Ethereum Fusaka Upgrade Explained

The Ethereum Fusaka upgrade is the network-wide update that went live on Ethereum mainnet on December 3, 2025. Its headline feature, PeerDAS, changed how Ethereum nodes handle the data that layer 2 networks post to the main chain, so Ethereum could carry far more of that data without forcing every node to download all of it. The result was more room for rollups such as Arbitrum, Optimism and Base, and lower costs for their users.

This guide explains the Ethereum Fusaka upgrade in plain English: what it changed, why it mattered, what the Blob Parameter Only forks that followed it did, what it meant for ETH holders, stakers and developers, and how it fits into Ethereum’s longer roadmap. If you hold ETH, you did not need to do anything, and anyone who told you otherwise was attempting a scam.

If you are new to Ethereum, our crypto for beginners guide explains blockchains, wallets and fees first.

Fusaka at a Glance

Before going into detail, here are the key facts in one table. Dates and figures come from the Ethereum Foundation’s mainnet announcement and ethereum.org.

Item	Detail
Name	Fusaka, combining Fulu (consensus layer) and Osaka (execution layer)
Mainnet activation	December 3, 2025, at 21:49:11 UTC (epoch 411392)
Previous upgrade	Pectra, May 2025
Main feature	PeerDAS (EIP-7594), data availability sampling for blobs
Blob limits at activation	Target 6, maximum 9 per block
BPO1	December 9, 2025: target 10, maximum 15
BPO2	January 7, 2026: target 14, maximum 21
Default gas limit	Raised to 60 million (EIP-7935)
Per-transaction gas cap	16,777,216 gas (EIP-7825)
New signature support	secp256r1 precompile for passkeys (EIP-7951)
Action needed by ETH holders	None
Next major upgrade	Glamsterdam

What Is Fusaka, and Where Does the Name Come From?

Ethereum runs as two connected layers of software. The execution layer processes transactions and smart contracts, and the consensus layer, also called the beacon chain, handles proof of stake and agrees on which blocks are valid. Each upgrade therefore has two names, one for each layer.

For this upgrade, the execution layer part was named Osaka, following Ethereum’s custom of naming execution upgrades after cities that have hosted its Devcon conference. The consensus layer part was named Fulu, after a star, following the tradition of naming consensus upgrades after stars. Combine the two and you get Fusaka.

How Fusaka fits with earlier upgrades

Fusaka is the latest step in a series of upgrades that have steadily shifted Ethereum toward a “rollup-centric” design, where most everyday transactions happen on layer 2 networks that then post their data back to Ethereum.

Upgrade	Year	What it did
The Merge	2022	Moved Ethereum from proof of work to proof of stake
Shapella	2023	Allowed stakers to withdraw their ETH
Dencun	2024	Introduced blobs (EIP-4844), a cheaper way for rollups to post data
Pectra	2025	Raised the blob target to 6, improved staking and wallet features
Fusaka	December 2025	Added PeerDAS so blob capacity could grow much further
Glamsterdam	Planned	Focuses on block building, parallel processing and state pricing

Why Ethereum Needed Fusaka

To understand the Ethereum Fusaka upgrade, it helps to understand the problem it solved. Rollups bundle thousands of transactions together and post the data to Ethereum so anyone can check it. Since the Dencun upgrade in 2024, they have posted this data in “blobs,” special data packets that are cheaper than regular transaction data and are deleted after a few weeks.

Blobs worked well, and demand grew. But there was a catch. Every full Ethereum node had to download and store every blob. Raising the number of blobs per block would have increased the bandwidth and storage needed to run a node, pushing out home operators and making the network more centralized. Ethereum needed a way to add more blob space without making each node do proportionally more work.

The data availability problem

When a rollup posts data, the network must make sure that data was actually published, so that anyone can rebuild the rollup’s state and challenge errors. This is called data availability. Before Fusaka, the only way a node could be sure of data availability was to download everything. Fusaka introduced a smarter method.

PeerDAS: The Heart of Fusaka

PeerDAS, short for Peer Data Availability Sampling and defined in EIP-7594, lets nodes confirm that blob data is available by downloading only small samples of it rather than the whole thing.

How PeerDAS works

The idea is easier to grasp than the math behind it. Each blob is extended with extra coded data, a technique called erasure coding, and then split into columns. Those columns are spread across 128 subnets, which are groups of nodes in Ethereum’s peer-to-peer network.

