform is engineered for maximum throughput and I/O density, utilizing the latest generation of server-grade silicon. All specifications listed below are based on the standard reference configuration (SKU: AX1-STD-2024Q3).
1.1 Core Processing Unit (CPU)
The platform supports dual-socket configurations utilizing the latest generation of Intel Xeon Scalable Processors (Sapphire Rapids architecture, specifically the 4th Gen Xeon Scalable series).
Parameter | Specification (Per Socket) | Notes |
---|---|---|
Processor Family | 4th Gen Intel Xeon Scalable (Sapphire Rapids) | Socket compatibility guaranteed for LGA 4677. |
Maximum Cores per Socket | 60 Cores | Total of 120 physical cores in dual configuration. |
Base Clock Frequency | 2.2 GHz (Config Dependent) | Varies based on SKU (e.g., Platinum 8480+). |
Max Turbo Frequency | Up to 3.8 GHz (Single Core) | Dependent on thermal headroom and power limits (TDP). |
L3 Cache | 112.5 MB (Shared) | Total of 225 MB L3 cache across the dual-socket system. |
Supported Instruction Sets | AVX-512, VNNI, AMX, DL Boost | Critical for HPC and AI workloads. |
PCIe Lanes | 80 Lanes (Per CPU) | Total 160 usable PCIe Gen 5 lanes for expansion. |
1.2 Memory Subsystem (RAM)
The Aetheria X1 features 32 DIMM slots, supporting both DDR5 ECC Registered (RDIMM) and Load-Reduced (LRDIMM) modules.
Parameter | Specification | OS Impact |
---|---|---|
Memory Type | DDR5 ECC RDIMM/LRDIMM | Requires OS kernel support for DDR5 memory controllers. |
Maximum Capacity | 8 TB (Using 256GB LRDIMMs) | Tested up to 4TB in standard deployments. |
DIMM Slots | 32 (16 per CPU) | Dual-channel configuration per CPU socket. |
Memory Speed Support | Up to 4800 MT/s (JEDEC Standard) | Higher speeds may require specific BIOS/UEFI settings. |
Memory Addressing | 64-bit architecture (PAE/PAE2 supported) | Full support for memory sizes exceeding 4GB is mandatory. |
Note: The memory controller relies heavily on UEFI firmware for initialization sequence integrity. Incompatible OS installations may fail to POST if memory training is unsuccessful.
1.3 Storage Controllers and Bays
The platform supports a flexible storage architecture, primarily focused on NVMe performance.
Interface | Quantity | Role/Protocol |
---|---|---|
M.2 NVMe (PCIe Gen 4/5) | 4 (Internal, Boot/OS Drive) | Dedicated slots for OS installation, typically configured in RAID1 or RAID10 via software or hardware controller. |
U.2 NVMe (PCIe Gen 4/5) | 8 Hot-Swappable Bays | Primary storage pool for high-speed data access. |
SATA III Ports | 8 (Via PCH) | Reserved typically for legacy HDD/SSD or optical drives. |
Hardware RAID Controller | Optional: Broadcom MegaRAID 9680-8i (or equivalent) | Supports RAID 0, 1, 5, 6, 10, 50, 60. |
1.4 Network Interface Controllers (NICs)
The integrated network capabilities are essential for high-throughput data movement, particularly critical for NIC drivers in operating systems.
Port | Speed | Controller Chipset |
---|---|---|
LOM (Baseboard) | 2 x 10GbE | Intel E810-XXV (Varies by revision) |
Expansion Slot (PCIe Gen 5) | Up to 4 x 25GbE or 2 x 100GbE | Dependent on installed PCIe adapter cards. |
1.5 Platform Firmware and Management
The system relies on a robust BMC implementation for out-of-band management.
- **BIOS/UEFI:** AMI Aptio V (or equivalent), supporting CSM (Compatibility Support Module) for legacy OS booting, though native UEFI mode is strongly recommended for modern OSes.
- **BMC:** ASPEED AST2600, supporting IPMI 2.0 and Redfish API v1.10.
