AMD Ryzen Threadripper PRO 9995WX 96-core Workstation Processor

The AMD Ryzen Threadripper PRO 9995WX is AMD's flagship workstation processor built on the Zen 5 microarchitecture, delivering 96 cores and 192 threads within the sTR5 platform. Fabricated on TSMC's 4nm process node, the processor integrates AMD's latest front-end and execution engine improvements, including wider instruction decode, enhanced branch prediction, and doubled AVX-512 throughput compared to Zen 4, making it exceptionally well-suited for compute-intensive professional workloads requiring both high single-threaded responsiveness and massive parallel throughput.

With support for eight-channel DDR5 memory and up to 2TB of registered ECC DDR5 capacity, the Threadripper PRO 9995WX provides the memory bandwidth and capacity headroom demanded by large AI model training, simulation datasets, and in-memory analytics workloads. The processor exposes up to 160 PCIe 5.0 lanes, enabling simultaneous high-bandwidth connectivity to multiple professional GPUs, NVMe storage arrays, and networking adapters without contention—a critical differentiator for workstation-class multi-accelerator configurations.

Designed as a direct fit for AMD WRX90 platform workstations, the 9995WX carries a 350W TDP that reflects its role as an uncompromising performance platform. The processor's deep integration with AMD PRO technologies supports enterprise manageability and security features relevant to regulated and high-assurance compute environments. As the Zen 5 generation flagship for single-socket workstations, the 9995WX targets next-generation AI development pipelines, computational fluid dynamics, structural simulation, high-resolution visual effects rendering, and genomic research workflows where core count, memory capacity, and I/O bandwidth all simultaneously constrain throughput.

Ideal for

• Large-scale AI and deep learning model training on-premises, leveraging 96 cores and wide AVX-512 execution for data preprocessing and CPU-side inference pipeline acceleration alongside multi-GPU configurations
• Computational fluid dynamics and finite element analysis simulations requiring high core counts, ECC memory integrity, and large addressable working sets up to 2TB DDR5
• High-resolution visual effects, 3D rendering, and real-time ray tracing production workloads in film, broadcast, and architectural visualization studios
• Genomics, proteomics, and bioinformatics pipelines processing large reference datasets in-memory, benefiting from eight-channel DDR5 bandwidth and ECC reliability
• Multi-stream real-time video transcoding and encoding at professional broadcast quality, where per-core performance and aggregate thread count directly determine throughput
• Software development and continuous integration build farms requiring maximum parallel compilation throughput within a single-socket workstation footprint

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.