According to ethereum.org, the key points are:

• Each regular full node keeps about one-eighth of the data, by subscribing to 4 of the 128 subnets.
• Any 50% of the extended data is enough to rebuild the rest, so the network can recover data even if many nodes go offline.
• The chance of wrongly accepting missing data is tiny, estimated at roughly one in 10^20 to one in 10^24.
• Nodes running validators hold more data, subscribing to at least 8 subnets, and very large stakers become “supernodes” that store every column.

An everyday comparison

Imagine a huge book that thousands of people must confirm was printed in full. Before Fusaka, every person had to read the whole book. With PeerDAS, each person reads a few random pages, and because the book was printed with built-in redundancy, the group as a whole can be confident every page exists and can be reconstructed. Each reader does far less work, yet the group is just as sure.

What PeerDAS made possible

Because each node handles only a fraction of the blob data, Ethereum can increase blob capacity without a matching rise in the load on each node. Ethereum.org describes this as enabling a theoretical increase of up to eight times in blob throughput. It also notes that for regular full nodes, download and disk requirements for blob data dropped substantially, roughly 80%.

Blob Parameter Only (BPO) Forks

PeerDAS created room for more blobs, but Ethereum’s developers did not raise the limits all at once. Instead, Fusaka introduced Blob Parameter Only forks, defined in EIP-7892.

What a BPO fork is

Normally, changing Ethereum’s rules requires a full coordinated upgrade, with months of testing and new client releases. A BPO fork is a lightweight upgrade that changes only two numbers: the blob target and the blob maximum per block. Clients are programmed in advance to switch to the new numbers at a set time, so the network can grow capacity step by step and pause if problems appear.

The first two BPO forks

The Ethereum Foundation scheduled two BPO forks shortly after Fusaka went live.

Stage	Date (UTC)	Blob target	Blob maximum
Fusaka activation	December 3, 2025	6	9
BPO1	December 9, 2025	10	15
BPO2	January 7, 2026	14	21

By January 2026, the blob target had more than doubled compared with the level at Fusaka’s launch, and was nearly five times the target of 3 set when blobs were first introduced in 2024. Future BPO forks can raise the numbers again as the network shows it can handle more.

The blob fee floor

Fusaka also added EIP-7918, which ties the minimum blob fee to the cost of regular execution. Before this change, when blob demand was low, the blob fee could fall almost to zero, which meant the fee stopped signaling how busy the network was. The new floor keeps blob pricing meaningful and makes sure rollups pay a fair share of the resources their data uses.

Other Changes in Fusaka

PeerDAS got most of the attention, but the Ethereum Fusaka upgrade bundled together a dozen other improvements. Most affect developers and node operators rather than everyday users.

EIP	What it does, in simple terms
EIP-7594	PeerDAS: nodes sample blob data instead of downloading all of it
EIP-7642	eth/69 networking: drops old pre-Merge data from peer messages, helping nodes save space
EIP-7823	Sets an upper limit on the size of inputs to the MODEXP operation
EIP-7825	Caps the gas a single transaction can use at 16,777,216
EIP-7883	Raises the gas cost of MODEXP to better match its real cost
EIP-7892	Creates Blob Parameter Only forks
EIP-7910	Adds an eth_config method so operators can check fork settings
EIP-7917	Makes upcoming block proposers known in advance
EIP-7918	Sets a floor on the blob base fee, tied to execution costs
EIP-7934	Caps the size of execution blocks
EIP-7935	Raises the default gas limit to 60 million
EIP-7939	Adds a “count leading zeros” (CLZ) instruction to the EVM
EIP-7951	Adds native support for secp256r1 signatures, used by passkeys

A higher gas limit for layer 1

EIP-7935 set the default gas limit to 60 million, which the Ethereum Foundation said core developers believed layer 1 could safely handle. A higher gas limit means more room for transactions on Ethereum itself, not just on rollups.

A cap on single transactions

EIP-7825 limits any single transaction to 16,777,216 gas. This prevents one transaction from filling an entire block, which supports future plans to process transactions in parallel and keeps blocks more predictable. Very large contract deployments or complex batches may need to be split into smaller transactions.

Passkeys and smarter wallets

EIP-7951 adds a built-in way for Ethereum to check secp256r1 signatures, the type used by passkeys and by the secure chips in many phones and laptops, such as Apple’s Secure Enclave and Android’s Keystore. This makes it easier to build wallets that let users sign with Face ID, a fingerprint or a hardware security key instead of managing a seed phrase. It does not replace seed phrases overnight, but it opens the door to simpler and safer wallet designs.

