SC95 — Dual-Socket 64-Core High-Frequency CFD HPC Node with 256 GB Memory and 22.5 TB NVMe Storage for ANSYS Fluent

The SC95 node brings a rare and powerful combination to the dual-socket lineup: elite per-core clock speeds paired with an uncompromised enterprise storage array. Powered by two AMD EPYC 9335 processors, it delivers a highly focused 64 physical cores and 128 threads with an aggressive 4.4 GHz boost clock. Memory sits at a comfortable 256 GB, while the storage array features our flagship configuration: six 3.75 TB NVMe SSDs, yielding roughly 22.5 TB of ultra-high-speed usable capacity.

By deploying fewer cores per socket but clocking them significantly higher, SC95 offers a massive 24 memory channels feeding just 64 processing pipelines. For CFD workflows burdened by incredibly complex physics, demanding transient acoustics, or large-scale data writing, this node provides extreme bandwidth per core, blistering single-thread speed, and the storage capacity to hold months of output data without a bottleneck in sight.

Specifications

Compared to our high-core-count nodes (like the 128-core SC93), SC95 trades raw MPI width for brute-force per-core performance. With 256 GB of memory spread across 64 cores, it offers a very healthy 4 GB of RAM per core. This makes it an ideal platform for meshes that require deep physics calculations per cell, rather than just purely massive cell counts.

What this node is built for

SC95 is the high-frequency and zero-compromise I/O tier. It suits:

• Compute-bound, complex physics — Combustion, heavy multiphase, discrete phase models (DPM), and reacting flows require far more math per cell. High 4.4 GHz clock speeds chew through this overhead much faster than simply throwing more slow cores at it.

• Heavy transient and acoustic workflows — Large Eddy Simulation (LES), Detached Eddy Simulation (DES), and aeroacoustics where the solver generates massive data files every few iterations and absolutely cannot wait on slow disk writes.

• Bandwidth-bound solves — 24 memory channels across two sockets feeding only 64 cores means the memory bandwidth per core is exceptionally high, practically eliminating data starvation.

• Long transient campaigns with massive output — 22.5 TB of pure NVMe storage accommodates massive time-step data and full DOE sweeps natively on the drive with instantaneous write speeds.

Understanding the dual-socket layout

A two-socket machine is not simply a bigger one-socket machine. Each processor has its own memory controllers and its own local RAM, and a core reaching data attached to the other socket pays a latency penalty crossing the interconnect between them.

For Fluent, this is highly manageable and worth knowing about. Partitioning the mesh so each socket works mostly on locally-held data — and pinning MPI ranks accordingly — keeps that cross-socket traffic low. Run without any NUMA awareness and you can lose some of the bandwidth advantage the second socket provides. We configure this for you as part of the setup, so in practice it's a detail we handle rather than one you have to manage.

Choosing SC95

Choose SC95 when:

• Your solve is gated by complex physics that benefit more from a 4.4 GHz boost clock than from scaling past 64 cores, or

• You are running heavy transients that demand both the peak speed of NVMe and the massive 22+ TB capacity usually restricted to SATA drives, or

• You need exceptionally high memory bandwidth per core (24 channels feeding just 64 cores).

If your case scales beautifully to 128 or 192 cores and doesn't rely heavily on per-core clock speed, nodes like the SC93 or SC91 will finish the job faster. If your mesh requires more than 256 GB of RAM, look to our 384 GB tiers. But for pure per-core muscle combined with the most powerful storage array available, SC95 is unparalleled.

Storage: capacity and peak speed combined

Historically, CFD engineers running massive transients had to choose between the high capacity of SATA or the sheer speed of NVMe. SC95 eliminates the compromise entirely. Six 3.75 TB NVMe SSDs provide 22.5 TB of space with extraordinary sequential write speeds. For large transient runs where the solver generates gigabytes of data every few time steps, this array allows Fluent to write autosaves instantly, ensuring your 4.4 GHz cores stay focused on calculating physics rather than hanging on disk operations.

How parallel solving works on this node

When you run Fluent in parallel, the solver partitions your mesh across all 64 cores. On this node, that means each core is tackling a larger chunk of cells compared to higher-core nodes, but it processes them significantly faster due to the 4.4 GHz boost clock and the immense memory bandwidth (roughly 1.2 TB/s aggregate) provided by the 24 dual-socket memory channels. Sensible partitioning and rank placement keep cross-socket traffic to a minimum, ensuring your heavy physics simulations run at peak efficiency.

Why rent instead of buy

How it works — you're in full control

SC95 is yours to drive directly. You connect to the node over a secure remote desktop connection and run everything yourself, exactly as you would on a local workstation:

1. Connect via remote desktop — you get full remote access to the SC95 node.

2. Set up your simulation — load your geometry and case, configure your solver settings, and run ANSYS Fluent with all 64 cores at your command.

3. Run and monitor in real time — start, pause, adjust, and watch the solution as it progresses, with complete control throughout.

4. Keep your results — post-process on the node or download your solution and data files whenever you're ready.

You have full authority over the node for the duration of your rental: we provide the hardware and the support, and you run the simulations your way.

Frequently asked questions

How is SC95 different from the single-socket SC90? While both nodes feature high-frequency processors (4.4 GHz) and 256 GB of RAM, SC95 is a dual-socket node. It offers double the core count (64 vs 32), double the memory bandwidth (24 channels vs 12), and more than ten times the ultra-fast NVMe storage (22.5 TB vs 2 TB). SC95 is built for much larger, highly transient workflows.

How is SC95 different from the 128-core SC93? SC93 scales wider (128 cores at 4.2 GHz), while SC95 punches harder per core (64 cores at 4.4 GHz) and offers double the memory per core (4 GB vs 2 GB). If your mesh is moderately sized but physically complex (e.g., combustion), SC95's higher clock speed and memory-per-core ratio will likely outperform the higher core count of the SC93.

Is 4 GB of RAM per core enough? Yes, it is highly optimal. This ratio comfortably handles dense physics, moving meshes, and complex multiphase models without bottlenecking the system. You can comfortably load meshes up to roughly 200 million cells for standard steady-state runs, though it shines brightest on highly complex models in the 50–100 million cell range.

Is my data kept confidential? Yes.

Get started with SC95

If your simulations require the aggressive single-thread performance of a 4.4 GHz processor alongside a massive, uncompromised NVMe storage array to warehouse terabytes of transient data, SC95 is engineered specifically for your workflow.

Contact MR CFD to run your ANSYS Fluent case on SC95 — and we'll help you configure the ideal environment for your complex physics simulations.sc95