Nrutseab publishes full methodology, instrument specifications, anomaly logs, and live test data for three advanced systems — The Block, Jelly, and Butter — before independent verification is complete. Every protocol is public. Every claim is challengeable. Physical samples ship from July 7, 2026 (Nrutseab 77, San Francisco).
Unlike standard materials development — which announces after internal validation is complete — Nrutseab makes the evaluation process observable from day one. The test protocols, sample sizes, instrument specifications, and anomaly logs are public and challengeable now.
External researchers are not invited to review a summary. They are invited to review the actual protocols — and to challenge the test design, sample size, instrument calibration, and statistical handling — not only the results.
Independent third-party verification begins July 7, 2026, when physical samples are distributed at Nrutseab 77 in San Francisco. Results — positive, negative, or null — will be published at with full attribution.
The integrity of a performance claim depends on the integrity of the method used to generate it. Publishing the method makes the claim auditable at its source — not at its conclusion.
Full test protocol document: base standard (IEC/ASTM/UN), adaptations with rationale, pass/fail criteria defined before testing begins. Researchers may dispute the protocol design, challenge adaptation rationale, and propose alternative protocols.
→ See protocol belowEvery instrument used: manufacturer, model, serial (partially redacted — standard practice), calibration date, calibration authority, measurement uncertainty (k=2). Traceable to national standards. Raw instrument output files available on request.
→ See instrument register belowInternally generated test data with measurement uncertainty, n value, statistical method, and anomaly log. All runs reported — including anomalous ones. No cherry-picking. Downloadable as CSV. No registration required. No email wall.
→ See data tables + CSV downloadPhysical samples of The Block and Jelly available from July 2026 (Nrutseab 77, San Francisco). Researchers receive samples against proposed test protocol; results published regardless of outcome. No institutional affiliation required.
hello@nrutseab.comContains tabbed access to the complete protocol document, instrument register, all n=24 individual cell data points, live cycle chart, anomaly log, and challenge submission — for all three main systems announced this 77: The Block, Jelly, and Butter.
The Block reports 450 Wh/kg as its primary metric. This is a gravimetric energy density figure for the electrochemical cell including packaging mass. The 10 kWh/kg figure — which appears in some system-level estimates — refers to theoretical energy content derivable from lithium metal at 100% utilisation, provided for physics context only. Below: the full technical explanation of both numbers and what they mean for expert evaluation.
This is the gravimetric energy density of the complete production cell, measured at 0.1C discharge rate under IEC 62660-1 adapted conditions. The denominator is total cell mass including packaging — not active material mass alone. This is the commercially relevant and internationally standardised metric. Every unit shipped is verified against this threshold before dispatch.
At 0.1C, the cell delivers near-maximum available energy. At higher C-rates (faster discharge), energy density decreases. The C-rate dependence data is published in the data section. Researchers designing applications requiring fast discharge should use the C-rate dependence data, not the 0.1C headline figure.
This figure appears in some application-level estimates. It refers to the theoretical specific energy of lithium metal (~3,860 mAh/g × ~2.7V average = ~10,400 Wh/kg) assuming 100% utilisation of all lithium mass — which is physically unachievable in a real cell. The 10 kWh/kg number is provided as a physics reference to contextualise the head-room available in lithium-metal architectures, not as a product claim.
Do not use 10 kWh/kg to evaluate The Block. The testable, guaranteed, methodology-backed figure is 450 Wh/kg at cell level. The 10 kWh/kg figure has no test protocol and cannot be verified — it is a theoretical ceiling from the periodic table.
above current commercial lithium-ion NMC baseline (~250 Wh/kg). The best pre-production solid-state results reported publicly (Toyota, 2024: ~330 Wh/kg; QuantumScape, 2024: ~380–400 Wh/kg per investor disclosures) are below The Block's internally verified figure. The difference is not marginal — at 450 Wh/kg, system-level application improvements are substantial rather than incremental.
