Lenovo 30HH004BUS ThinkStation P8 AMD Threadripper PRO 7975WX 4.00GHZ 128MB Windows 11 Pro 64

Lenovo 30HH004BUS ThinkStation P8 AMD Threadripper PRO Workstation

Overview

The Lenovo ThinkStation P8 (30HH004BUS) is a tower workstation built around AMD's Threadripper PRO 7975WX — a 32-core, 64-thread processor with a 5.3 GHz boost clock and 128 MB of L3 cache. This configuration pairs that CPU with 128 GB of DDR5-4800 RAM, an NVIDIA RTX 5000 Ada Generation GPU with 32 GB of VRAM, and a 4 TB PCIe Gen4 M.2 NVMe SSD. It ships with Windows 11 Pro 64-bit and includes a 3-Year Premier Next Business Day on-site warranty. For professional workstations handling GPU-accelerated compute, AI inferencing, or multi-stream video analysis, this configuration sits at the high end of what a single-socket tower can deliver.

Key Features

• 32-Core AMD Threadripper PRO 7975WX: 32 physical cores and 64 threads with a 5.3 GHz boost frequency mean this processor handles highly parallelized workloads — video transcoding, 3D rendering, large-scale data processing — without the bottlenecks you hit on mainstream desktop CPUs. The 128 MB L3 cache keeps frequently accessed data close to compute, reducing latency on cache-sensitive workloads like physics simulation and ML training pipelines.
• 128 GB DDR5-4800 RAM (4 x 32 GB, 8-slot board): DDR5 at 4800 MHz delivers higher bandwidth than DDR4, which matters when feeding a 32-core CPU and a high-VRAM GPU simultaneously. The 4 x 32 GB configuration leaves four DIMM slots open, so you can scale to 256 GB or beyond — up to 1 TB maximum — without replacing existing modules. That headroom is worth considering if your workload is growing.
• NVIDIA RTX 5000 Ada Generation with 32 GB VRAM: 32 GB of GPU memory is the threshold that matters for large AI model inference, high-resolution 3D scene rendering, and multi-stream video analytics. The Ada Lovelace architecture includes dedicated hardware for ray tracing and tensor operations, making this relevant for any workflow that uses GPU-accelerated frameworks like CUDA, TensorRT, or OpenCL.
• 4 TB PCIe Gen4 TLC Opal SSD (M.2 2280): Gen4 NVMe delivers sequential read/write throughput roughly double that of Gen3, which is relevant when working with large project files, video datasets, or virtual machine images that must be loaded quickly into RAM or VRAM. The Opal specification means hardware-level self-encryption is available — relevant in regulated environments where data-at-rest protection is a compliance requirement.
• Up to 1 TB Maximum RAM: The Socket sTR5 platform and eight DIMM slots support up to 1 TB of installed memory. For workloads like in-memory databases, large-scale simulation, or multi-VM hosting, that ceiling matters. Most desktop platforms cap at 128–192 GB; this one does not.
• Wi-Fi 6E (RZ616 2x2 AX) + Bluetooth 5.1: Built-in Wi-Fi 6E handles high-throughput wireless connectivity without a PCIe adapter, preserving expansion slots for other cards. BT 5.1 covers peripheral connectivity at low power. Neither replaces wired Ethernet for server-facing traffic, but both reduce cable management complexity in mixed-use deployments.
• PCIe, USB, and Ethernet Interfaces: The ThinkStation P8 platform exposes PCIe expansion slots (for additional GPUs, capture cards, or accelerators), multiple USB ports for peripheral integration, and Gigabit or faster Ethernet for network connectivity. This breadth of interface coverage matters when integrating with existing network infrastructure or adding specialized hardware.
• Windows 11 Pro 64-bit: Ships with a full commercial OS license — not Home, not an OEM-restricted SKU. Windows 11 Pro includes BitLocker, Hyper-V, Remote Desktop host, and domain join. For enterprise deployment and software licensing compliance, this is the correct OS tier.
• 3-Year Premier Next Business Day Warranty: Premier NBD means on-site technician response the next business day after a confirmed hardware fault — not depot return, not mail-in. For a workstation in active production use, this is the support tier that minimizes downtime. Verify the coverage geography applies to your site before ordering.