- **Trusted Platform Module (TPM):** TPM 2.0 compliant, essential for OS features like Windows BitLocker and Linux dm-verity.
---
2. Performance Characteristics
Operating System choice directly impacts the platform's ability to utilize advanced hardware features such as AVX-512 and direct memory access (DMA) optimization. Benchmarks confirm optimal performance scaling across supported OS families.
2.1 CPU Feature Utilization
The performance heavily relies on the OS kernel's scheduler and driver quality to expose the CPU's capabilities.
- **AMX (Advanced Matrix Extensions):** Requires kernel support (Linux Kernel 5.18+ or Windows Server 2022+) to expose these instructions for AI inferencing acceleration.
- **PCIe Gen 5:** Requires OS driver stacks capable of negotiating Gen 5 link speeds. Older OSes (e.g., RHEL 7) may negotiate down to Gen 3/4 speeds, severely impacting storage and network throughput.
2.2 Benchmark Results (VM Density)
The following table summarizes typical performance metrics for virtualization workloads, where the OS configuration (Hypervisor vs. Bare Metal) plays a crucial role.
Configuration | Hypervisor OS | Guest OS (Average) | VMs Supported (vCPU/Core Ratio 1.5:1) | Latency (Avg. Microseconds) |
---|---|---|---|---|
Aetheria X1 (Dual 60C) | VMware ESXi 8.0 U2 | Windows Server 2022 | 180 | 15.2 µs |
Aetheria X1 (Dual 60C) | KVM (RHEL 9.3) | Ubuntu 24.04 LTS | 195 | 14.9 µs |
Aetheria X1 (Dual 60C) | Microsoft Hyper-V (Server 2022) | Windows Server 2022 | 175 | 16.1 µs |
The KVM configuration demonstrates slightly higher density due to the mature integration of the KVM scheduler with the Sapphire Rapids architecture.
2.3 Storage I/O Throughput
Storage performance is bottlenecked by the slowest component, which is often the OS storage stack (e.g., Windows Storage Spaces vs. Linux mdadm/LVM).
OS/Driver Stack | Sequential Read (GB/s) | Random Read IOPS (4K QD32) | Latency (Read ms) |
---|---|---|---|
Windows Server 2022 (Native Drivers) | 28.5 | 13.1 Million | 0.045 ms |
RHEL 9.3 (NVMe-CLI) | 31.2 | 14.5 Million | 0.039 ms |
VMware ESXi 8.0 (VMFS 6) | 29.8 | 12.9 Million | 0.051 ms |
The superior handling of asynchronous I/O queues by modern Linux kernels often results in higher raw IOPS figures compared to the Windows stack, provided the appropriate I/O scheduler (e.g., none/mq-deadline) is selected.
---
3. Recommended Use Cases
The Aetheria X1 configuration is optimized for environments demanding high core count, massive memory capacity, and cutting-edge I/O speed.
3.1 Enterprise Virtualization and Cloud Infrastructure
The 120-core density, combined with high RAM capacity (up to 8TB), makes this platform ideal for hosting large-scale Virtual Machines (VMs) or containers.
- **Recommended OS:** VMware ESXi 8.x, Microsoft Windows Server 2022 (Hyper-V role), or RHEL/Rocky Linux with KVM.
- **Rationale:** These hypervisors have validated support for the platform's Intel VMX extensions and provide robust management layers for resource allocation across 120 physical cores.
3.2 High-Performance Computing (HPC)
The extensive PCIe Gen 5 lanes (160 total) allow for the integration of multiple high-speed accelerators (GPUs/FPGAs) and high-bandwidth networking (InfiniBand or 100GbE).
- **Recommended OS:** Specialized Linux distributions (e.g., SUSE HPC, CentOS Stream) utilizing the latest kernels (6.x+) for optimal NUMA awareness and topology mapping.
- **Key Requirement:** The OS must correctly map memory channels to the appropriate CPU socket to avoid cross-socket latency penalties, especially when running MPI jobs.
3.3 Large-Scale Database Systems (OLTP/OLAP)
Databases like Oracle, SQL Server, and massive PostgreSQL/MySQL instances thrive on high core counts and fast, low-latency storage access.
