A NEW KIND OF CHIP COMPANY

AI-native
Fabless.

Our AI agent fleet and silicon experts work together to design complex chips—faster, at lower cost, and with rigorous testing throughout.

Conceptual silicon die on a light oak presentation block
The Melten Design-Agent System

Chip design across the
AI infrastructure stack.

Our silicon experts and design-agent system take customer chip projects from a product idea to a tested design, with Melten responsible for the engineering work.

AI INFRASTRUCTURE — DESIGN CONTEXT Melten design domainsSELECT A BLOCK ↗
Host processingHost CPUSystem orchestration & control
ConnectivityPCIe SwitchRouting & transaction controlRead story AI compute MULTIPLE DEVICESNPUCompute engines & memory dataflowsRead story Storage MULTIPLE DEVICESNVMe ControllerCommand processing & data pathsRead story
Data processingDPUProtocol processing & offloadExplore design focus ↗

DPU Design

Design focus: data movement, protocol processing, and hardware offload. This domain illustrates data-processing design challenges within AI infrastructure. Project scope and validation evidence will be detailed in its engineering story.

Engineering story coming soon
Network fabricData Center NetworkConnectivity across the data center
Conceptual system view illustrating chip-design domains. Stacked cards represent multiple PCIe-connected devices. Individual projects have separate scopes and validation results. Highlighted domains do not imply a jointly validated system.
Melten Silicon Designs

Complex real-world designs.
Delivered in days, not months.

Explore FPGA implementations across AI compute, connectivity, and storage, with each project’s scope, measurements, and validation results documented.

PCIe Controller / NVMe10 min read

A custom PCIe Gen4 x16 controller, in seven agent-days.

Melten’s agents built a custom PCIe Gen4 x16 interface on an FPGA whose built-in controller stops at Gen3. An encrypted NVMe drive demonstrates 28.4 GB/s reads and Linux boot, with reusable interface, encryption, and search IP for future customer designs.

28.4 GB/sENCRYPTED 128 KiB READS
7.3 agent-daysCAMPAIGN TO CLOSURE
Read engineering story
NVMe Controller6 min read

An NVMe SSD front end, built in 7 days.

Directed by one designer, Melten’s design-agent system built an NVMe SSD front end on an AMD Versal FPGA. Over PCIe Gen4 x8 it moved 13.9 GB/s each way and 3.3 million 4 KiB I/Os per second.

13.9 GB/sEACH DIRECTION
3.3M IOPS4 KiB RANDOM I/O
Read engineering story
PCIe Switch5 min read

A PCIe Gen4 switch, built in 17 days.

Directed by one designer, Melten’s design-agent system built a three-port PCIe Gen4 switch on an AMD Versal FPGA. Two off-the-shelf NVMe SSDs ran through it at 87–99% of their rated speeds.

17 daysSPECIFICATION TO WORKING SWITCH
87–99%OF EACH SSD’S RATED SPEED
Read engineering story
NPU / N116 min read

N11: From Dense Qwen2.5 to Modern Gemma 4—An FPGA NPU Re-Architected in 52 Hours

Melten’s design-agent system re-architected an FPGA NPU for Gemma 4 26B-A4B, routed experts, mixed attention, and 32K context—with 11,226 exact hardware/reference output checks.

52 hoursCAMPAIGN TO CLOSURE
11,226 / 11,226Hardware outputs matched the fixed-point reference.
Read engineering story
NPU / N108 min read

An LLM accelerator, built in five days.

Melten’s design-agent system took a human-set requirement through architecture, RTL, software, verification, implementation, lab bring-up, and independent closure in 122 hours.

122 hoursCAMPAIGN TO CLOSURE
3 modelsONE HARDWARE IMAGE
Read engineering story
View all posts
  • Storage controllers

    Organize how data is stored and retrieved.

    • NVMe controller core
    • NVMe-oF storage offload
  • Connectivity & memory

    Connect processors to the data they need.

    • PCIe virtualization
    • RoCEv2 / RDMA service
    • Shared-memory fabric
  • Secure processors

    Protect the systems people depend on.

    • Secure RISC-V SoC
    • Inline AES-GCM security
AI INFRASTRUCTURE FIRST. NEW INDUSTRIES NEXT.

More ideas can
become real products.

When chips take less time and money to design, custom hardware becomes practical for more applications. We’re starting with AI infrastructure—and building toward the needs of new industries.

Robotics

Chips that help machines sense and act.

Industrial automation

Smarter equipment with real-time control.

Aerospace

Specialized computing for demanding environments.

PROJECT INQUIRIES

Your next chip
starts here.

Tell us what you want your next chip to do.

contact@melten.ai