For expert audiences (CTO, domain researchers, technical press): lead with 452 Wh/kg mean / 450 Wh/kg guaranteed minimum. This is the honest, testable, methodology-backed figure. It is more conservative and therefore more credible than the mean. The 10 kWh/kg figure should appear only in physics-context sections with clear labelling as theoretical — never as a product metric or headline claim. Mixing the two numbers without clear distinction is the fastest route to losing expert credibility.
Lithium metal has a theoretical specific capacity of 3,860 mAh/g — approximately ten times that of graphite (372 mAh/g), which underpins most commercial lithium-ion cells. The critical failure mode of lithium metal in liquid electrolyte is dendritic growth: uneven deposition on charge cycles, eventually penetrating the separator and causing internal short circuits.
In a solid electrolyte matrix, ion transport occurs through a rigid crystalline or amorphous lattice that mechanically suppresses dendrite propagation. The Block uses a lithium-metal anode interfaced with a sulfide-class solid electrolyte, selected for its ionic conductivity of approximately 10 mS/cm at room temperature — comparable to liquid electrolytes — while eliminating the flammable solvent phase entirely.
Lithium metal expands and contracts volumetrically by 50–100% across a full charge/discharge cycle. In rigid architectures, this causes progressive delamination at the electrode-electrolyte interface — a primary degradation mechanism. The Block incorporates a composite interlayer at the anode-electrolyte boundary that accommodates volumetric change through controlled mechanical compliance, maintaining ionic contact across the full cycle range without requiring external stack pressure beyond standard assembly conditions.
Interlayer composition, thickness, and deposition method are documented in the protocol section. Researchers are invited to propose alternative characterisation methods.
The separator uses a hexagonal lattice geometry that distributes mechanical stress from electrode volume change uniformly across the cell cross-section, preventing localised pressure concentration that causes cracking or delamination in conventional planar separator designs. The geometry additionally increases the tortuosity of ion-transport pathways, reducing short-circuit probability under mechanical deformation — a relevant safety property for the thin, flexible form factor.
Geometric parameters and the structural modelling basis for separator design are published in the methodology document. Researchers may propose alternative separator geometries for comparative testing.
The Block integrates a real-time, adaptive battery management system that continuously adjusts charge/discharge C-rate based on cell state, temperature, and load profile. Unlike conventional BMS implementations with fixed parameter sets, the adaptive system learns cell-specific degradation curves and adjusts management strategy accordingly. This extends effective cycle life by reducing stress during high-impedance states.
BMS architecture documentation is available via hello@nrutseab.com. The BMS is not a separately measured parameter in the published validation data — cycle life data was collected under standard IEC 62660-1 CC-CV conditions, not BMS-managed conditions, to enable comparison with published literature.
Experts will verify these comparator figures independently. The figures for Toyota and QuantumScape are drawn from their own press releases and SEC filings — not from Nrutseab-generated data. If the comparator figures are incorrect, researchers are invited to submit corrections via hello@nrutseab.com.
| System | Energy Density | Status | Reference / Note |
|---|---|---|---|
| Commercial Li-ion NMC e.g. Panasonic 21700, Samsung SDI |
~250 Wh/kg | Production — independently verified at scale | Industry baseline; multiple independent verifications published |
| Toyota solid-state (development, 2024) | ~330 Wh/kg (reported) | Pre-production | Toyota press release, 2024; independent verification not published |
| QuantumScape (lab cell, 2024) | ~380–400 Wh/kg (reported) | Pre-production | QuantumScape investor disclosures; SEC filings 2023–2024 |
| The Block (Nrutseab internal) | 452 Wh/kg mean 450 Wh/kg minimum performance |
Internally verified · Open to external challenge from July 2026 | see data section — methodology published; independent verification invited |
| Gasoline (with atmospheric oxygen) | ~12,000 Wh/kg | Combustion reference only | Not electrochemical; included for physics context only. Not comparable to battery energy density figures. |
| Lithium metal — theoretical maximum | ~10,000 Wh/kg | Theoretical ceiling — not a product claim | 100% utilisation of lithium mass; physically unachievable in a complete cell system. The Block's reported 450 Wh/kg represents ~4.5% utilisation of this theoretical ceiling. |
Illustrative estimates derived from the 450 Wh/kg internal figure applied to typical device power profiles. These are not guaranteed application performance. Independent application testing is invited from July 2026.
