SC89 — Dual-Socket 64-Core CFD HPC Node with 384 GB Memory and 6 TB NVMe Storage for ANSYS Fluent

The SC89 node brings massive parallel scale and exceptional memory capacity to our dual-socket lineup. Powered by two AMD EPYC 9354 processors, it delivers 64 cores and 128 threads with a 3.8 GHz boost. Memory capacity expands to an enormous 384 GB, and storage is driven by two 3.01 TB NVMe SSDs — giving you roughly 6 TB of lightning-fast usable capacity.

Two sockets provide dual memory controllers: 24 memory channels in total, roughly 1.2 TB/s of aggregate bandwidth, and a massive 512 MB of L3 cache across the pair. For CFD, that combination of extreme memory capacity, wide bandwidth, and high core count is the ultimate trifecta for solving large, complex models without compromise.

Specifications

Compared to standard dual-socket configurations, SC89 provides 64 physical cores alongside 384 GB of memory. Per core, that translates to exactly 6 GB of RAM — an exceptionally healthy ratio that ensures even when scaling across all 64 ranks, memory per rank will not bottleneck your solve.

What this node is built for

SC89 is the extreme-capacity and high-throughput tier. It suits:

• Massive meshes that require up to 384 GB — 384 GB holds colossal cell counts (well into the hundreds of millions), opening up enterprise-scale simulations that simply cannot be run on smaller machines.

• Bandwidth-bound solves — 24 memory channels across two sockets provide the widest memory path available, rapidly feeding data to all 64 cores.

• High-frequency transient saves — Dual NVMe storage offers massive sequential write speeds, preventing the solver from hanging while writing massive data files at every time step.

• Heavy physics on large cases — Multiphase, combustion, and moving mesh cases that demand intense compute scaling alongside huge memory pools.

• Cases that scale beautifully to 64 ranks — Enough physical cores to slash solve times on massive models.

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 SC89

Choose SC89 when:

• Your case needs more than 256 GB of memory, or

• You have a massive cell count that scales efficiently across 64 cores, or

• Your transient workflow demands the peak I/O speed of dual NVMe SSDs rather than just raw storage bulk.

If your priority is keeping months of output data natively on the drive and you don't need 384 GB of RAM, a massive-storage SATA tier might be a better fit. But if your bottleneck is keeping the solver fed with data, holding an immense mesh in memory, and writing massive chunks of transient data to disk instantly, SC89 is the undisputed choice.

As a rough guide, Fluent needs about 1 GB of RAM per million cells for a typical steady single-phase case. With 384 GB, you can theoretically open and run meshes approaching 300+ million cells — and with 64 high-speed cores, you have the parallel compute power to actually drive a mesh of that size.

Storage: peak speed with NVMe

Two 3.01 TB NVMe SSDs represent a distinct performance advantage. While standard SATA drives excel at cheap bulk capacity, NVMe drives are built for extreme I/O throughput. When you are running a 300-million cell case, each autosave file is massive. NVMe storage allows Fluent to dump those gigabytes of data to the drive in seconds rather than minutes, meaning your solver spends its time calculating physics, not waiting on disk writes.

How parallel solving works on this node

When you run Fluent in parallel, the solver partitions your mesh across all 64 cores (using algorithms such as METIS), each core solving its share of cells and exchanging boundary data every iteration via MPI. On a dual-socket node, most of that communication still stays on-board rather than crossing a network — a real advantage over two separate machines — while each socket's 12 memory channels feed its own cores from local memory. Sensible partitioning and rank placement keep cross-socket traffic to a minimum, which is what lets the doubled bandwidth actually show up in your solve times.

Why rent instead of buy

How it works — you're in full control

SC89 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 SC89 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 SC89 different from the SC99 node? SC89 scales up the compute and memory (64 cores and 384 GB RAM vs. 48 cores and 256 GB RAM), but trades SC99's massive 22 TB SATA storage array for 6 TB of much faster NVMe storage. If you need extreme mesh capacity and solve speed, choose SC89. If you are doing long parameter sweeps and need maximum storage bulk, choose SC99.

Does the second socket slow anything down? Only if the job is set up without regard for it. Cores reaching memory attached to the other socket pay a latency penalty, so partitioning and rank placement matter. We configure this as part of your setup to ensure maximum efficiency.

Why NVMe instead of SATA storage? Speed. Huge meshes generate huge result files. NVMe prevents the solver from pausing for long periods during transient autosaves by writing data at vastly higher sequential speeds than SATA.

How large a mesh can it handle? With 384 GB of RAM, plan for meshes comfortably reaching 300+ million cells for simpler steady single-phase cases, or fewer for heavy transient/multiphase physics (about 1 GB of RAM per million cells is a safe starting estimate). Send us your case for a precise answer.

Is my data kept confidential? Yes.

Get started with SC89

If your cases demand the absolute maximum in memory capacity, core count scaling, and high-speed NVMe storage, SC89 is built to push your simulations through faster.

Contact MR CFD to run your ANSYS Fluent case on SC89 — and we'll help you provision the ultimate compute node for your largest simulation workflows.