Ethereum price forecast: Glamsterdam achieves 64 million Gas on Sepolia instead of 2 million, and wh

Ethereum's "Glamstam" upgrade has been launched on the Sepolia test network

Since Tuesday afternoon, Ethereum's "Glamstam" upgrade has been launched on the Sepolia test network. The fork started on time at 13:53:36 UTC, exactly the same time as the developer announced three weeks ago. This is a rare opportunity for anyone who owns Ethereum: the network, designed to host the main network in a few weeks or months, is now publicly running and being tested.

That's exactly what we did, and the results were different from what was said on the headlines. The core figure recently reported is a 200 million (200 million) Gas cap, more than triple the previous 60 million. But about an hour after the fork occurred on Tuesday afternoon, the Gas cap on Sepolia was 64,238,034. This is not contradictory, but it is not the same thing, and this difference explains why the main online launch date is still missing.

Glamstam on Sepolia: Forking Details

The Ethereum Foundation sets the fork at epoch 353,024 and slot 11,296,768, corresponding to UTC on October 6, 2026 at 13:53:36. This slot does become the first block under the new rule; at the execution level, it corresponds to a block number of 11,856,337.

You can identify this by two previously unseen fields in the block header: one is the `blockAccessList` Hash belonging to the "Block-Level Access Lists", and the other is the declared `slotNumber`. The first block with these fields happens to contain the slot number 11,296,768 in the announcement. Therefore, a fork is not only an assertion, but also a readable fact on the chain.

The block header is a block header record: a short list of key data that each node checks before accepting a block. When a new field is added, it means that all relevant programs have adopted the same new rule. This is exactly what the test network is for.

This analysis was performed by cryptoticker.io on October 6, 2026. It is based on the public block header data of the Sepolia test network; it counted 24 blocks between 13:29 and 14:49, as well as blocks before and after the fork.

Sepolia Gas Upper Limit: It was 60 million before the fork, and rose to 64.2 million an hour later.

Before the fork, the Gas Upper Limit on Sepolia remained at integer digits for a long time: 60,000,000, which was completely consistent between blocks. This is normal when the verifier operator sets the same value.

As the bifurcation begins, the values begin to change. Five minutes later, the Gas limit reached 60,647,496; forty minutes later, it was 62,568,310; and less than an hour later, it reached 64,238,034. Among the 218 blocks, approximately 4 million Gas was added, with an average of approximately 18,600 per block.

This process has two distinctive characteristics: First, the value does not increase uniformly, but in a step-like manner, occasionally even falling slightly. Secondly, multiple counted blocks are almost full: one uses 60,611,247/60,647,496 Gas, and another subsequent block uses 63,827,010/63,925,416 Gas. The filling rate exceeds 99%. On test.com, this was not accidental, but intentional. Anyone who wants to test a higher cap needs to fully load the network.

Increase by 1/1024 per block: How to move towards 200 million Gas Limit

The Gas upper limit is not a number set by a switch. Each proposed block can adjust the upper limit of the previous block by up to 1/1024, whether it is upward or downward. At a base of 64 million, this means about 62,500 Gas changes per block, with a theoretical upper limit of about 500,000 per minute based on a 12-second exit time.

The measured 18,600 per block is much lower than this because not every proposer has set a new target value. Extrapolated the 136 million Gas that was still missing from the 200 million target by Tuesday afternoon, which would take about a day at the current rate. As more verifiers switch, the speed can increase; otherwise, it slows down and may stall halfway.

Important background: The figure of 200 million appeared in the Ethereum Foundation announcement as a configuration recommendation for Prysm and Teku verifier programs, rather than a fixed value written into the upgrade specification. Therefore, the number describes what is to be tried on the test web, not what will be applied on the main web. For details, see the Ethereum Foundation's testnet announcement.

EIP-7928 and EIP-7732: block-level access lists and ePBS entry headers

Glamestead implemented two major changes. EIP-7928 introduces block-level access lists: a block declares in advance which accounts and storage slots it will touch. This allows nodes to perform transactions in parallel rather than strictly one after another. This is a prerequisite for significantly increasing the Gas upper limit, because without parallel processing, a block that triples in size will simply crush the machine.

EIP-7732称为“内嵌提案者-构建者分离”（Enshrined Proposer-Builder Separation，简称ePBS）。在此之前，区块构建和区块提议通过外部中继进行，即通过协议之外的中介。通过ePBS，这种分工进入了协议本身。对投资者而言，这主要是可靠性问题：网络越不依赖自愿的中介，中断导致区块生产放缓的风险就越低。

此外还有对Gas价格的调整，使网络存储访问变得更昂贵，而某些计算步骤更便宜。这种重新分配意味着更高的上限并不自动意味着转账便宜三倍。

Sepolia与主网对比：测试网分叉告诉以太坊持有者什么

Sepolia是一个测试网。那里的以太币没有价值，验证者由可管理的圈子运营，错误不会造成金钱损失。正因如此，周二的分叉不能作为主网已准备好的证据。

但它证明了另一件事，且非常有价值：各个开发团队的程序已就相同的新规则达成一致，并共同生成区块。当测试网分叉失败时，路线图通常会推迟数周。当它成功时，下一阶段就该到来。

下一个阶段是Hoodi测试网，之后才是主网。两者目前都没有日期。如果有人在最近几周读到升级将在10月到来，那是混淆了测试网激活与主网。

以太坊价格：决定10月的水平

On Tuesday afternoon, Ethereum was quoted at around $2,712 (about 2,410 euros), up 0.25% in 24 hours. It fell 0.75% in one week and rose 8.6% in one month. The market value is approximately US$331 billion.