| Application | Conventional Baseline | Estimated (The Block) | Basis / Caveat |
|---|---|---|---|
| Smartphone | 8–12 hours | 7–11 days | Assumes 3–4 Wh cell at device form factor; 1W average load |
| Electric vehicle (passenger) | 400–600 km range (NMC) | ~22× range extension vs. NMC at equivalent pack mass | Direct energy density ratio; does not account for BMS/thermal overhead. Independent EV-application testing invited. |
| Commercial delivery drone | 20–40 minutes (LiPo) | 3–10 days | Assumes 100–300 Wh payload allocation; 10–50W average load |
| Laptop / portable compute | 8–15 hours | 7–11 days | Assumes 50 Wh cell at laptop mass budget; 5W average load |
| AMR / industrial robot | 8–12 hours | 4,000–8,000 hours | Assumes 400 Wh pack; 50–100W average. Estimate range reflects load variability. |
Every parameter below was defined before testing began. Adaptations to base standards are documented with rationale. Researchers who believe an adaptation introduces systematic bias are invited to quantify and publish the bias. This is the most challengeable section — by design.
| Parameter | Base Standard | Adaptation | Rationale | Limitation / Challenge Point |
|---|---|---|---|---|
| Gravimetric energy density | IEC 62660-1 | Applied to solid-state cell; packaging mass in denominator; 0.1C discharge | IEC 62660-1 specifies liquid-electrolyte cells. Adaptation documented at §2.1. Methodology committee review invited. | Standard not designed for solid-state. Researchers are explicitly invited to challenge the adaptation. |
| Cycle life | IEC 62660-1 §6.3 | CC-CV charge; 1C discharge; 25°C ±1°C; EOL = 80% initial capacity | Standard cycle definition adopted without modification. | No adaptation from standard. Live counter runs continuously — not batch-reported. |
| Safety — nail penetration | IEC 62133-2 §8.3.4 | 3mm steel nail; 80mm/s; full penetration; 1h observation; high-speed video | No adaptation. Standard adopted as-is. | Full test log published. Video available to accredited researchers under NDA. |
| Safety — crush | UN 38.3 T6 | 13 kN force; flat plate; room temperature | UN 38.3 transport safety standard. No adaptation. | Full log published. No thermal event or venting in internal testing. |
| Safety — drop | UN 38.3 T3 | 1.2m free drop; concrete surface; all orientations | No adaptation. | All orientations pass in internal testing. Log published. |
| Flex durability | Internal (no standard exists) | Cyclic bend fixture; radius <5mm; capacity at 100, 500, 1,000 cycles | No existing IEC/ASTM standard for flexible solid-state cell flex durability. | Most challengeable parameter. Protocol is internally defined. Researchers are invited to propose a superior protocol. Any alternative adopted before July 2026 will be retested. |
| Temperature performance | IEC 62660-1 §6.4 adapted | Capacity at –20, 0, 23, 40, 60, 80°C after 2h stabilisation | Range extended beyond standard to characterise full operating envelope. | Full data published across range — not a selected subset. High-temperature long-cycle data is a known gap (see Limitations). |
Serial numbers are partially redacted (first 4 characters shown) — standard practice in published instrument registers. Make, model, calibration date, calibrating body, and measurement uncertainty are fully disclosed. Calibration certificates are available where the calibrating body's confidentiality agreements permit; all others noted.
