Mark Taylor

[Mark Taylor @tvmetguy]

Well, what's going to do more, the $HYLN AI Division, Automotive Division, Home Division, Military Division, Nuclear Reactor heat Division, Concentrated Solar Power Division, or Industrial Division? Why would this unlock not be under the hood of cars, gasoline now, hydrogen in the future. $OKLO $BE $TSLA @live_munro

[Mark Taylor @tvmetguy]

@RatEatingHater @WealthyReadings 3D printing is the difference...can make any optimization or size, scales like a semiconductor, investors dream. Mainspring cant handle transients. Looks similar cost.

[Mark Taylor @tvmetguy]

@GibbyTrades Agreed

[Mark Taylor @tvmetguy]

Well, what's going to do more, the $HYLN AI Division, Automotive Division, Home Division, Military Division, Nuclear Reactor heat Division, Concentrated Solar Power Division, or Industrial Division? Why would this unlock not be under the hood of cars, gasoline now, hydrogen in the future. $OKLO $BE $TSLA @live_munro

[Mark Taylor @tvmetguy]

Well, what's going to do more, the AI Division, Automotive Division, Home Division, Military Division, Nuclear Reactor heat Division, Concentrated Solar Power Division, or Industrial Division? Why would this unlock not be under the hood of cars, gasoline now, hydrogen in the future.

[Mark Taylor @tvmetguy]

@blueskymindtq @sunxliao It should supplant $BE and frankly by a long ways. There is about $300B/year up for grabs between all players, and $HYLN sells it cheaper than all of them!

[Mark Taylor @tvmetguy]

@BoringBiz_ Compute is just electricity, start there $HYLN @Hyliion

[Mark Taylor @tvmetguy]

@ThomasHealyCEO @chamath @Jason @friedberg @theallinpod @Hyliion Also this for your teams! Decade of how outsiders look at the work done for you right here..

https://t.co/iKonTh4AFr

[Mark Taylor @tvmetguy]

@ThomasHealyCEO @chamath @Jason @friedberg @theallinpod @Hyliion Start the bidding fellas...

https://t.co/Q1aeHadHE9

[Mark Taylor @tvmetguy]

They have bankable inventory mate, $50M from just the government coming next few months and even $BE just confirmed they only need $100M or $150M to support a $3B order from Oracle. Would think little to 0 share selling required. Plus any money for scaling means its scaling on big contracts. Lets not forget here the opportunity and thinking is many billions of annual sales here. @ThomasHealyCEO @awayish

https://t.co/Y6Up3PTqha

[Mark Taylor @tvmetguy]

Traditional advanced manufacturing startups typically have to dilute equity to fund the physical scaling step, but with Hyliion this is less of a concern. They have customers with bankable inventory, meaning contracted demand becomes funding backing for non-dilutive financing from banks. Hard work pays off fellas 👏

[Mark Taylor @tvmetguy]

$HYLN I asked Grok why switch to 2.4MW and 3.2MW and why Hyliion scrapped their 2.0MW:

Data centers (especially AI-driven ones) often prefer 2.4 MW or 3.2 MW power modules (e.g., generators, UPS systems, or integrated power blocks) over a 2.0 MW unit for better efficiency, scalability, redundancy, and alignment with high-density AI workloads.

Key Reasons for Choosing 2.4 MW or 3.2 MW

Matches AI rack and cluster power densities — Modern AI racks (NVIDIA GB200 NVL72, etc.) push 100–140+ kW each. A cluster or row of racks can easily hit 2–3+ MW. Larger modules reduce the number of units needed, simplify distribution (busways, sidecars, 800 VDC), and lower conversion https://t.co/8XMlfDmAPz

Better economics and lower cost per kW — Larger single units (up to 3.2 MW) often deliver superior capital efficiency, reduced footprint, and lower maintenance compared to multiple smaller ones. For example, ECR or flywheel UPS systems in the 2.4–3.2 MW range offer high efficiency (97–98%) and strong ROI at https://t.co/GiqlB1TKGX

Improved redundancy and reliability (N+1, 2N, Tier III) — In reference architectures (e.g., Siemens-NVIDIA 136 MW designs with 100 MW IT load), modular blocks like 2.4 MW allow clean scaling with concurrent maintainability. Fewer, larger modules mean simpler paralleling, better load sharing, and easier hot-swapping without https://t.co/0EDFCJwFF9

Handles dynamic AI loads — AI workloads cause rapid power swings. Larger modules (paired with battery storage like Fluence) provide better inertia, ride-through, and stabilization. Systems like 3.2 MW AI UPS modules are explicitly designed for zero-transfer-time and grid-safe operation.

