← Back to home

Fleet Hardware Inventory & Production Over Time

📊 Cross-Referenced EOC Production Summary (User ID: 1358046)

Total Points: 9,970,042,939 | Total WUs: 19,633

EOC Current 24-Hour Average: 55,096,108 PPD

EOC Monthly Production (2026):

Feb 2026: 850.6M pts (~30.4M PPD/day)

Mar 2026: 1,048.8M pts (~33.8M PPD/day)

Apr 2026: 964.9M pts (~32.2M PPD/day)

May 2026: 1,031.5M pts (~33.3M PPD/day)

Jun 2026: 988.9M pts (~33.0M PPD/day)

Jul 2026 (to date): 870.9M pts (~58.1M PPD/day)

---

🖥️ Fleet Hardware & Best Estimate Average PPD Speed (Feb 2026 – Present)

---

🔬 Detailed Host Configuration Specifications

1. itztli

Role: Primary Flagship Node & Master Telemetry Aggregator & User Desktop/Gaming Rig

Physical CPU: AMD Ryzen 9 9950X3D (Zen 5 16-Core / 32-Thread, 1 Socket AM5)

System Memory (RAM): 64GB DDR5 (2x 32GB DDR5-6000 EXPO)

PCIe Platform Generation: PCIe Gen 4.0 / Gen 5.0 (24 PCIe 5.0 lanes from CPU)

Discrete GPU: 1x NVIDIA GeForce RTX 4090 (24GB VRAM, PCIe 4.0 x16, capped at 315W)

iGPU: Integrated AMD Radeon Graphics (used for display, non-folding)

2. origami

Role: High-Density Folding and Local LLM Node

Physical CPU: Intel Core i9-7940X (Skylake-X 14-Core / 28-Thread, 1 Socket LGA2066)

System Memory (RAM): 128GB DDR4 (8x 16GB DDR4-3200 Quad-Channel)

PCIe Platform Generation: PCIe Gen 3.0 (44 PCIe 3.0 lanes from CPU, X299 Platform)

Discrete GPUs & Slot Link Speeds (3 Total):

Slot 0: 1x NVIDIA RTX PRO 4000 Blackwell (24GB VRAM, PCIe Gen 3.0 x16, 145W cap) (Swapped May 18)

Slot 1: 1x NVIDIA GeForce RTX 5070 (12GB VRAM, PCIe Gen 3.0 x8, 180W cap)

Slot 2: 1x NVIDIA GeForce RTX 5070 (12GB VRAM, PCIe Gen 3.0 x8, 180W cap)

3. smallpieces

Role: Multi-GPU Octominer Folding Host

Physical CPU: Intel Core i7-7700 (Kaby Lake 4-Core / 8-Thread, 1 Socket LGA1151) (Upgraded Jun 30 from i7-4790 Haswell on AX-B85M-ETH board)

System Memory (RAM): 32GB DDR3L (2x 16GB DDR3L SODIMM, upgraded Jun 30 alongside Octominer X12 Ultra migration from legacy AX-B85M-ETH 8GB SODIMM limit)

PCIe Platform Generation: PCIe Gen 3.0 / Gen 2.0 (Octominer Riser Bus Topology)

Discrete GPUs & Slot Link Speeds (5 Live Total):

Bus 01:00.0: 1x NVIDIA GeForce RTX 5060 Ti (8GB VRAM, PCIe Gen 3.0 x8, 150W cap)

Bus 02:00.0: 1x NVIDIA GeForce RTX 4070 (12GB VRAM, PCIe Gen 3.0 x1, 145W cap)

Bus 04:00.0: 1x NVIDIA GeForce RTX 5060 Ti (8GB VRAM, PCIe Gen 3.0 x1, 150W cap)

Bus 05:00.0: 1x NVIDIA GeForce RTX 5060 Ti (8GB VRAM, PCIe Gen 3.0 x1, 150W cap)

Bus 06:00.0: 1x NVIDIA GeForce RTX 5060 Ti (8GB VRAM, PCIe Gen 3.0 x1, 150W cap)

(Retired Legacy Cards: 1x RTX 3060 Ti a76d84a0, 1x RTX 2060 SUPER, 1x RTX 2060)

---

📈 EOC Production Analysis & Milestones

1. Feb – May Baseline (~30M – 33M PPD):

During Feb, Mar, Apr, and May 2026, the fleet averaged 30.4M – 33.8M PPD.

The addition of itztli (RTX 4090) in April and the Blackwell RTX PRO 4000 swap on origami in May increased theoretical capacity, but overall output was constrained by legacy Haswell hardware on smallpieces.

2. Late June Production Breakout (+70% Output):

Between June 18 and June 30, production surged from ~33M PPD to ~54.7M PPD (+65-70% increase).

Key Driver: Platform upgrade of smallpieces (i7-7700 Kaby Lake + Octominer riser expansion), allowing multi-GPU parallel folding without PCIe bus bottlenecks.

3. July Peak & Modern Efficiency Standard (~57.5M EOC PPD / 102.9M QRB Peak):

Jul 13 Peak: Reached 60.55M PPD on EOC daily stats.

Jul 14 Consolidation: Replacing older legacy GPUs (3060 Ti / 2060S / 2060) on smallpieces with 4x RTX 5060 Ti + 1x RTX 4070 achieved ~57.5M EOC PPD with 2 fewer total GPUs, maximizing power efficiency and QRB bonus multipliers.