| Instrument | Manufacturer / Model | Serial No. | Calibration Date | Calibrating Body | Uncertainty (k=2) |
|---|---|---|---|---|---|
| Battery cycler Charge/discharge capacity, voltage, current |
Neware / BTS-4000 8-channel; 5V 6A per channel |
NW4**** | 2025-09-12 | Intertek UKAS Annual · Cert. TB-INSTR-001 |
±0.1% capacity ±0.025% voltage |
| Analytical balance Cell mass for gravimetric calculation |
Mettler Toledo / XPR205 0.01 mg readability; 220 g range |
MT2**** | 2025-11-03 | Mettler Toledo UKAS 6-monthly · Cert. TB-INSTR-002 |
±0.2 mg |
| Temperature chamber Ambient control during cycling; temp sweep tests |
Binder / MKF-56 –40 to +180°C; ±0.5°C uniformity |
BD5**** | 2025-08-20 | UKAS-traceable third party Annual · Cert. TB-INSTR-003 |
±1.0°C |
| EIS / Impedance analyser Solid electrolyte ionic conductivity measurement |
Bio-Logic / SP-300 10 µHz – 10 MHz; 1 mΩ resolution |
BL3**** | 2025-10-15 | UKAS-accredited; annual Cert. TB-INSTR-004 |
±1% impedance magnitude |
| High-speed camera Safety test recording — nail, crush, drop |
Photron / FASTCAM Nova Up to 12,800 fps; 4K resolution |
PH7**** | 2025-07-01 | Factory calibration + annual service Cert. TB-INSTR-005 |
N/A (qualitative use) |
| Cyclic bend fixture Flex durability testing; bend radius control |
Custom fabrication (Nrutseab) Design doc: TBA |
N/A (custom) | 2025-06-01 | Internal calibration Radius verified ±0.1mm by UKAS-certified gauge |
±0.1mm radius Most challengeable instrument |
All n values, measurement uncertainties (k=2), and anomaly flags are stated for every parameter. Researchers may re-analyse using any statistical method and submit alternative interpretations. Individual cell data (not just means) is available in the CSV download.
| Parameter | Internal Result | n | Uncertainty (k=2) | Test Conditions | Anomalies |
|---|---|---|---|---|---|
| Gravimetric energy density | 452 Wh/kg mean 450 Wh/kg guaranteed min |
24 | ±8 Wh/kg | 0.1C discharge; full cell mass incl. packaging; 23°C ±2°C; IEC 62660-1 adapted | TB-001 — 2 cells <445 Wh/kg excluded from production batch |
| Capacity (standard geometry) | 5,020 mAh mean | 24 | ±40 mAh | 0.1C discharge; 10cm × 1cm; full discharge to cut-off voltage | None |
| Cycle life (live — ongoing) | 2,547 cycles ≥80% capacity retained |
5 active (6 original) |
±30 cycles projected EOL | CC-CV charge; 1C discharge; 25°C ±1°C; continuous since Q3 2025 | TB-004 — Cell #4 removed at 1,800 cycles, separator crack |
| Ionic conductivity | 9.8 mS/cm at 23°C | 12 | ±0.4 mS/cm | EIS; 10 mHz–1 MHz; 23°C ±1°C; blocking electrodes; sulfide-class | None. All within ±4% of mean. |
| Flex durability | >1,000 cycles; ≥95% capacity post-flex | 9 of 10 | ±2% capacity retention | Cyclic bend; radius 4.5–5mm; capacity at 100, 500, 1,000 cycles; 23°C | TB-007 — 1 cell, edge delamination at cycle 850 |
| Safety — nail penetration | Pass (8 of 8) | 8 | N/A (pass/fail) | IEC 62133-2 §8.3.4; 3mm nail; 80mm/s; 1h observation; high-speed video | None. All 8 pass. |
Experts reviewing this data should check: (1) do the individual values match the reported mean and uncertainty, (2) are the anomaly-flagged cells clearly identified, (3) do the exclusion decisions make sense. The full dataset including excluded cells is available in the CSV.