- **Recommended OS:** Windows Server 2022 Datacenter (for SQL Server) or RHEL/Oracle Linux (for Oracle DB).
- **Tuning:** OS configuration must prioritize minimizing context switching and ensuring large memory pages (HugePages) are properly configured within the OS kernel to prevent TLB misses across the large physical memory footprint.
3.4 AI/Machine Learning Training
While specialized GPU servers exist, the Aetheria X1 serves as an excellent host for inference workloads or smaller-scale model training, leveraging the built-in AMX instructions.
- **Recommended OS:** Ubuntu 24.04 LTS or RHEL 9.x, ensuring the NVIDIA CUDA Toolkit and associated drivers are fully compatible with the kernel version.
---
- 4. Supported Operating Systems Matrix
This section details the officially validated operating systems, categorized by vendor and version. Compatibility is defined by successful installation, driver availability for all integrated hardware components (NICs, RAID, BMC), and sustained performance meeting defined SLAs.
4.1 Microsoft Windows Server Family
Windows Server support relies heavily on the availability of the latest chipset and network drivers from the OEM/Intel.
Version | Installation Method | Key Driver Support Status | UEFI/Secure Boot Support | Recommended Use Case |
---|---|---|---|---|
Windows Server 2025 (Insider/Preview) | UEFI Native | Full (Requires Beta Drivers) | Yes | Testing/Early Adoption |
Windows Server 2022 (LTS) | UEFI Native (Recommended) / Legacy BIOS | Full (Stable Drivers) | Yes | Virtualization, Database |
Windows Server 2019 | UEFI Native / Legacy BIOS | Full (Requires Driver Pack 3.0+) | Limited (TPM 2.0 issues without patching) | Legacy Workloads |
Note on Windows Installation: For clean UEFI installs, the installation media must include drivers compatible with the Intel C741 Chipset PCH, particularly for the NVMe boot devices. CSM must be disabled in the BIOS for optimal performance and security features (like VBS).
4.2 Linux Distributions
Linux support is generally broader due to the open-source nature of kernel development, but specific hardware features (like AMX) require newer kernel releases.
Distribution/Version | Kernel Version (Minimum) | Key Hardware Support Notes | Features Enabled |
---|---|---|---|
Red Hat Enterprise Linux (RHEL) 9.x | 5.14+ | Full support for PCIe Gen 5 and E810 NICs. | AMX, SR-IOV, DPDK |
Red Hat Enterprise Linux (RHEL) 8.x | 4.18+ | Requires backported drivers for optimal DDR5 performance. | SR-IOV (Limited) |
Ubuntu Server 24.04 LTS | 6.8+ | Excellent KVM integration; strong community support for new hardware. | Full AMX, eBPF acceleration |
Ubuntu Server 22.04 LTS | 5.15+ | Stable, but CPU topology mapping may require manual tuning for high core counts. | Standard Hypervisor Load |
SUSE Linux Enterprise Server (SLES) 15 SP6+ | 6.x Series | Optimized for SAP workloads; excellent NUMA balancing. | SAP HANA Certification Path |
4.3 Virtualization Hypervisors
These platforms are primary targets for bare-metal hypervisors.
Hypervisor | Minimum Version | Hardware Virtualization Support | Notes on I/O Passthrough (VT-d) |
---|---|---|---|
VMware ESXi | 8.0 Update 2 | Full VT-d/IOMMU support | Excellent SR-IOV performance for E810 NICs. |
Microsoft Hyper-V (Role) | Windows Server 2022 | Full Intel VMX support | Requires specific integration services for optimal performance. |
Proxmox VE (KVM/QEMU) | 8.2+ | Full IOMMU support | Requires updated QEMU binaries to fully expose all CPU features. |
4.4 Unsupported/Deprecated Operating Systems
The following operating systems are **not supported** and attempting installation will likely result in hardware initialization failure or severe performance degradation due to missing modern driver support for PCIe Gen 5, DDR5, and TPM 2.0.
- Windows Server 2016 and earlier.
