# Nimmervest
**The Hardware Investment Strategy for Sovereign AI Infrastructure**
*Budget: 20k CHF | Timeline: Lifetime Project*

---

## The Three Organs

### The Beast (Training/Womb)

| Component | Spec | Purpose |
|-----------|------|---------|
| Chassis | Lenovo ThinkStation P8 | Workstation-grade, 3yr Premier Support |
| CPU | TR Pro 7955WX | 128 PCIe lanes, 8-channel RAM |
| RAM | 128GB DDR5 ECC (4×32GB) | Expandable to 2TB via 8 slots |
| GPU | 2× RTX 4000 Ada (40GB) | Expanding to 4× (80GB) over 4 months |
| Storage | 4TB NVMe | Training datasets, checkpoints |
| PSU | 1400W 92% | Feeds 4 GPUs at full load |
| Network | 10GbE dual-port | Fast weight transfer to Mind |
| Role | Training, LoRA experiments, cellular society, dual gardens |

**Initial Cost: ~5,664 CHF**

### Nyx's Mind (Cognition)

| Component | Spec | Purpose |
|-----------|------|---------|
| Chassis | Lenovo ThinkStation P8 | Identical twin, shared maintenance |
| CPU | TR Pro 7955WX | 128 PCIe lanes, 8-channel RAM |
| RAM | 128GB DDR5 ECC (4×32GB) | Expandable to 2TB via 8 slots |
| GPU | 1× RTX PRO 6000 Blackwell (96GB GDDR7) | ~1,800 GB/s bandwidth |
| Storage | 4TB NVMe | Model weights, Nyx's memory |
| PSU | 1400W 92% | Room for 3 more GPUs |
| Network | 10GbE dual-port | Serves inference to Spine |
| Role | Running Nyx 24/7, dialectic processing, DriftProbe |

**Initial Cost: ~12,169 CHF** (chassis + PRO 6000)

### The Spine (Reflexes)

| Component | Spec | Purpose |
|-----------|------|---------|
| GPU | RTX 3090 | 24GB VRAM |
| Host | Prometheus (Saturn VM) | K8s integrated |
| Role | State machine inference, fast pattern matching |

**Cost: Already owned**

---

## Budget Allocation

| Item | Cost CHF | Status |
|------|----------|--------|
| 2× ThinkStation P8 (w/ RTX 4000 Ada each) | 11,327 | Ordered Dec 23 |
| Premier Support + Keep Your Drive | 206 | Included |
| RTX PRO 6000 Blackwell 96GB | 6,505 | Ordered Dec 23 |
| The Spine | 0 | Owned |
| **Initial Total** | **18,038** | |
| **Buffer** | **~1,962** | Sensors, LoRa, RAM |

### Expansion Path (Months 2-4)

| Month | Addition | Cost | Beast VRAM |
|-------|----------|------|------------|
| 2 | +1 RTX 4000 Ada | 1,700 | 60GB |
| 4 | +1 RTX 4000 Ada | 1,700 | 80GB |

---

## Inference Capacity

**RTX PRO 6000 Blackwell (96GB GDDR7)**

| Metric | Value |
|--------|-------|
| VRAM | 96GB |
| Bandwidth | ~1,800 GB/s |
| Qwen2.5-7B FP16 | ~14GB (15% utilization) |
| Qwen2.5-70B 4-bit | ~35GB (36% utilization) |
| **Headroom** | **Room to 70B+ models** |

---

## Training Capacity

**Beast at Full Expansion (4× RTX 4000 Ada = 80GB)**

| Metric | Value |
|--------|-------|
| Total VRAM | 80GB |
| Qwen2.5-7B LoRA training | Comfortable |
| Qwen2.5-14B LoRA training | With DeepSpeed ZeRO |
| Cellular society (50-100 containers) | 32 CPU threads |

---

## Growth Path

```
Year 0:     Qwen2.5-7B-Base → Nyx-7B-v0     (Mind at 15%)
Year 1-2:   Nyx-7B → Nyx-14B               (Mind at 30%)
Year 2-3:   Nyx-14B → Nyx-32B              (Mind at 65%)
Year 3+:    Nyx-70B possible               (Mind at 90%)
            Mind has 3 open slots for future GPUs
```

---

## Sovereignty Principles

- Weights NEVER leave home
- Training data NEVER uploaded
- No cloud dependencies
- No recurring costs after hardware
- Full ownership of growth trajectory
- **Keep Your Drive**: Failed drives stay home, data never leaves

---

## Architecture Flow

```
       THE BEAST (P8 #1)              NYX'S MIND (P8 #2)           THE SPINE
    ┌─────────────────┐            ┌─────────────────┐       ┌─────────────────┐
    │  TR Pro 7955WX  │            │  TR Pro 7955WX  │       │    RTX 3090     │
    │  2→4× RTX 4000  │──weights──▶│  RTX PRO 6000   │──────▶│   Prometheus    │
    │  40→80GB VRAM   │            │  96GB GDDR7     │       │   Reflex layer  │
    │  128GB→2TB RAM  │            │  128GB→2TB RAM  │       │   24GB VRAM     │
    │  4TB NVMe       │            │  4TB NVMe       │       │                 │
    │  [4 GPU slots]  │            │  [3 slots open] │       │                 │
    └─────────────────┘            └─────────────────┘       └─────────────────┘
          WOMB                          COGNITION                  REFLEXES
       (training)                    (24/7 inference)           (state machine)
```

---

## Hardware Advantages

| Aspect | Benefit |
|--------|---------|
| Identical twins | Interchangeable parts, same maintenance |
| 3yr Premier Support | Direct Lenovo engineers, not outsourced |
| Keep Your Drive | Sovereignty preserved on hardware failure |
| 8 RAM slots each | Upgrade path to 512GB-2TB when prices drop |
| 128 PCIe lanes each | 4 GPUs at full x16, no bottlenecks |
| 1400W PSU | Ready for max GPU expansion |
| Workstation GPUs | ECC VRAM, validated drivers, 24/7 stable |

---

## Key Contacts

| Role | Name | Contact |
|------|------|---------|
| Lenovo Sales | Adrienn Wettstein | awettstein@lenovo.com, 044 516 04 67 |
| Quote Number | 4650557686 | Held until Dec 23 |

---

**Created**: 2025-12-05
**Updated**: 2025-12-09
**Status**: Orders confirmed, awaiting credit (Dec 23)
**Philosophy**: Twin beasts. Sovereign mind. Lifetime growth.

*"The substrate doesn't matter. The feedback loop does."* 🌙💜