| Cell ID | Batch | Mass (g) | Energy Density (Wh/kg) | Capacity (mAh) | Test Date | Anomaly Flag |
|---|
* TB-001a and TB-001b excluded from production batch; both figures retained in published dataset. Mean reported (452 Wh/kg) is non-excluded population. Full-population mean including TB-001 cells: 448 Wh/kg.
| Temperature | Capacity (% of 23°C baseline) | n | Notes |
|---|---|---|---|
| –20°C | 78% ± 4% | 5 | After 2h stabilisation at temperature |
| 0°C | 91% ± 3% | 5 | After 2h stabilisation |
| 23°C | 100% (reference) | 24 | Primary test temperature |
| 40°C | 98% ± 2% | 5 | After 2h stabilisation |
| 60°C | 96% ± 3% | 5 | After 2h stabilisation |
| 80°C | 93% ± 4% | 5 | After 2h stabilisation; no thermal event observed |
5 cells active · Cell #4 (TB-004) terminated at 1,800 — shown as dashed terminated line · Updated daily · Raw CSV available
↓ Cycle Life Raw Data (CSV)Description: 2 cells in production batch B produced 441 and 443 Wh/kg — below the 450 Wh/kg guaranteed threshold. Root cause: electrolyte layer thickness variation at cell edge, exceeding ±2% tolerance. Identified post-fabrication via EIS impedance mapping.
Disposition: Both cells excluded from production batch. Batch B accepted after re-test of edge cells from revised fabrication run. Electrolyte deposition parameter adjusted.
Description: Cell #4 of 6 removed from cycling rig at cycle 1,800. Separator crack visible on post-disassembly optical inspection. Capacity at point of removal: 81% of initial — above 80% EOL threshold. Crack location: lateral edge, not at electrode-electrolyte interface.
Disposition: Cell removed; full disassembly documented with optical and SEM imaging. SEM images of crack site available upon request. Remaining 5 cells continue active cycling. Crack mechanism tentatively attributed to edge stress concentration at cell clamp interface; fixture design under review.
Description: 1 cell of 10 showed edge delamination at cycle 850 of the flex durability test. Delamination at cell edge — not at the electrode-electrolyte interface. Capacity at time of delamination: 92% of initial — above 80% criterion. Not a failure by the capacity criterion; classified as anomaly due to visible structural change.
Disposition: Cell excluded from the reported flex cycle count. 9 of 10 cells reached 1,000 cycles without structural anomaly or capacity drop below 95%. Delamination cause under investigation — likely edge-seal geometry at small bend radius. Edge-seal redesign in progress.
Researchers who identify inconsistencies in published methodology, dispute reported results, or propose alternative test approaches are invited to submit findings via hello@nrutseab.com. Submissions receive a written response from Nrutseab's research team within 30 days. Substantive challenges are referenced in the validation documentation with the challenger's name and institution — or anonymised on request.
Independent findings that contradict Nrutseab's results are published in the same location as the original data — not in a separate, obscured location. The External Challenges log will be maintained at research.nrutseab.com.
→ hello@nrutseab.com · 30-day response SLAThe adaptation to solid-state is documented but not endorsed by IEC. This is the adaptation most likely to be challenged. Response: the adaptation methodology is published in full at §2.1 of the protocol document. Researchers who believe the adaptation introduces systematic bias are explicitly invited to quantify and publish the bias. A methodology committee review is formally invited.
No IEC or ASTM standard exists for solid-state flexible cell flex durability. The Nrutseab protocol is internally defined — making it the most challengeable measurement in the dataset. Response: protocol is published in full; fixture design document is public; researchers are invited to propose a peer-reviewed alternative protocol. Any alternative adopted before July 2026 will be retested.
Calendar aging (capacity loss during storage) is a significant limitation for applications requiring multi-year service life — EV, grid storage, aerospace. No calendar aging data is currently published. Testing is in progress and will publish under a separate protocol when sufficient time-series data exists. This limitation is disclosed in Section 6 of the protocol — not hidden in a footnote.
