AMCM has integrated PanOptimization's PanX thermomechanical simulation engine directly into EOSPRINT, according to a partnership announcement on PanOptimization's company site, folding pre-build thermal correction into the same software metal printer operators already use to prepare a job. Rather than treating simulation as a separate step run in outside software before a file is handed off to the printer, the integration lives inside the build-preparation pipeline itself — reading the machine's native build file, running the physics, and writing the corrections back before a single layer gets exposed.
The move targets a problem that gets worse, not better, as parts get bigger. AMCM's metal laser powder bed fusion systems span EOS's familiar M 290 platform up through AMCM's own large-format M 4K and M 8K machines, built to cover build volumes far beyond what a standard EOS system handles. Bigger build plates and taller parts mean more opportunities for heat to accumulate unevenly across a build, and that accumulated heat is what drives warping, residual stress, and outright part failure on tall or geometrically complex metal components.
What the Integration Actually Does
The mechanics, as described in reporting from VoxelMatters, are more specific than a typical vendor integration announcement. PanX reads AMCM's openjz build files — the native format EOSPRINT produces when a job is sliced and ready to print — and pulls geometry data, build-plate layout, processing parameters, and layer-by-layer processing time straight out of EOSPRINT through its API. That's the part that matters for anyone who has fought with simulation tools in the past: no manual re-export of STEP files, no rebuilding support structures in a separate program, no translating process parameters by hand between two pieces of software that don't talk to each other.
Once PanX has that data, it runs its thermomechanical model and writes results back into the build file: corrected dwell times between layers, compensated geometry to offset predicted distortion, and adjusted laser power where the simulation flags a risk of overheating. The corrected build comes back to EOSPRINT as a TIFF image stack paired with a Smart Fusion Replay file — formats the pipeline already expects, rather than a proprietary export that has to be manually reconciled. All of that happens before the build starts, not as a post-mortem after a part comes out warped or a recoater blade catches a raised feature mid-print.
PanX comes from PanOptimization, a simulation company co-founded by Erik Denlinger, who holds the title of chief engineer there. "AM is moving into a phase where simulation has to become a standard part of the manufacturing process," Denlinger said of the shift, cited by VoxelMatters. He added that manufacturers who want to print and qualify high-value metal AM parts with confidence "need physics-based models that help them understand and optimize what will happen before the build begins." The partnership gives AMCM customers access to that modeling without leaving the EOS software ecosystem — a meaningful distinction for shops that have standardized their workflow around EOSPRINT and don't want a second license, a second file format, or a second point of failure in the pipeline between CAD and laser.
Why AMCM Needed This
AMCM's pitch has always centered on scale — printing metal parts too large for a standard industrial LPBF machine to touch. That scale is exactly where thermal simulation stops being a nice-to-have and starts being load-bearing. Tobias Petzinger, an AMCM application specialist, put it plainly in comments cited by VoxelMatters: "Especially for the large and complex applications our customers manufacture on the AMCM M 4K and M 8K, thermomechanical simulation is a key tool to counteract overheating and thermally induced distortion." On a small M 290 build, a modest heat imbalance might cost you surface finish or a dimensional tolerance. On an M 8K build spanning a meter or more, the same imbalance can mean a part that cracks, delaminates, or simply doesn't fit its assembly once it cools.
PanOptimization's own materials point to two reference projects run through PanX in support of the AMCM partnership: a 765mm-tall Aerospike-style part and a separate 1.2-meter-tall component whose build was specifically optimized for dwell time. For the aerospike, PanOptimization says its Multi-Grid modeling approach handled a full-part mesh of more than 26 million nodes, solved in roughly 3.5 hours on a standard desktop workstation; the dwell-time optimization on the 1.2-meter component ran in about an hour on comparable hardware. Both are the kind of geometry — tall, thin-walled, thermally isolated from the build plate as height increases — that has historically been hardest to predict without full-scale simulation, and both sit squarely in AMCM's large-format wheelhouse rather than in the smaller M 290 class. In its own statement on the partnership, AMCM said it turned to PanOptimization "to push the boundaries of additive manufacturing," adding that PanX "allows us to model the entire AM process with exceptional resolution and performance — even for very large components."
What It Means for Makers
This integration is aimed at AMCM's customer base — production shops and service bureaus running EOS-derived metal systems at industrial scale — not at desktop FDM or resin users. But the underlying shift is one worth watching regardless of what's on your bench. Thermal simulation for metal AM has traditionally lived in standalone software: separate licenses, separate file formats, and a workflow where getting a corrected build file back into your slicer required manual re-import and re-verification. Bolting that simulation directly onto the build-preparation software, reading and writing the native file format automatically, removes an entire category of human error and turns a specialist's tool into something closer to a checkbox in the normal slicing workflow.
That pattern — simulation and correction happening inside the slicer rather than bolted on afterward — is the same direction desktop and prosumer slicing software has slowly been moving, from built-in support generation that accounts for overhangs to warping predictions for large FDM parts. For makers running multi-material or large-format polymer printers where warping and layer adhesion are already a headache, AMCM's approach is a useful data point on where automated thermal correction is headed as it works its way down from industrial metal printing toward more accessible hardware. It's also a reminder that as build volumes grow — whether in metal LPBF or in oversized FDM machines — thermal management stops being something you tune by feel and becomes something that needs to be modeled before the print even starts.
