Two jobs, one machine: a viewport that stays smooth when the scene gets dense, and render power that turns an overnight job into a lunch break. Most 3D artists start at FOCUS MAX. When the render queue is what holds up the week, FOCUS X and HYPERFOCUS take it from there.
4.9 Stars on GoogleFrom more than 4,200 customer reviews
3-Year WarrantyIndustry-leading full-system cover, extendable to 4
Lifetime SupportWe're here for as long as you own it
Award-Winning BuildersReliability proven by four years of awards
Built in AustraliaBuilt and tested by the people who support it
On-Site ServicingLess downtime — we come to youFour machines on one ladder. FOCUS MAX is one strong card for the artist seat. FOCUS X adds a second when the render queue is the bottleneck. HYPERFOCUS takes up to six cards when the machine is the farm. CREATOR COMPACT is fourteen litres for the modelling desk. All four can be configured before we build them.
Focus and Creator run the same performance at each level — the difference is the chassis and how much of it you want on the desk. The full range shows every model and level side by side.
3D asks for two machines in one: the one you model on and the one that renders. Here is where each program puts the load, what it looks like when a machine is short of it, and which model we would start you on.
| Program | What it depends on | How it shows up | Start with |
|---|---|---|---|
 Blender | GPU and VRAM for Cycles, then cores for simulation bakes and the CPU side of the pipeline. | A scene that renders until it exceeds the card's memory, then falls back to the processor and takes hours. | Focus MaxOne strong card for the artist seat |
 Maya | Single-core speed for rig evaluation, then GPU for Arnold and Redshift batch renders. | Animation playback that drops below real time on a heavy rig, and a shot sequence that runs past the deadline. | UltrafocusTwo cards on the render queue |
 Unreal Engine | One strong card, not several. Real-time rendering does not divide across GPUs the way a batch render does. | Frame rate in the editor, Lumen and Nanite headroom, and how long a shader compile blocks the build. | Focus MaxA second card only for live output |
 Cinema 4D | GPU and VRAM for Redshift, with single-core speed for the serial parts of the scene. | Viewport lag as the MoGraph stack deepens, and sequence render times against a delivery date. | Focus MaxFocus X once sequences are daily |
 Houdini | Cores, memory and disk for simulation — then GPU on top for Karma XPU and Solaris. | Sim times, cache sizes that fill a drive in an afternoon, and volumes too large to hold in memory. | HyperfocusFilm-grade FX and rendering |
 Nuke | Cores and memory for the comp tree, with one card carrying the GPU nodes. | Viewer response on a deep graph, and full-resolution plates that force a drop to proxy. | Focus MaxMore memory before more GPU |
| Render engines generally | VRAM is a wall, not a slider. A scene either fits in the card's memory or it does not. | A render that was fast yesterday failing today because one more texture set pushed it over the limit. | Size the card to the sceneNot to the benchmark |
Where your recent work sits matters as much as which program you open. Stills and single shots run comfortably on one strong card, and the money is usually better spent on memory than on a second GPU.
Animation and lookdev on a deadline is where a second card starts paying for itself, because a render can run while you keep working. Simulation and film-grade VFX is the level at which the machine becomes the farm.
In the order it matters for 3D work — not the order it appears on a spec sheet.
Redshift, Octane, Cycles and V-Ray GPU do the entire job on the card, and the same card draws your viewport. Nothing else on this list moves render time as much.
Simulation is CPU work — Houdini solvers, cloth, fluids, destruction — and so are Arnold CPU, V-Ray CPU and Corona. Clock speed makes the viewport responsive; core count is what shortens a CPU render.
Scene loading, sim caching, and keeping a DCC, a browser and a comp open at once. Running out of system memory part-way through a solve can discard an afternoon's cache.
Sim caches and texture libraries are read and written constantly. Fast local NVMe is the difference between a cache that scrubs in real time and one that stutters — network storage is not a substitute here.
A render is a sustained full-load run, not a benchmark burst. Cooling is what keeps clocks where they should be in hour six, not just hour one.
Focus Max · Ultrafocus · Hyperfocus · Creator Compact
Custom-built workstations for professional 3D creators. Maya, Blender, Cinema 4D, Houdini, Unreal Engine — there is a model built for each stage: FOCUS MAX for the artist seat, Ultrafocus when renders hold up the week, HYPERFOCUS when the machine is the farm.
A 3D workstation is really two machines sharing a case. Modelling, layout and animation are interactive — the wait is on one CPU thread evaluating the scene and on how fast the graphics card redraws a dense viewport. Final-frame rendering is the opposite: it loads every core and every card in the machine, for hours.
Those two jobs want opposite things, and the machine has to do both. Buy on core count alone and it feels sluggish to work in. Buy around one fast card and it stalls when a sequence render lands. FOCUS MAX is specified as the balance point between them, and it is where we suggest most 3D artists start.