The intraday volatility range is narrow: low of US$2,680.92, high of US$2,721.34. These two values are the most recent support and resistance levels. Beyond them is the entire area near $2,800, where ether was blocked several times in September. Judging from the historical high of US$4,946.05 on August 24, 2025, the price is still about 45% away.

Upgrade on the test website itself will not cause significant price fluctuations, and the price response on Tuesday will be correspondingly small. The technology roadmap will have a real impact on prices only after the main network date is determined. We described this situation in our September 30 assessment of Glamsterdam-main website dates, and the core content remains unchanged.

Fourth quarter mainnet: Roadmap without confirmed date

The official roadmap still states that Glamsterdam will be on the main net as the fourth quarter of 2026, and clearly states that the date has not yet been confirmed. For Hoodi and Mainnet, the announcement stated that once the development team makes a decision, the activation time will be announced.

This leads to a simple rule of thumb: In past Ethereum upgrades, forking from the test network to the main network was usually several weeks apart, because there was a second test network after Sepolia, followed by an observation period. October dates will be unusually rapid; November or December dates will fit the pattern. Therefore, anyone who links purchasing decisions to upgrades should count them on a weekly basis rather than a daily basis.

质押队列：146万ETH进入，等待25天

除了技术路线图外，还有一个比任何分叉都更具体地影响你规划的发展：质押队列两端都很长。入队队列中近期约有146万枚以太币，等待时间约25天；出队队列中有786,275枚以太币，等待时间约13天16小时。总共有约4370万枚以太币被质押，约占流通供应量的35.8%，预计年化收益率为2.63%。

Coindesk于10月5日根据ValidatorQueue数据汇编了这些数字。

实际的后果是一个日历上不显示的截止日期：任何人现在想要质押，只需承诺其以太币在约三周内激活；任何人想要退出，需等待约两周释放。这两者都与价格无关。如果在此期间市场下跌，你无法立即反应。

长的出队队列在此情况下并非自动卖出信号。部分释放的以太币直接重新投入质押，例如当提供商重建基础设施时。目前，该数字只显示了一件事：目前退出的路径有多长。

持有期、门槛、质押奖励：德国税务局的要求

For German investors, a second calculation is attached to the pledge. Gains from the sale of cryptocurrency assets are exempt from tax after one year of the holding period; this is stipulated in Section 23 of the Income Tax Act. If sold earlier, the annual threshold applicable to all private disposal transactions is € 1,000. The threshold means that all proceeds of € 999 are exempt from tax; when it reaches € 1,000, all proceeds become taxable, not just the part above the line.

Rewards from the pledge itself fall into different drawers. In the view of the German Federal Ministry of Finance, when privately owned, they are usually counted as service income under Article 22, paragraph 3, of the Income Tax Act, with their own annual threshold of € 256. They are valued at the time they are received, i.e., the exchange rate at which the award is credited.

The important point is where the most false claims are made: the pledge will not extend the one-year holding period of the coins used. Ether that has been held for one year before pledge can still be sold duty-free thereafter. However, any individual buying multiple batches at different prices does need to be clearly documented as proof of each inflow.

Gas costs, Layer 2 and hardware: What they mean for your tokens

A higher Gas cap initially only meant more computing work could be put into a block. As long as demand does not rise in the same proportion, this is good for costs: more space under the same demand will reduce base fees. This is most evident on Layer 2 networks, which bundle data on Ethereum and currently account for the largest share of space.

The synchronous reallocation of Gas prices works in the opposite direction. Network storage access becomes more expensive because permanently increased storage increases costs for each node operator. As a result, simple ether transfers do not change much; for complex contracts, calculations may swing around depending on their construction.

Anyone running their own nodes should pay attention to progress on Sepolia. A 200 million Gas block requires more memory, disk throughput, and bandwidth than a 60 million Gas block. This is the purpose of gradual increase: the only way to show at which value the first node will fall behind. So, the validator on Sepolia slowly raises the number instead of skipping directly to the target is not hesitation, but part of the program.

Glamstam in Sepolia: Summary points

Distinguish between test nets and mainnets. The fork on October 6 involved Sepolia. There are no dates for Hoodi and Mainnet; the roadmap only points to the fourth quarter of 2026. If you tie it to your purchase decision, count it weekly. Where you buy has a greater impact on returns than upgrades.

Consider waiting time before pledging. Currently, activation takes about 25 days, and release takes only about 14 days, which is normal. Anyone who wants to sell during this period cannot do so. Which providers offer what yields and exit routes can be found in the comparison.

Record each inflow. A one-year period and a threshold of € 1,000 applies to sales; a separate threshold of € 256 applies to pledge rewards, based on the rate on the day of receipt. Without complete records, neither can be proved in subsequent tax filings.

(As of October 6, 2026. This article does not constitute investment advice. Price and fee structures are subject to change; check terms with your provider before purchasing.)