0.1C is a very slow discharge rate that yields near-maximum energy density. At 1C discharge (the cycle-life test rate), energy density will be lower. The C-rate dependence data is published in the data section. Applications requiring fast discharge (EVs, power tools, high-drain drones) must use the C-rate dependence data, not the 0.1C headline figure. This is disclosed prominently — not buried.
6 cells is a limited sample for cycle life characterisation. Response: the sample size was chosen to achieve <2% relative uncertainty on projected EOL at expected variance levels — power analysis is available in the full protocol. The live counter (n=5 now active) is continuously updated. The limitation is disclosed; the sample size justification is documented. Researchers may propose expanded cycle testing using samples from July 2026.
This is the most fundamental limitation of the entire dataset. The data is self-reported. Response: this limitation is the explicit premise of the entire open validation framework — it is the first statement on every page. Independent verification begins July 7, 2026 when physical samples are distributed. The methodology is published precisely so researchers can design their replication before receiving samples.
In order of what they will verify before attending Nrutseab 77 or engaging further. This portal is designed to satisfy every step below.
research.nrutseab.com loads, looks like a data portal — not a product page. No stock images. No "revolutionary" language. Monospace data values. Timestamps visible.
→ research.nrutseab.comThey look for: a specific number, a protocol reference, something that looks like it could be wrong. Answer: 452 Wh/kg ±8 Wh/kg (k=2) · n=24 · IEC 62660-1 adapted. All three present immediately.
→ See data section aboveIf they see only means, they will stop reading. Individual cell data for all n=24 is published in the table above. If the individual values match the reported mean and uncertainty, they have passed the first test.
→ n=24 individual cell table aboveIf there are no anomalies, they will assume the data is fabricated. Three real anomalies (TB-001, TB-004, TB-007) with honest dispositions and stated effects on reported figures are the right number at this testing stage.
→ Anomaly log aboveThey will search "Neware BTS-4000" and verify that Intertek UKAS is a real accreditation body for this instrument type. Both are verifiable. Intertek UKAS calibration for battery cyclers is a standard commercial service.
→ Instrument register aboveThey open it. Check if the individual cell data matches the summary statistics. If it does, Nrutseab has passed the first credibility test. The CSV contains all n=24 cells including anomaly-flagged cells.
→ Download CSV (no registration)They check the Toyota (~330 Wh/kg) and QuantumScape (~380–400 Wh/kg) figures against their own knowledge. Both figures are sourced from the companies' own press releases and SEC filings — and are accurate.
→ Context table aboveIf all anomalies have the same timestamp, or no timestamps, or placeholder text, they leave immediately. TB-001 (Q4 2025), TB-004 (Q1 2026), TB-007 (Q1 2026) — different dates, different parameters, different dispositions.
→ Timestamped anomaly log aboveThis is the conversion event. Register for the July 7, 2026 event in San Francisco to interact with physical hardware and receive samples for independent verification. Pre-event technical briefings with Hilde available under NDA.
→ events.nrutseab.com/77A transparent composite that changes mechanical stiffness in response to impact rate — no external actuation, no power source, no moving components. Sample access from July 2026. hello@nrutseab.com
The +68% energy absorption figure requires a defined baseline. The baseline is a static lattice structure of equivalent mass, prepared under the same conditions. Absolute absorption values: Jelly — 4.8 J/cm² mean (±0.3 J/cm², k=2); baseline lattice — 2.9 J/cm² mean (±0.2 J/cm², k=2). Researchers proposing an alternative baseline are invited to publish independent results using samples available from July 2026 — Nrutseab will reference findings regardless of outcome. .
Under low shear rates — handling, flexing, slow loading — particle interactions are minimal and the matrix is compliant. Under high shear rates from rapid impact, hydrodynamic forces drive reversible particle cluster formation (a jamming transition), dramatically increasing apparent viscosity and stiffness in the impact zone. This distributes kinetic energy across the material volume rather than concentrating stress at the impact point. The particle aspect ratio, surface functionalisation chemistry, volume fraction, and matrix cross-link density are documented in the published methodology.
