SC102 — 128-Core CFD HPC Node with 512 MB Cache and 12-Channel Memory for ANSYS Fluent

The SC102 node condenses a small cluster into one socket. It runs a single AMD EPYC 9755 — 128 cores and 256 threads, every one of them at 4.1 GHz — backed by 512 MB of L3 cache and twelve DDR5 memory channels. With 128 GB of RAM and 2 × 1.88 TB NVMe SSD, it delivers more total compute per hour than anything else in the lineup, and does it without a network sitting between the ranks.

A detail worth stating early: AMD rates this chip at 4.1 GHz for both single-core boost and all-core boost. Load all 128 cores and the clock doesn't fall — the speed you get on one core is the speed you get on every core.

Specifications

512 MB of L3 is by a wide margin the largest cache we offer, and the 4.1 GHz all-core figure matches our 48- and 96-core tiers — so the additional cores cost you nothing in clock speed. Twelve memory channels supply roughly 600 GB/s.

Sizing note: read this before choosing SC102

SC102 spreads 128 GB of RAM across 128 cores — about 1 GB per core, the tightest ratio in the lineup — and its 600 GB/s of bandwidth is the same figure our 64-core node carries, now shared among twice as many cores. Both facts lead to one conclusion: this node is built for running many jobs simultaneously, not for finishing a single job as quickly as possible.

A single Fluent case split into 128 MPI ranks sits well beyond where most CFD problems scale usefully. Fluent needs roughly 1 GB of RAM per million cells for a typical steady single-phase case, so a mesh sized to fit in 128 GB gives each of 128 ranks only a small share of cells — the point where exchanging boundary data between partitions begins to cost more than the solving. On that same case, a 24- or 32-core node frequently finishes first.

What SC102 does better than anything else here is concurrency: eight cases on 16 cores apiece, four on 32, or a DOE sweep processed in parallel batches. The aggregate throughput is unmatched; what differs from the smaller tiers is how that throughput gets distributed.

Have one very large case instead of many? Ask us before booking. Depending on the mesh, a 32- or 64-core tier may solve it faster.

What this node is built for

SC102 is a maximum-capacity throughput tier. It suits:

• Running many cases at once — the clearest use of 128 cores: a full DOE sweep, a mesh-independence study, or a batch of design variants, several running side by side on separate core groups.

• Teams sharing one machine — several engineers running their own cases at the same time without queuing behind each other.

• Very large cases that genuinely scale past 96 ranks — uncommon, but this is the node for them.

• Time-critical batch work — compressing a week of sequential runs into a day.

• Cluster-scale parallelism without a cluster — 128 ranks on one socket, no network hop, no MPI-over-network tuning.

• Heavy I/O batch runs — NVMe storage keeps up with many concurrent jobs writing data files and autosaves.

Choosing SC102

Choose SC102 when:

• You have many cases to run rather than one, or

• You have a genuinely enormous case, shown to scale past 96 ranks, that still fits within 128 GB.

For a single moderate case, almost any smaller tier serves you better. That's the honest positioning of this node, not fine print. Fluent scales well only while each core still has enough cells to stay busy relative to the data it exchanges with neighbouring partitions, and 128 ranks pushes past that threshold for most CFD work. The value here is aggregate throughput.

Tell us how many cases you're running and how big they are, and we'll tell you whether SC102 or a smaller tier gets your results sooner.

Fast NVMe storage

The two 1.88 TB NVMe SSDs offer far higher read/write throughput than SATA SSDs, plus nearly 3.76 TB of combined capacity. On a node running many jobs at once that matters more than usual: several cases writing data files and autosaves concurrently would saturate slower storage, and the capacity holds a full sweep's output without constant clearing.

How parallel solving works on this node

When you run Fluent in parallel, the solver partitions your mesh across whichever cores you assign it (using algorithms such as METIS), each core solving its share of cells and exchanging boundary data every iteration via MPI. On SC102 you'll often want to assign a subset — 16 or 32 cores per case — and run several cases alongside each other rather than stretching one across all 128. When you do commit the full core count to a single case, the single-socket layout is a genuine advantage: MPI traffic stays on fast internal channels instead of crossing a network, keeping overhead lower than the same rank count spread over separate machines.

Why rent instead of buy

A 128-core server is a large purchase that most CFD teams leave underused between projects. Renting it for the batch weeks and dropping to a smaller tier the rest of the time matches how the work actually arrives.

How it works — you're in full control

SC102 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 SC102 node.

2. Set up your simulations — load your geometry and cases, configure solver settings, and run ANSYS Fluent with all 128 cores at your command, on one case or several.

3. Run and monitor in real time — start, pause, adjust, and watch your solutions as they progress, 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 SC102 different from our 64- or 96-core nodes? More cores and considerably more cache, at the same 4.1 GHz all-core clock and the same total memory bandwidth. Because that bandwidth and the 128 GB of RAM are divided among more cores, SC102's edge lies in running more jobs at once rather than running one job faster.

Will 128 cores speed up my single case? Usually not. Most CFD cases stop scaling well long before 128 ranks, as inter-partition communication starts to outweigh the solving work. Send us your case and we'll check — a 32- or 64-core node is often the faster answer for one case.

So what is this node actually for? Volume. A DOE sweep, a batch of design variants, a mesh-independence study, or a team sharing one machine — several cases running concurrently on separate core groups, where 128 cores delivers throughput nothing else here can match.

Is 128 GB enough memory for 128 cores? For running multiple moderate cases concurrently, yes. For one very large case, memory becomes the binding limit before core count does. If your mesh needs more than 128 GB, our high-memory nodes are the right tier.

Does the clock drop when all 128 cores are working? No — AMD rates this chip at 4.1 GHz for both single-core and all-core boost. The clock on one core is the clock on all of them.

Is my data kept confidential? Yes.

Get started with SC102

If you have a batch of cases rather than a single one — a sweep, a study, or a team's worth of concurrent runs — SC102 gives you more total throughput than any other node in the lineup.

Contact MR CFD to run your ANSYS Fluent cases on SC102 — and we'll help you decide whether 128 cores or a smaller, faster-per-core tier gets your results sooner.