[ For Research Partners ]

Research-grade AI. Production-grade infrastructure.

The compute platform research teams build on when AI has to run on real hardware. Constrained, sovereign, or air-gapped. Same model. Your environment.

Talk to the team How we engage
  • NVIDIA Inception
  • AWS Partner
  • Microsoft for Startups
  • Garage&Co

Program names and logos are property of their respective owners.

[ Two Paths ]

Shrink the model. Or keep it.

The conventional path makes the model smaller until it fits the hardware. We take the opposite path. Keep the model. Make the hardware reach further.

The conventional path

Shrink the model

Distil, quantise, prune. The model that ships is not the model that was trained.

The Sector88 path

Keep the model. Tier the memory.

We orchestrate model weights across whatever memory your system has available. The model that ships is the model that was trained.

[ The Hard Part ]

A systems problem. Years in the making.

Memory management, kernel selection, hardware abstraction, deployment, fleet telemetry, security hardening. All of it has to work together, on every device, in every environment.

What it looks like

The duct-tape stack

Patched inference servers, custom CUDA kernels, bash scripts for deployment. Works on one machine until the hardware changes.

What we built

One integrated platform

Memory tiering, engine selection, fleet control, air-gapped deploy, audit, identity. Runs across Jetson, x86, ARM, GPU and CPU-only machines.

[ On the machine ]

Memory hierarchy.
In production.

Sector88 probes the machine, picks the engine, and tiers the model across whatever memory is available. GPU and system RAM. p95 latency stays predictable. The model that was trained is the model that serves.

Same software on a Jetson at a ground station, a research workstation, and a server rack in a SCIF.

Explore Runtime
s88 serve --model Llama-3.3-70B-AWQ
INITIALIZING

Llama-3.3-70B-AWQ

AWQ INT4 70B · TP=2
Detected

Backend Selection

Auto
llama.cpp vLLM Native

Memory Hierarchy

PASS

2x RTX 4090 VRAM (hot)

44.1 / 48 GB

System RAM (warm)

8.0 / 64 GB

Serving

localhost:8088/v1/chat/completions

Throughput

34.8 tok/s

Latency

29 ms

OOM Events

0

Uptime

0s

[ The Platform ]

Three layers. One platform.

Fig 1.1

Fleet control

Across the fleet

The control plane. Manage models, nodes, and deployments across every site. Audit, RBAC, telemetry, version control.

Explore Hub

Fig 1.2

Memory tiering

On the machine

The execution layer. Tiered memory across VRAM and RAM. Models that should not fit run stable. GPU or CPU.

Explore Runtime

Fig 1.3

Air-gapped deployment

Deploy

Single Helm chart. Air-gapped install from media. No external dependencies. Same artifact from lab to ground station to facility.

Explore Deploy

[ Across the fleet ]

Every node. One pane.

Pin a model, promote it, roll it back. Same model promoted across the lab, the field, and the facility.

Metadata-only audit trails. Role-based access. Telemetry stays local and resyncs when the link returns.

Explore Hub
Sector88
Illustrative Live

Nodes

128

Serving

124

Fleet Uptime

99.98%

OOM Events

0

Active Deployments

ground-station-08

Svalbard, Norway
Llama-3.3-70B llama.cpp
VRAM 43.2/48 tok/s 26.4 64d up

ops-center-03

Edwards AFB, CA
Qwen3-32B vLLM
VRAM 71.8/80 tok/s 586.2 31d up

rig-platform-11

North Sea, Offshore
YOLO11x · 6 cams TensorRT
VRAM 9.6/24 fps 180.4 112d up

datacenter-sg-02

Singapore, APAC Warming
Llama-3.1-405B TensorRT-LLM
VRAM -- tok/s -- 0s up

Activity

2m agoSite pin promoted to 42 sites. 0 rollbacks.
5m agoPreflight passed on datacenter-sg-02. Loading Llama-3.1-405B.
18m agoVision workload pinned on rig-platform-11. 6 camera streams.

[ How We Engage ]

Three ways in.

We work with research partners in a few different shapes. Each one starts the same way: a real environment and a real problem.

01

Hands-on access

We give your team direct access to the platform in a real environment. You run the models, we sit alongside as you go.

02

Co-funded research

We co-apply on research programs where you bring the academic side and we bring the industry side. National and international funds.

03

Embedded partnership

Longer-term research engagements where we deploy alongside your program. Shape and terms scoped to the relationship.

[ For the Engineers ]

The technical shape.

Fig 3.1

Memory orchestration

Memory orchestration

Weight tiering across VRAM and system RAM. p95 latency stays predictable under load.

Fig 3.2

Model support

Model support

Open-weight transformers, custom fine-tunes, multimodal. vLLM, TensorRT-LLM and llama.cpp serving paths.

Fig 3.3

Validated hardware

Validated hardware

Jetson Orin and Orin NX, industrial edge PCs, x86 with or without GPUs, CPU-only environments.

Fig 3.4

Air-gapped install

Air-gapped install

Single Helm chart. Offline model registry. Signed bundles. Zero external dependencies.

Fig 3.5

Audit and identity

Audit and identity

Metadata-only audit trails. Admin, operator and viewer roles.

Fig 3.6

Sovereign deployment

Sovereign deployment

Data stays on premises. Models stay on premises. Built for no phone-home and no outbound calls.

[ Where It Applies ]

Real environments. Real constraints.

Fig 4.1

Ground station deployment

Ground station

Vision-language model on a Jetson Orin. Imagery processed at the edge. Same artifact promotes to a server rack.

Fig 4.2

Air-gapped sovereign facility

Sovereign facility

LLM on classified hardware. Air-gapped install from media. Offline licence. Built for no outbound calls.

Fig 4.3

University lab fleet

University lab

Multi-model benchmarking across GPU generations. Auto-detect, auto-tier, version-tracked results.

[ Research Output ]

Publishable work. Not just a deployment.

Benchmarks

Benchmark data

Latency, throughput, and memory data across hardware. Reproducible and citable.

Co-authored papers

Research publications

Memory-tiered compute, edge deployment, sovereign AI. Your research, our experimental foundation.

Grant support

Grant evidence

Deployment evidence, validation data, and industry letters for funding applications.

Build with us.

We work with a small number of research partners at a time. Tell us about the environment, the hardware, and the model you are trying to run.

Talk to the team View partner programs