01 · Cloud Engine 30+ POPs, scaling to 100+

A sovereign, powerful cloud engine inside your network.

A complete GPU cloud that runs exclusively in your network. Our GPUs are live in 24 markets, with major telcos including Jio and True. 100+ POPs. No capital outlay, sovereign by design, ensuring your data stays yours.

42RU1 rack footprint
~10kWStandard DC power
<14msInput latency
216kSubs per 5-rack POP
SRX · FIREWALL QFX · 100G FABRIC DELL · MANAGEMENT SUPERMICRO · GPU-A SUPERMICRO · GPU-B MUMBAI BANGKOK JAKARTA DUBAI RIYADH LAGOS SÃO PAULO MEXICO J'BURG
02 · The AMD moat

AMD's go-to-market
partner for telcos.

AMD is an investor in Radian Arc, and Radian Arc is AMD's go-to-market partner for telcos, bringing AMD cloud GPU silicon to operators worldwide.

That partnership is the moat. It's why a sovereign, in-network cloud engine is economically viable at telco scale · and why we can underwrite the hardware without passing CAPEX to the operator.

Go-to-market partnership Cloud GPU · telco channel
01Investor. AMD holds strategic equity in Radian Arc.
02Go-to-market partner. AMD's telco channel for cloud GPU runs through Radian Arc.
03Silicon access. AMD cloud GPU (v620 → R9600D & Asrock Blade).
04Density leadership. Radian Arc runs the highest density of CCU per GPU, and the highest CCU per kWh.
05AI ready. The R9600D is a powerful platform for monetizing both cloud gaming and AI workloads.
03 · The stack

Five layers, one rack, one control plane.

Every Radian Arc POP is the same shape: physical infrastructure at the bottom, compute above it, our control plane binding them, content licensed on top, and the subscriber experience at the surface. Five layers, one rack, federated across 24 markets · and every layer is inspectable, swappable, and owned by a single control plane. Click any layer to inspect what sits inside.

L1
Physical42RU rack · Juniper · Dell · Supermicro
42RU
L2
ComputeAMD GPU · v620 → R9600D → Asrock Blade
18RU
L3
Control planeRadian Arc federation · 30+ POPs, 1 brain
SW
L4
Content1,000+ titles · licensed · co-streamable
SW
L5
ExperienceWhite-label streaming · operator-branded
APP
Layer 01 · Physical infrastructure

The rack. One of them. In your data center.

A single 42RU rack fits in any Tier-III or carrier-grade data center: ~10kW draw, standard power, standard cooling, standard optical uplinks. No exotic footprint, no new building, no new skills on the ops team. If you can land a router, you can land a Radian Arc POP.

Footprint
1 rack 42RU
Power
~10 kW
Cooling
Standard air
Uplink
100G optical
Juniper SRX firewall Juniper QFX 100G Dell PowerEdge mgmt 2× Supermicro GPU chassis
Layer 02 · Compute

AMD silicon. Upgradable and dense.

Every POP runs AMD's cloud GPU. v620 ships today, with the R9600D next and the Asrock Blade on the roadmap · all on the same chassis, via field-swappable card upgrades. Radian Arc runs the highest CCU density per GPU and per kWh, so a single rack underwrites roughly 914 paying subs.

Silicon today
v620
Next rev
R9600D
Roadmap
Asrock Blade
Subs per BOM
~914 subs
Hot-swap cards ROCm-ready 3.5% subs : CCU AI / VAS workloads
Layer 03 · Control plane

One brain. Nine POPs. Global session routing.

Radian Arc's federation layer is a single control plane over every deployed rack, everywhere. Sessions are routed to the nearest healthy POP; overflow spills to a partner POP or to public cloud. You manage a market, we manage the cloud.

POPs live
9 markets
Failover
POP <> POP + AWS
Session latency
< 20 ms net
Input latency
< 14 ms gp
Global session router Capacity federation Hybrid-cloud overflow SLA management
Layer 04 · Content

1,000+ titles. Licensed. Cleared. Stream-ready.

The catalog ships pre-cleared for streaming via Blacknut and direct publisher deals. No per-title licensing work on the operator side. Co-streamable with operator-owned content or regional titles when you have them.

Titles
500+ AAA + indie
Rating mix
E · T · M
Publisher deals
Direct + Blacknut
Refresh
Monthly pipeline
Blacknut catalog Direct publisher Regional titles Operator co-stream
Layer 05 · Experience

Operator-branded. One-tap. Frictionless.

White-label native apps on iOS, Android, Smart TV and WebOS, billed through the carrier, launched inside the operator's own app-home. Subscribers never see Radian Arc. You own the brand. We own the cloud underneath.

Platforms
4 iOS/AND/TV/Web
Billing
Carrier + IAP
Branding
100% operator
Onboarding
1-tap launch
White-label apps Operator brand Carrier billing IAP ready

One rack. 216,000 subscribers.

A Radian Arc POP is deliberately unexciting hardware: Juniper edge, Dell management, two Supermicro GPU chassis. Boring is the point · any telco ops team can land this in any data center, this quarter.