Integration and Compatibility

The 30HH004BUS is positioned for integration into professional computing environments where both CPU thread count and GPU VRAM are simultaneously constrained resources. The PCIe expansion architecture on the Threadripper PRO sTR5 platform supports multi-GPU configurations and high-bandwidth add-in cards. For security and surveillance applications, this workstation-class hardware handles demanding VMS server roles — running GPU-accelerated analytics across many camera streams simultaneously — that exceed what entry NVR appliances can process. The DDR5 memory subsystem and Gen4 NVMe storage also make this a fit for AI-assisted video analytics software that benefits from fast model loading and large in-memory frame buffers.

Connectivity via PCIe, USB, and Ethernet means it integrates with standard enterprise rack and desktop environments without proprietary cabling. The Opal-compliant SSD supports software-defined encryption management tools common in enterprise IT environments. For teams evaluating the broader Lenovo workstation line, this configuration represents the high-core-count, high-VRAM tier of the ThinkStation P8 family — distinct from lower-core or NVIDIA A-series GPU configurations in the same platform.

Frequently Asked Questions

Q: What processor does the Lenovo ThinkStation P8 30HH004BUS use?

A: It uses the AMD Ryzen Threadripper PRO 7975WX — a 32-core, 64-thread processor on the Socket sTR5 platform, with a 5.3 GHz boost clock and 128 MB of L3 cache.

Q: How much RAM does the 30HH004BUS ship with, and can it be expanded?

A: It ships with 128 GB of DDR5-4800 SDRAM installed as 4 x 32 GB modules across an 8-slot DIMM board. The platform supports up to 1 TB maximum installed memory, leaving four slots open for future expansion without replacing existing modules.

Q: What GPU is included in this configuration?

A: The NVIDIA RTX 5000 Ada Generation with 32 GB of VRAM. This GPU uses the Ada Lovelace architecture with dedicated hardware tensor and ray-tracing cores, relevant for AI inferencing, GPU-accelerated analytics, and professional visualization workloads.

Q: What storage does the 30HH004BUS include?

A: A single 4 TB M.2 2280 PCIe Gen4 Performance TLC SSD with Opal (hardware encryption) support. Gen4 NVMe delivers approximately double the sequential throughput of Gen3, and the Opal spec enables hardware-level data-at-rest encryption.

Q: What warranty is included?

A: A 3-Year Premier Next Business Day on-site warranty is included. This covers on-site technician dispatch the next business day following a confirmed hardware fault — verify that your site location is within the covered service area.

Q: What operating system does it ship with?

A: Windows 11 Pro 64-bit — a full commercial license that includes BitLocker encryption, Hyper-V, Remote Desktop host capability, and domain join support.

Ted Perry

The 30HH004BUS sits at the intersection of high core count and high GPU VRAM — 32 cores at 5.3 GHz boost paired with 32 GB of Ada-generation GPU memory is a combination you don't see on mainstream workstations, and it's what makes this configuration relevant for GPU-accelerated AI analytics or dense VMS deployments rather than general office compute.

Technical Highlights:

• 128 MB L3 Cache on Threadripper PRO 7975WX: That cache size matters for workloads with large working sets — simulation, ML training, and video frame processing all benefit from keeping data resident near the cores rather than waiting on RAM round-trips.
• DDR5-4800 with 4 open DIMM slots: Ships at 128 GB but the board accommodates 8 DIMMs total. Upgrading to 256 GB or more is a drop-in operation — you're not replacing what's already there, just filling empty slots. The 1 TB ceiling gives this platform a long useful life as memory-hungry workloads grow.
• PCIe Gen4 NVMe + Opal Encryption: Gen4 bandwidth is approximately 7 GB/s sequential read on quality drives — fast enough that storage is rarely the bottleneck when loading large model weights or video datasets. The Opal self-encryption means BitLocker (included in Windows 11 Pro) can offload encryption to hardware, with no measurable throughput penalty.

Deployment Considerations:

• The Threadripper PRO sTR5 platform is a single-socket architecture — you get 32 cores in one socket, which is strong for most AI/analytics deployments, but this is not a dual-socket system. If your workload genuinely requires more than 32 cores and cannot be distributed across multiple nodes, plan accordingly.
• The RTX 5000 Ada is a professional-tier GPU but draws significant power. Verify your facility's circuit capacity and UPS sizing before deployment — a system with this GPU and CPU under load will draw substantially more than a standard office workstation.

This configuration is well-matched to a GPU-accelerated VMS analytics server role — handling deep learning inference across 20–50+ camera streams simultaneously — or as a primary workstation for engineering teams running CUDA-based simulation or large 3D scene rendering where 32 GB of VRAM eliminates out-of-core rendering bottlenecks.