Preparing for preconfirmations

EIP-7917 lets the network know which validators will propose upcoming blocks. This sounds technical, but it supports “preconfirmations,” where a block proposer promises in advance to include a transaction. Over time, that could give users faster assurance that their transactions will go through.

What the Ethereum Fusaka Upgrade Means for You

The impact of the upgrade depends on how you use Ethereum. For most people, nothing changed in how they hold or send ETH.

ETH holders

If you hold ETH on an exchange or in a wallet, you did not need to do anything. Your balance and address stayed the same. Ethereum.org explicitly warns that anyone asking you to “upgrade” or “migrate” your ETH because of Fusaka is trying to scam you. There is no new ETH token and no swap.

Layer 2 users

Rollup users were the main beneficiaries. More blob space means rollups can post data more cheaply and handle more transactions. Reports after the BPO forks, such as a January 2026 analysis from BlockEden, described typical transaction costs on major layer 2 networks in the range of fractions of a cent, and cited a drop of about 40% in the cost of posting rollup data to layer 1. These figures come from secondary sources and change with demand, so treat them as a general picture rather than a guarantee.

Stakers

Validators had to update both their beacon node and validator client before the fork. Because nodes that run validators store more blob columns than regular nodes, stakers with large balances took on more data work, though generally still less than before PeerDAS. Running up-to-date clients remains the most important task for any staker.

Node operators

Operators of non-staking nodes needed to update both their execution and consensus clients. After that, many saw lower bandwidth and storage needs for blob data. The Ethereum Foundation listed supported client versions, including Geth 1.16.7, Nethermind 1.35.3, Besu 25.11.0, Erigon 3.3.0 and Reth 1.9.3 on the execution side, and Lighthouse 8.0.1, Prysm 7.0.0, Teku 25.11.1, Nimbus 25.11.1, Lodestar 1.36.0 and Grandine 2.0.0 on the consensus side.

Developers

Developers needed to check whether any of their transactions could exceed the new per-transaction gas cap, and to account for the higher cost of MODEXP calls. Teams posting blob data needed to adapt to PeerDAS changes in how blob transactions are submitted. Wallet builders gained new options thanks to passkey support.

Did Fusaka Change the Price of ETH?

The upgrade was a technical change, not a monetary policy change. It did not alter ETH’s issuance rate or create new tokens. Any effect on price would come indirectly, through changes in network usage, fees and market sentiment, and those effects are very hard to separate from wider market moves.

Blob fees and ETH burning

Since 2021, part of every transaction fee on Ethereum has been burned, removing ETH from supply. The blob fee floor introduced by Fusaka means blob fees are less likely to fall to almost nothing, which can add to the amount of ETH burned when rollup activity is high. Whether this is significant depends on demand. Growthepie, an Ethereum data site, tracks Fusaka-related blob fees for anyone who wants live figures.

A note on predictions

Some articles before Fusaka predicted large price moves. Upgrades can influence sentiment, but they do not guarantee any price outcome. If you are considering ETH as an investment, focus on whether you understand the asset and can accept its volatility, not on short-term upgrade hype.

Why Rollups Matter for Ethereum’s Future

Fusaka makes most sense once you see why Ethereum leans so heavily on rollups. Ethereum’s main chain is designed to be secure and decentralized, which means thousands of independent computers must be able to keep up with it. That limits how many transactions it can process directly.

Rollups take most transactions off the main chain, process them quickly and cheaply, and then post compressed data and proofs back to Ethereum. Users get low fees, while still relying on Ethereum’s security to settle disputes and protect funds. The cheaper and more plentiful the space for that posted data, the cheaper rollups can be.

What this means in practice

For everyday users, the effect shows up as lower fees and faster confirmations when using apps on networks such as Arbitrum, Optimism, Base and others. For Ethereum as a whole, it means the network can serve far more people without asking every node operator to buy bigger hardware. Fusaka is one of the most important steps so far toward that goal, because it changed how data availability scales rather than simply raising a limit.

The trade-offs

Relying on rollups also brings challenges. Activity and fees are spread across many networks, moving funds between them can be confusing, and each rollup has its own security setup. Ethereum’s developers and rollup teams continue to work on making these networks easier to use together, but beginners should check which network an app runs on before sending funds.