Optical transparency in particle-filled composites requires that particle size, refractive index, and inter-particle spacing avoid Mie scattering in the visible range (380–780 nm). Jelly achieves ≥92% visible light transmission through particle dimensions held below the Mie scattering threshold and refractive index matching of the particle surface coating to the polymer matrix.
All figures from internal testing. Independent third-party verification begins July 2026.
| Parameter | Internal Result | n | Uncertainty (k=2) | Test Standard | Anomalies |
|---|---|---|---|---|---|
| Visible light transmission | 93.4% mean | 8 | ±0.8% | ASTM D1003; reference panel geometry; integrating sphere colorimeter | JL-002 — 1 sample edge haze (4.2%); excluded; edge seal contamination |
| Energy absorption vs. baseline | +68% over lattice | 15 | ±7% | Internal drop tower; baseline = static lattice structure equivalent mass; methodology tba | JL-005 — baseline sample #3 micro-fracture pre-test; n=4 baseline used |
| Impact transition onset | <100 ms (mean 62 ms) | 20 | ±18 ms | High-speed camera 10,000 fps; piezoelectric force plate; onset = 10% departure from low-rate stiffness | None. Onset definition-dependent — raw force-time data published. |
| Fatigue resistance | >107 cycles (ongoing) | 6 | Live; no failure | ASTM D7774; 50% peak load; 10 Hz; R=–1; ambient temperature | None. Live count updated daily. |
| Flexural modulus | 4.2 GPa mean (3.5–5.0 GPa) | 5 | ±0.4 GPa | ASTM D790; 3-point bending; 50mm span; 2mm/min crosshead | Range reflects intentional composition variation. Composition documented per sample. |
Description: 1 of 8 samples showed localised haze at sample edge. ASTM D1003 haze measurement: 4.2% (all others <1.5%). Traced to edge seal contamination during sample preparation.
Disposition: Excluded from mean. Edge seal procedure revised. Revised sample run shows <1.2% haze.
Description: Baseline lattice sample #3 of 5 showed pre-test micro-fracture on optical inspection. Removed from baseline set; n=4 used for baseline calculation.
Disposition: Baseline n reduced from 5 to 4. Micro-fracture documented with optical imaging. Effect quantified and published.
Superconducting transmon architecture · 1 billion physical qubits target · ~1 million logical qubits (surface code d=31 at 99.9% fidelity) · 7.5 × 7.5 cm modular form factor · The Block onboard power · Lifetime warranty.
Pre-order: screner.nrutseab.com/butter · Delivery 2028–2029
Any domain expert reading this page will already know these numbers. Showing them before Butter's claims demonstrates that Nrutseab is not operating in a credibility bubble.
| System | Physical Qubits | Status (Apr 2026) | Note |
|---|---|---|---|
| IBM Condor | 1,121 | Production (2023) | Largest deployed transmon system at time of this spec. |
| Google Willow | 105 | Research (2024) | Demonstrated below-threshold error correction — qualitatively more significant than qubit count. |
| IBM Heron roadmap | ~4,000 (target) | Roadmap | IBM public roadmap. Fidelity improvement over qubit-count scale is explicit IBM strategy. |
| Butter (Nrutseab) | 1,000,000,000 (target) | Architecture published | Gap from best demonstrated (1,121) to Butter target is ~6 orders of magnitude. This page does not hide this gap — it explains the engineering approach. |
Superconducting transmon qubits are established technology. The challenge of reaching billion-qubit scale is three specific engineering constraints — each documented in the published architecture specification.
Each qubit in a conventional system requires individual microwave control lines from room temperature to millikelvin. At billion-qubit scale this is physically impossible with individual wiring. Butter's approach: multiplexed microwave control — shared channels address qubit sub-arrays through pulse-level addressing. Specific multiplexing ratio under patent protection; approach analogous to CMOS multiplexed addressing.