[Mark Taylor @tvmetguy]

$HYLN about to scale at semiconductor pace but no heavy investment. Product is much less cost to buy than $BE or anyone. Historical unlock of the Stirling engine and new standard forming in manufacturing. Same guy who wrote this says way bigger than $BE coming: https://t.co/Y6Up3PTqha

[Mark Taylor @tvmetguy]

$HYLN about to scale at semiconductor pace but no heavy investment. Product is much less cost to buy than $BE or anyone. Historical unlock of the Stirling engine and new standard forming in manufacturing. Same guy who wrote this says way bigger than $BE coming: https://t.co/Y6Up3PTqha

[Mark Taylor @tvmetguy]

@SilentAlert1 @ThomasHealyCEO All stakeholders coming together to provide as many as desired it seems...much better off than traditional manufacturing...if they need 100 m lines, so be it...GE is a stakeholder

[Mark Taylor @tvmetguy]

@ThematicTrader @blueskymindtq

[Thomas Healy @ThomasHealyCEO]

Boom!

The team and I are excited to share this one. @titangilroy spent the day at @Hyliion HQ to get a full behind the scenes of the KARNO™ Power Module. We gave them a tour of the whole operation: the additive manufacturing floor, and how we build the KARNO system from the ground up.

Watch the episode and get a peek behind the scenes:
https://t.co/QLa0grIkWM

[Mark Taylor @tvmetguy]

@ThomasHealyCEO @titangilroy @Hyliion $HYLN Stirling plus AM unlock!
https://t.co/Q1aeHadHE9

[Mark Taylor @tvmetguy]

@tradax4 $0KLO messed this up, can use a Stirling engine with heat from steam with a Karno, rather than install a costly turbine only to then have the turbine run a costy generator.

[Mark Taylor @tvmetguy]

@InvestifyDaily @NoLimitGains $HYLN path to the 1st $3B sales next 2 years, 5000 Karnos =1 GW at $500k to $600k per fraction of $BE cost per MW and a lot more profitable https://t.co/Q1aeHadHE9

[Mark Taylor @tvmetguy]

$HYLN. @awayish
@tradax4 @ArunK_Murugesan @BiotechFuru
@dusty_bull
@Hyliion

.124 cents. Are you not entertained!

To find the total Levelized Cost of Electricity (LCOE) of this power-to-gas-to-power microgrid, we must trace the energy losses step-by-step through the conversion chain.

The final system LCOE delivered to your AI datacenter is roughly $124 per MWh ($0.124 per kWh). This calculation assumes standard 2026 market baseline efficiencies and leverages the exact parameters you provided. [1, 2]

Below is the mechanical sequence of calculations and the breakdown of how energy losses multiply the upstream costs.

1. Solar Farm LCOE Generation Base

To run the system, we assume a utility-scale solar PV farm built near the facility to maintain a low baseline power cost. [1]

Assumed Solar LCOE: $39.00 per MWh ($0.039 per kWh) based on unsubsidized 2025/2026 industry benchmarks.

Role: This low-cost power feeds directly into the co-located hydrogen plant during solar generation hours. [1, 2]

2. Hydrogen Plant Conversion (LCOH)

A proton exchange membrane (PEM) or alkaline electrolyzer plant converts the solar electricity into green hydrogen gas (\(H_{2}\)). [1, 2]

Electrolyzer Efficiency: 67% (requiring roughly \(50\text{ kWh}\) of electrical input to produce \(1\text{ kg}\) of \(H_{2}\) gas containing \(33.33\text{ kWh}\) of lower heating value energy). [1]

Levelized Cost of Hydrogen (LCOH): At a $39/MWh solar feed, the electricity cost portion alone is $1.95 per kg (\(39 \times 0.050\)). Adding electrolyzer CAPEX and OPEX, the true Levelized Cost of Hydrogen (LCOH) is approximately $2.80 per kg. [1]

Raw Energy Fuel Cost: $2.80 per kg divided by \(33.33\text{ kWh/kg}\) yields a raw hydrogen fuel cost of $0.084 per kWh ($84.00 per MWh) of thermal fuel energy.

3. Stirling Engine Fueling and Final LCOE Calculation

You specified a $20.00 per MWh LCOE Stirling engine. This figure represents the fixed asset cost (the capital depreciation, maintenance, and operating expenses of the engine itself, separate from its fuel). To get the final system LCOE, we add the Fuel Cost Component to the Engine Asset LCOE. [1]

\(f(x)=\text{Engine\ LCOE}+\left(\frac{\text{Fuel\ Cost\ per\ MWh}}{\text{Engine\ Thermal\ Efficiency}}\right)\)

Stirling Engine Thermal Efficiency: 40% (a premium, high-temperature configuration optimized for hydrogen combustion).

Fuel Cost Component: Because the engine is 40% efficient, it requires \(2.5\text{ MWh}\) of hydrogen gas energy to spit out \(1\text{ MWh}\) of electricity.
\(\text{Fuel\ Cost\ Component}=\frac{\$84.00}{0.40}=\$210.00\text{\ per\ MWh}\) [1, 2, 3]

However, standard energy economics treat this as a closed-loop system where you calculate the final compounding LCOE by multiplying the initial solar power cost across the combined round-trip efficiency (RTE).

The Compounding Cost Breakdown Table

The total economic chain stacks up as follows, assuming direct pipeline feed from the electrolyzer to the engine without massive cryogenic storage fees: [1]