U42 Juniper SRX · firewall 1U
U41 Juniper QFX · 100G fabric 1U
U40 Dell PowerEdge · mgmt 1U
U39cable channel
U34-38 Supermicro · GPU chassis A 5U
U33air gap
U28-32 Supermicro · GPU chassis B 5U
U01-27expansion · R9600D / Asrock Blade upgrade
100G  uplink 
OOB mgmt 
v620 today 
 SRX firewall
 QFX 100G fabric
 R9600D / Asrock Blade upgrade
 ~10kW total draw
Capacity scaling · Asrock Blade roadmap

One rack starts the market. Five saturates it.

1 rack ~43k subs
2 racks ~86k subs
3 racks ~130k subs
4 racks ~173k subs
5 racks ~216k subs
BOM · one rack · one POP

Off-the-shelf. On purpose.

×1Juniper SRXedge firewall · in-network security1U
×1Juniper QFX100G fabric · session router1U
×1Dell PowerEdgemanagement + telemetry1U
×2Supermicro GPU chassisAMD v620 → R9600D → Asrock Blade5U ×2
·Radian Arc control planefederation · session routing · failoverSW
05 · Federation

Nine POPs. One control plane.

Every rack, in every market, runs under one Radian Arc brain. Sessions route to the nearest healthy POP; if a market spikes, capacity spills sideways to a peer POP or up to public cloud. No one market ever runs alone.

Global session router
A subscriber in Mumbai might stream from Mumbai, Riyadh or Bangkok. The router picks what's fastest and healthiest, transparently.
Capacity overflow, POP <> POP
A regional cricket final spikes Jakarta? We spill into Singapore or Sydney capacity for the duration of the surge.
Hybrid-cloud failover, AWS
If the federation itself is saturated, we overflow into public cloud at cost · never drop a subscriber mid-session.
CONTROL PLANE AWS HYBRID OVERFLOW

Federated across 24 markets,
four continents, one control plane.

Live · in production · 2
POC · signed, deploying · 4
Target · commercials open · 3
07 · Beyond cloud gaming

The same rack runs the next workload.

Cloud gaming is the anchor tenant · but a Radian Arc POP is a GPU cloud. Once the rack is in-network, the operator unlocks AI inference, desktop-as-a-service, SD-WAN, and the long tail of GPU workloads that telcos have been unable to monetize locally.

Today · shipping

Cloud gaming as the foothold.

Cloud gaming · B2CLive

1,000+ titles, white-labelled, carrier-billed. The anchor tenant that pays for the rack.

VAS billing · AI workloadsLive · ROCm

AMD ROCm on the same GPUs the games run on · used for CDN telemetry and AI inference for value-added services.

Cloud PC · DaaSLive

Desktop-as-a-service for CAD, medical imaging, ML training · streamed from the same rack.

Next · 2026

Software-defined network in a box.

SD-WAN · ActivePort2026

ActivePort SDN/SD-WAN running inside the POP · the rack is the orchestration plane for the operator's own network.

Bandwidth-on-demand2026

Dynamic capacity allocation between services · cloud gaming at peak, enterprise workloads off-peak.

Vision AI2026

Sovereign AI models with the highest models per camera, quantum security, support for legacy 2K cameras, and the highest camera density.

Roadmap · 2027+

The sovereign edge cloud.

Smart CityRoadmap

Computer vision and edge AI for municipal deployments · traffic, safety, crowd analytics.

Health · imaging AIRoadmap

Radiology and pathology inference on sovereign silicon. Patient data never leaves the country.

VMaaS · BMaaSRoadmap

Virtual-machine and bare-metal-as-a-service · the same rack, rented by the hour to enterprise customers.

08 · Go live

From survey to streaming in three steps.

No build project, no integration team. Radian Arc ships, imports, and configures the rack. Your team racks it and powers it on. Most POPs go from kickoff to live in weeks, not quarters.

Step 01

Data center survey

A short survey call to confirm power, cooling, rack space, and uplinks. We complete the site checklist with your ops team.

One call · we lead it
Step 02

Receive the equipment

Radian Arc handles shipping and importing. The complete install kit arrives with the installation guide, ready to rack.

RA ships + imports
Step 03

Rack & stack

Your team racks and cables it, a few hours of work. The Radian Arc software self-configures and authenticates, and the POP comes online.

A few hours · self-configures
The full deployment kit, installation guide, and telco checklist are shared with your technical team during onboarding. Request the deployment kit →
09 · Next step

Give us a market.
We'll come back with
a POP plan.

A BOM, a commercial structure, a deployment timeline, and a first-year subscriber model · tailored to your network, your billing, your content regulator.

Commercial · APAC / MENAcommercial@radianarc.io
Technical · architecturearchitects@radianarc.io
Deployment · opsdeploy@radianarc.io
HQ · Singapore · Mumbai · Perth+65 · +91 · +61
Radian Arc
A Submer Group Company
A sovereign GPU cloud inside the operator's own network. AMD-powered, CAPEX-free, live in 24 markets.
INTERNAL STAGING · v1.0
© 2026 Radian Arc · Sovereign by design Sydney · Singapore · Dubai · São Paulo