- RHEL 7.x and earlier (Lacks necessary kernel features for modern CPU schedulers).
- Any 32-bit OS (The platform is strictly 64-bit capable).
---
- 5. Maintenance Considerations
Proper maintenance is crucial for sustained performance, especially given the high power density and core count of the Aetheria X1. OS configuration plays a role in managing thermal envelopes and power states.
5.1 Power and Thermal Management
The dual-socket configuration, when fully loaded with high-speed RAM and multiple PCIe Gen 5 expansion cards, can draw significant power.
- **TDP Envelope:** The system supports CPUs up to 350W TDP each. A fully loaded system can approach 1500W peak draw.
- **OS Power Management:** Modern OSes (Windows Server 2022, RHEL 9+) utilize Intel Speed Shift and ACPI states effectively. Disabling C-states or P-states in the OS scheduler (e.g., setting power profile to 'High Performance' constantly) will prevent thermal throttling but increase idle power consumption significantly.
- **Cooling Requirement:** Requires a minimum of N+1 redundant cooling capacity rated for at least 2.5 kW per rack unit. The BMC monitors multiple thermal zones, and the OS should expose these readings via appropriate monitoring agents (e.g., `lm-sensors` on Linux).
5.2 Firmware and Driver Updating
Maintaining synchronization between the UEFI and the OS driver stack is paramount for stability.
1. **UEFI Update:** Always update the UEFI firmware *before* installing or upgrading the OS. New firmware often introduces fixes for memory training, PCIe enumeration timing, and secure boot validation, which the OS installer relies upon. 2. **Driver Installation Order (Windows):** Chipset -> Storage (RAID/HBA) -> Network -> BMC/Management. 3. **Driver Installation Order (Linux):** Kernel modules (often integrated), followed by proprietary vendor drivers (e.g., Mellanox OFED, specific storage drivers) if not included in the mainline kernel.
5.3 Licensing and Virtualization Overhead
Operating system licensing structures (especially Microsoft Server licenses based on physical cores) must be factored into the cost model, as this hardware has 120 physical cores.
- For virtualization environments, licensing for the hypervisor (VMware, Microsoft) and the guest OS must account for the total physical core count of the host machine, regardless of the number of active VMs.
5.4 BMC and Out-of-Band Management
The BMC (AST2600) provides essential remote access. Ensure the OS recognizes the BMC management interface (often via specific PCIe/SMBus bridges).
- **Redfish API:** Modern OSes should be configured to utilize the Redfish API for telemetry, allowing centralized monitoring platforms to gather detailed hardware health data independent of the main OS health status. This is crucial for predictive maintenance strategies.
---
- 6. Comparison with Similar Configurations
To contextualize the Aetheria X1's positioning, we compare it against a previous generation platform (Aetheria X0, based on Ice Lake) and a denser, lower-power alternative (Aetheria S1, single-socket optimized).
6.1 Feature Set Comparison Table
Feature | Aetheria X1 (Current) | Aetheria X0 (Previous Gen) | Aetheria S1 (Single Socket) |
---|---|---|---|
CPU Architecture | Sapphire Rapids (4th Gen Xeon) | Ice Lake (3rd Gen Xeon) | Sapphire Rapids (4th Gen Xeon) |
Max Cores (Dual System) | 120 | 80 | 60 (Single Socket Max) |
Memory Speed (Max) | DDR5-4800 MT/s | DDR4-3200 MT/s | DDR5-4800 MT/s |
PCIe Generation | Gen 5.0 (160 Lanes) | Gen 4.0 (128 Lanes) | Gen 5.0 (80 Lanes) |
Integrated Storage Protocol | NVMe 1.4/1.5 Capable | NVMe 1.3 Capable | NVMe 1.4/1.5 Capable |
Instruction Set Focus | AMX, AVX-512 | AVX-512 | AMX, AVX-512 |
6.2 OS Driver Dependency Comparison
The generational leap in PCIe and memory technology significantly impacts OS compatibility requirements.
