We needed to run trusted workloads on untrusted host machines.
So over a year ago, we started building the Targon Virtual Machine to enable Confidential TEEs in production.
Today we're sharing our white paper written alongside @intel: Decentralized Compute on Untrusted Hardware Using Intel® TDX and Encrypted CVMs
With exploits on the rise, plan for the host getting breached.
Targon Virtual Machines run on Intel TDX.
Memory is encrypted with keys the host never sees, so even a compromised server can’t see your data, your weights or your code.
Every few hours, it seems like there’s a new software exploit.
This week, a critical PHP flaw forced government websites across Europe offline, and the NSA told anyone who couldn’t patch to take their systems down, and one attacker used AI agents to break into the websites of 100+ companies for about $50 a hack.
Exploits that used to take months now take minutes. And a standard VM still assumes you trust the host. The hypervisor, the admins, anyone who gets that far can read what’s in memory.
The hardware to close that gap already exists. The question is no longer how fast you can patch, it’s whether anyone else can see inside your workload while it runs.
Talking Targon, Ep. 01 is now live.
Bittensor co-founder Jacob Steeves aka @const_reborn joins us to break down eTAO, the upgrade that lets subnet owners decide how their emissions are split.
We get into subnet economics, the guardrails against exploits, and what it unlocks for compute on Targon.
Bittensor is adapting, and passing more control to the network participants.
eTAO is yet another step toward true decentralization.
“Inside of a competitive network, where you're being governed by the market.” @const_reborn on Talking Targon
Talking Targon, Ep. 01 is now live.
Bittensor co-founder Jacob Steeves aka @const_reborn joins us to break down eTAO, the upgrade that lets subnet owners decide how their emissions are split.
We get into subnet economics, the guardrails against exploits, and what it unlocks for compute on Targon.
This Thursday on Novelty Search :: Subnet 4 :: @TargonCompute
Targon is building a decentralized GPU cloud on Bittensor, providing on-demand compute for AI training and inference.
Its confidential VMs use Intel and NVIDIA hardware security to protect sensitive models and data, even from the providers operating the servers.
Live on Thursday :: 9PM UTC :: via Bittensor Discord
Hosted by @const_reborn
Had an awesome time last week representing @TargonCompute in Montreal for @ExploitSummit!
I always leave any conference with Bittensor folks feeling even more bullish, the energy that this community has is contagious.
Great seeing old friends and making some new ones, future is bright!
Last week in Montreal, we sat down with a few of our fellow builders working to push the boundaries of permissionless intelligence on Bittensor.
Innovation is happening rapidly, and we're proud to be shipping alongside such a talented group of passionate leaders.
Thank you to everyone who joined us, we look forward to continue accelerating the growth and capabilities of decentralized AI.
NEWS: @TargonCompute says NVIDIA B300 GPUs are now available on demand on Targon. This adds an on-demand Blackwell deployment option.
https://t.co/1FqtoYVhLS
NVIDIA B300s are now available on demand.
Secure GPUs are selling out fast on Targon, Ready to deploy on Blackwell?
Grab yours now at https://t.co/siuXJPbZtw
The Manifold team is enjoying Montreal at Exploit Summit, where we're announcing new features and roadmaps.
As part of our new release rollout, we've launched Bare Metal and Sandboxes on Targon: a full physical machine when you need the whole box, and disposable Linux environments when you don't.
Thanks for building with us.
We're excited to share that Bare Metal and Sandboxes are now live on https://t.co/RxzEHJPN0t
Bare Metal
→ An entire physical machine dedicated to your organization
→ No hypervisor, no container runtime, no noisy neighbors
→ Full control of kernel, drivers and firmware-level GPU settings
→ Every operator is KYC-verified, keeping workloads secure without TVM
Sandboxes
→ Short-lived, isolated Linux environments for development, testing, previews and agents
→ Provisioning to running within seconds
→ Browser terminal, graphical Linux desktop or SSH
→ Fork independent copies for parallel experiments and agent tasks
→ Publish a template from any setup worth repeating
→ Pause when idle to stop active compute billing, resume anytime
The Targon team has been focused on expanding features and enhancing the Targon experience, and will continue rolling out new releases.
Happy to help! Proud to be trusted with running critical workloads such as validators.
Very excited to see the talented team at DeSci Labs join the Bittensor ecosystem, their research and expertise are sure to add long-lasting value to the network. 🔬
Our validator for SN111 is running on hardware provided by @TargonCompute.
We're super grateful for the fast, reliable setup and the team's amazing support.
Thank you, guys!