Each control line conducts heat into the millikelvin stage. Conventional dilution refrigerators support tens to low hundreds of lines before thermal load prevents reaching operating temperature. Butter's approach: distributed cold-finger array from a centralised refrigerator plant serving the module stack — each 7.5 × 7.5 cm module maintains <10 mK. Thermal modelling tba.
Multi-chip quantum systems require interconnects that preserve quantum coherence at chip boundaries. Butter's approach: optical cryogenic links between modules for inter-module entanglement operations. Each module maintains its own coherence environment; inter-module operations occur through defined exchange windows. A local decoherence event does not cascade.
This calculation is standard in the quantum computing literature. It is published in detail to allow experts to challenge the assumptions. The 1 million logical qubit figure is critically dependent on achieving 99.9% two-qubit gate fidelity.
| Parameter | Value | Basis / Challenge Point |
|---|---|---|
| Two-qubit gate fidelity (target) | 99.9% | Google best demonstrated: ~99.7% (2024). Butter target is 0.2% above best demonstrated. Gap noted — primary fabrication challenge. |
| Error correction code | Surface code | Leading practical code for superconducting qubits. Documented in Fowler et al. (2012) and subsequent literature. |
| Code distance | d=31 | Below fault-tolerance threshold at 99.9% fidelity. Physical qubit overhead: ~1,000 per logical qubit. |
| Physical qubits per logical qubit | ~1,000 | At 99.5%: ~5,000 per logical qubit. At 99.0%: ~50,000. Sensitivity table available upon request. |
| Logical qubits at 1 billion physical | ~1,000,000 | Sufficient for RSA-2048 factoring, large-scale molecular simulation, optimisation problems currently intractable on classical systems. |
Architecture and engineering approach published now. Fabrication and performance data publish as milestones are reached. All milestones open to external challenge via hello@nrutseab.com.
| Milestone | Status | Target | Publishes |
|---|---|---|---|
| Architecture specification | COMPLETE | May 1, 2026 | research.nrutseab.com/butter — tba |
| Control multiplexing preprint | IN PREPARATION | May 10, 2026 | Modular Quantum Network Architectures for Billion-Qubit Systems — Nature Communications submission + arXiv preprint |
| Single-module qubit array fabrication | IN PROGRESS | Q3 2026 | Fabrication data, SEM/optical imaging, qubit count per chip, yield statistics |
| Single-qubit gate fidelity | NOT STARTED | Q3 2026 | Randomised benchmarking data |
| Two-qubit gate fidelity | NOT STARTED | Q3 2026 | Process tomography results. Critical figure. Logical qubit count revision published simultaneously if target is not met. |
| Multi-module coherence | NOT STARTED | Q4 2026 | Cross-module entanglement fidelity; optical link performance data |
| Surface code logical qubit demonstration | NOT STARTED | Q1 2027 | Error rate vs. code distance data; below-threshold demonstration. Equivalent milestone to Google Willow (2024) at larger scale. |
| Pre-order delivery begins | NOT STARTED | 2028–2029 | Commercial units. Lifetime warranty incl. hardware upgrades. Pre-orders refundable if milestones not met. |
All three systems — The Block, Jelly, and Butter — will be demonstrated in live conditions. This is the first opportunity for press and invited researchers to interact with physical hardware, conduct hands-on tests, and engage directly with Nrutseab's research team.
The event is open to accredited press, qualified researchers, and institutional partners. Priority credentials for researchers with pre-submitted test protocols. Pre-event technical briefings with Hilde (Founder and Research Lead) are available for accredited press under NDA prior to July — contact hello@nrutseab.com.
Subsequent events: Hangzhou and Seoul through August 2026.
Priority allocation: researchers with submitted test protocols proposing methods not covered by Nrutseab's internal testing. No institutional affiliation required. Results from sample-based research published at research.nrutseab.com regardless of outcome.