How Fusaka Was Tested and Rolled Out

Ethereum upgrades go through a long testing process before reaching mainnet. Fusaka was first run on short-lived developer networks, then activated on Ethereum’s public testnets, Holesky, Sepolia and Hoodi, during October 2025. Client teams fixed issues found along the way before the Ethereum Foundation announced the mainnet date in November 2025.

Why upgrades take so long

Ethereum has several independent client teams writing software in different programming languages. Every client must implement every change in exactly the same way, or the network could split. Months of testing reduce that risk. The staged BPO approach after Fusaka continued this cautious philosophy, raising blob limits in steps rather than all at once.

What Comes After Fusaka

Fusaka was not the end of Ethereum’s scaling plans. The next major upgrade, Glamsterdam, combines the execution layer upgrade Amsterdam with the consensus layer upgrade Gloas.

Glamsterdam’s main goals

Glamsterdam’s planned changes include:

• Enshrined proposer-builder separation (ePBS, EIP-7732), moving block building arrangements into the protocol and reducing reliance on outside relays.
• Block-level access lists (EIP-7928), which tell nodes in advance which parts of the state each transaction touches, enabling parallel processing.
• State pricing changes, adjusting costs for creating and accessing data to manage how fast Ethereum’s state grows.
• Changes to basic transaction gas (EIP-2780), which could make simple ETH transfers cheaper.

Glamsterdam went live on the Sepolia testnet on October 6, 2026. Its next test, on the Hoodi testnet, was planned for late October but was postponed on October 22 so developers could study how one change, EIP-8037, interacts with existing smart contracts. As of early November 2026, no mainnet date has been set. Further BPO forks may also raise blob capacity again over time.

The longer road

Over the next few years, Ethereum’s developers aim to keep increasing both layer 1 throughput and blob capacity, make running a node lighter, and improve wallets. PeerDAS is an early version of data availability sampling, and later upgrades are expected to build on it to raise capacity much further.

Key Terms Explained

The upgrade introduced or relied on several technical terms. Here they are in plain language.

Term	Meaning
Blob	A temporary data packet that rollups use to post their transaction data to Ethereum cheaply
Data availability	Proof that data was actually published so anyone can check it
Data availability sampling	Checking data availability by downloading small random pieces instead of everything
Erasure coding	Adding redundant data so the full data can be rebuilt from part of it
Subnet	A group of nodes in Ethereum’s network that share a particular slice of data
Supernode	A node that stores all blob columns, typically run by large stakers
BPO fork	A small upgrade that changes only the blob target and maximum
Gas limit	The maximum amount of computation allowed in one block
Precompile	A built-in function the network provides at a fixed address, faster than normal smart contract code
Rollup	A layer 2 network that processes transactions off the main chain and posts data back to Ethereum

Common Misunderstandings About Fusaka

Big upgrades attract confusion and, unfortunately, scams. These are the misunderstandings that come up most often.

“I need to swap my ETH for a new token”

False. No ETH holder needed to swap, migrate or “upgrade” anything. Any message saying otherwise was a scam designed to steal funds or recovery phrases, the same playbook used by fake crypto recovery services.

“Fusaka made layer 1 gas fees much cheaper”

Not directly. Fusaka mainly expanded blob space for rollups. The higher gas limit added room on layer 1, but ethereum.org notes that no significant direct change in layer 1 fees was expected. Fees on Ethereum depend mostly on demand.

“PeerDAS reduces security”

PeerDAS was designed so the network as a whole still guarantees that all blob data is available, with a vanishingly small chance of error. Individual nodes simply do less of the work.

“Blobs are stored forever”

Blob data is kept only for a limited period, a few weeks, which is long enough for anyone to check rollup data. That has been true since blobs were introduced in 2024, and Fusaka did not change it.

Frequently Asked Questions

These are the questions people ask most often about the upgrade.

Final Thoughts

The Ethereum Fusaka upgrade was a quiet but important step for the network. By introducing PeerDAS and Blob Parameter Only forks, it let Ethereum expand the data space that layer 2 networks rely on without forcing every node to do more work, and it added smaller improvements such as passkey support, a higher gas limit and a more meaningful blob fee.

For everyday ETH holders, nothing changed in how you hold or send your coins, and no action was ever required. For rollup users, it helped keep transactions cheap. Keep an eye on Glamsterdam as the next milestone, and remember that any message urging you to “upgrade” your ETH is a scam.

Sources