OS Component | Aetheria X1 (Gen 5/DDR5) | Aetheria X0 (Gen 4/DDR4) |
---|---|---|
Network Driver (E810) | Requires Kernel 5.x+ or Windows 2019+ with latest drivers. | Supported by older kernels (4.x). |
Storage Driver (NVMe) | High dependency on in-box NVMe driver supporting PCIe Gen 5 enumeration and power states. | Lower dependency; Gen 4 is well-established. |
UEFI/Secure Boot | Mandatory for full feature set (TPM 2.0 interaction). | Optional; Legacy BIOS support is robust. |
NUMA Awareness | Critical due to complex memory interleaving schemes (HBM/CXL potential in future revisions). | Important, but simpler 2-socket topology. |
The Aetheria X1 mandates modern operating systems (Server 2022/RHEL 9/Ubuntu 22.04+) to unlock its performance potential, whereas the X0 platform offered broader compatibility with slightly older, established OS versions.
---
Conclusion on OS Support
The Aetheria X1 server platform is a state-of-the-art compute node requiring operating systems released from **mid-2021 onwards** to ensure full access to hardware features, including AVX-512, TME, and high-speed PCIe Gen 5 I/O. While older OSes might boot using the Compatibility Support Module (CSM), performance will be severely degraded, and support for the 120-core topology will be suboptimal.
For mission-critical deployments, **VMware ESXi 8.0 U2** or **RHEL 9.3** are the most thoroughly validated operating environments for this hardware.
Intel-Based Server Configurations
Configuration | Specifications | Benchmark |
---|---|---|
Core i7-6700K/7700 Server | 64 GB DDR4, NVMe SSD 2 x 512 GB | CPU Benchmark: 8046 |
Core i7-8700 Server | 64 GB DDR4, NVMe SSD 2x1 TB | CPU Benchmark: 13124 |
Core i9-9900K Server | 128 GB DDR4, NVMe SSD 2 x 1 TB | CPU Benchmark: 49969 |
Core i9-13900 Server (64GB) | 64 GB RAM, 2x2 TB NVMe SSD | |
Core i9-13900 Server (128GB) | 128 GB RAM, 2x2 TB NVMe SSD | |
Core i5-13500 Server (64GB) | 64 GB RAM, 2x500 GB NVMe SSD | |
Core i5-13500 Server (128GB) | 128 GB RAM, 2x500 GB NVMe SSD | |
Core i5-13500 Workstation | 64 GB DDR5 RAM, 2 NVMe SSD, NVIDIA RTX 4000 |
AMD-Based Server Configurations
Configuration | Specifications | Benchmark |
---|---|---|
Ryzen 5 3600 Server | 64 GB RAM, 2x480 GB NVMe | CPU Benchmark: 17849 |
Ryzen 7 7700 Server | 64 GB DDR5 RAM, 2x1 TB NVMe | CPU Benchmark: 35224 |
Ryzen 9 5950X Server | 128 GB RAM, 2x4 TB NVMe | CPU Benchmark: 46045 |
Ryzen 9 7950X Server | 128 GB DDR5 ECC, 2x2 TB NVMe | CPU Benchmark: 63561 |
EPYC 7502P Server (128GB/1TB) | 128 GB RAM, 1 TB NVMe | CPU Benchmark: 48021 |
EPYC 7502P Server (128GB/2TB) | 128 GB RAM, 2 TB NVMe | CPU Benchmark: 48021 |
EPYC 7502P Server (128GB/4TB) | 128 GB RAM, 2x2 TB NVMe | CPU Benchmark: 48021 |
EPYC 7502P Server (256GB/1TB) | 256 GB RAM, 1 TB NVMe | CPU Benchmark: 48021 |
EPYC 7502P Server (256GB/4TB) | 256 GB RAM, 2x2 TB NVMe | CPU Benchmark: 48021 |
EPYC 9454P Server | 256 GB RAM, 2x2 TB NVMe |
Order Your Dedicated Server
Configure and order your ideal server configuration
Need Assistance?
- Telegram: @powervps Servers at a discounted price
⚠️ *Note: All benchmark scores are approximate and may vary based on configuration. Server availability subject to stock.* ⚠️