Modelling · layout · animation · lookdev
One CPU core doing scene evaluation, and the GPU drawing the viewport.
High clock speed, a card with enough memory to hold the scene, and RAM headroom so nothing starts swapping.
Extra cores. A 64-core CPU redraws your viewport no faster than a 16-core one — most DCC viewports still evaluate on a single thread.
Scrubbing the timeline stutters, or the viewport lags behind the mouse.
Final frames · sequences · simulation
Everything — every core on a CPU renderer, every card on a GPU renderer, for hours at a time.
GPU count for Redshift, Octane, Cycles and V-Ray GPU. Core count for Arnold CPU, V-Ray CPU and Corona.
Running out of video memory. The scene either fails outright or falls back to out-of-core and slows by a multiple.
The day is planned around a frame count rather than a shot list.
Every other spec on a workstation degrades gracefully. Video memory does not. If the scene fits on the card, it renders at full speed. If it does not, one of two things happens: the render fails, or the engine falls back to system memory and frame times go up by a multiple — not a percentage.
What fills it is predictable: 4K and 8K texture sets, displacement, dense hair and fur, volumetrics. A plain-looking interior with a full PBR texture library often sits heavier on the card than a hero character. Spec the card for the heaviest scene you ship, not the average one.
This is also why a second card is not the same purchase as a bigger card. On most GPU renderers the scene has to fit on every card — two cards render twice as fast, they do not hold twice the scene. Scenes that will not fit need more memory. Frames that take too long need more cards.
Each row is a different workload. The highlighted column is the component that repays extra budget for that work.
| Your workload | GPU countmore cards | Video memorybigger card | CPU coresmore threads | Local NVMefast scratch |
|---|---|---|---|---|
| GPU renderingRedshift, Octane, Cycles, V-Ray GPU | Primary factor | Primary factor | Secondary factor | Secondary factor |
| CPU renderingArnold CPU, V-Ray CPU, Corona | Minimal effect | Minimal effect | Primary factor | Secondary factor |
| SimulationHoudini solvers, cloth, fluids, RBD | Secondary factor | Secondary factor | Primary factor | Primary factor |
| Real-time & virtual productionUnreal Engine, archviz walkthroughs | Minimal effect | Primary factor | Secondary factor | Primary factor |
| Modelling, layout, animationThe interactive half of the day | Minimal effect | Secondary factor | Secondary factor | Secondary factor |
Anyone can order the same parts. What you are really choosing is who builds the machine, who tests it before it ships, and who answers the phone when something goes wrong two years from now. That is where we are different.
4.9 on Google from over 4,200 customer reviews
Best Custom Desktop PC Brand 2020–2023
8 years of workstation builds — universities, studios and practices
Every system is checked as it is built, then stress-tested and fully updated before it leaves us — so problems are found here, not at your desk. It arrives ready to work, hand-foamed for safe transit anywhere in Australia. Four years of national awards say we get that right.
Full-system cover for three years, extendable to four — the whole machine, not a chain of separate component warranties you have to chase. If something happens in transit or years later, a dedicated team gets you running again. Support carries on for as long as you own it.
Add on-site servicing and we come to you. No packing it up, no courier, no waiting while it travels both ways — a technician diagnoses and repairs it where it sits. It is the difference between days of downtime and hours of it.
A dedicated workstation team — not a call centre — with eight years of these builds behind it, trusted by universities, studios and practices. Call or email and an Australian expert helps you match the right tool to the job — before you buy, and long after.
What we actually check before a machine is packed, how we specify one in the first place, what happens in year three, and an honest comparison against buying off the shelf or building it yourself.
Real-time virtual sets, Epic’s own Unreal showcase, and aircraft firefighting simulators for the CFA. Rendering that has to hold up live, in front of people, with no second take.
Our Hyperfocus workstation powers Dreamscreen’s 12 × 4.5m LED virtual production stage at their 950m² Epping studio — one of Australia’s most advanced filmmaking environments. Rendering immersive virtual sets in real time demands hardware that never drops a frame.
Invited as the official technology partner for Unreal Fest, supporting Epic Games’ showcase of Unreal Engine across virtual production, game development and real-time experiences.
We partnered with the Country Fire Authority to power their state-of-the-art aircraft firefighting simulators — giving crews the tools to train for the critical, high-pressure moments of flying difficult firefighting aircraft before they ever need to do it for real.
Some of these relationships now run to several years and multiple deployments. Others began with a single machine.
Curated configurations held in stock and dispatched fast. Professionally built, tested and backed by Aftershock PC — no configuration required.
Start with the machine built for your software, then change what you like. We build, test and optimise it before it leaves us.
Tell us the camera, the codec and how many seats. We'll come back with a spec and a price — written enquiries answered within one business day.
