Shardwood Protocol ($SHWD) — Product Thesis, Token Distribution and Execution Risk

From attached source: message.txt

I researched Shardwood Protocol from a product, token-distribution and execution-risk perspective.

Shardwood is attempting to build a distributed AI inference network using independent GPUs.

The basic idea is that a large AI workload does not always need to run on one centralized machine. Instead, compatible GPUs can participate in an inference route, with different workers handling stages of the workload.

What makes the project interesting to me is that Shardwood is not only describing a marketplace for unused GPUs.

The more ambitious goal is to make distributed inference measurable and accountable enough that independent operators can participate in the same network.

According to the project's own materials, workers can participate in inference lanes, handoffs can be signed, and the system is designed to record which operator processed each stage.

However, there is an important limitation:

proving who handled a computation is not the same as proving that the computation itself was fully correct.

The current architecture appears stronger at attribution, ordering and metering than at full cryptographic verification of every AI calculation.

That distinction matters when evaluating the project.

PRE-TGE STATUS

Shardwood is still genuinely pre-TGE.

The project's own materials state that:

• the token is not deployed • there is no live token sale • contracts are still pre-audit • contract addresses are not yet final • the token launch depends on product usage, legal review and technical readiness

Because of that, I think Shardwood is more useful to study before market price and token speculation begin influencing the narrative.

TOKENOMICS

The proposed genesis supply is:

420,690 SHWD

The proposed allocation is:

35% — Network contributor incentives

17% — Ecosystem and community

17% — Core contributors and future hires

10% — Development / creator treasury

8% — Liquidity and market access

8% — Community bootstrap / security

5% — Technical partners and advisors

One of the most important points for me is that total supply and circulating supply are not the same thing.

The project has published proposed vesting schedules, but the exact TGE circulating supply has not yet been published.

There is also no public token price.

Therefore, there is currently no defensible FDV.

Any exact FDV calculation today would require using an assumed price rather than a verified market or sale price.

UNLOCK AND DILUTION RISK

The unlock structure is more important than the headline total supply.

The 35% network contributor allocation is proposed to release linearly over a long period.

The 8% community bootstrap/security allocation also begins releasing from T0.

Meanwhile, the 10% creator treasury is proposed to be available from T0 without a normal vesting schedule.

That does not mean the entire 10% automatically becomes circulating supply.

But it does create an important question:

Who controls those tokens, what restrictions exist, and how transparent are treasury movements?

At approximately the one-year point, several other allocations begin leaving their cliffs, including ecosystem, team and partner allocations.

For me, this means the most important dilution risk may not be TGE itself.

The more important period may begin later, when multiple previously locked allocations start releasing at the same time that contributor incentives are already entering circulation.

Even if maximum supply remains fixed, circulating supply can still expand significantly.

PRODUCT AND EXECUTION RISK

The largest product risk is whether distributed inference remains competitive outside a controlled environment.

Splitting AI workloads across independent machines can introduce:

latency

network instability

hardware compatibility issues

privacy concerns

coordination overhead

fault-detection problems

A decentralized GPU network is only economically useful if the cost and performance remain competitive with centralized infrastructure.

That means I would want to see reproducible benchmarks, not only architecture diagrams.

I would also want to understand how the system performs when a worker fails, becomes slow, or provides incorrect output.

DECENTRALIZATION RISK

Another area I would watch is how the network moves away from coordinator-led infrastructure.

Early-stage projects often need more centralized coordination to make the product usable.

That is understandable.

But the important diligence question is whether that control decreases over time.

I would want clear answers to:

Who controls routing?

Who approves workers?

Who controls settlement?

What happens if the coordinator goes offline?

Can new operators eventually participate permissionlessly?

If those functions remain permanently centralized, the network may be technically distributed without being meaningfully decentralized.

TEAM VERIFICATION

The project's technical materials are relatively detailed and openly identify several components as pre-audit, pending, experimental or not yet deployed.

I consider that transparency a positive signal.

However, I could not verify enough primary-source information to treat the team itself as fully de-risked.

Before assigning higher confidence, I would want stronger verification of the people responsible for:

protocol engineering

economic design

treasury control

security

and long-term governance

MY CURRENT VIEW

I think Shardwood is worth following because it is trying to build infrastructure around a real resource: distributed GPU compute.

But I would not evaluate the project primarily from the token.

The stronger evidence would be:

real inference workloads

reproducible performance benchmarks

recurring paying users

audited contracts

transparent treasury controls

clear team verification

and a final TGE circulating-supply schedule

If the network proves real demand first and introduces the token as a coordination mechanism second, the design becomes much more convincing.

If the token launches before the infrastructure and demand are proven, token-distribution and incentive risks become much more important.

Disclosure:

I currently have no SHWD position because the token is not deployed.

This research was prepared as part of The Republic / Eastern Campaigns research mission.

Primary sources:

https://shardwood.tech/

https://shardwood.tech/papers/shardwood-concept-paper-v0.2.pdf