序言：一直以来一神教布道者都有一种原教旨的狂热，凡是不理解他们观念的，被认为是荒谬的，愚蠢的，野蛮的，落后的。而事物的嬗变总是有过程有条件有节奏的，带着造物主的气定神闲。我们相信未来是属于BTC与黄金互锚的时代

TLDR

1. 痛点：婴儿潮一代理解BTC有困难，需要拥有链上支付网络的黄金凭证来对冲法币风险。

2. 方案：建立“去中心化的布雷顿森林”——用链上黄金（实体锚）+美元稳定币（流动性）+BTC（心理锚）三元共存。

3. 黄金和BTC双飞论：发黄金Token，买黄金，提高BTC相对比价，拉BTC，扩大Crypto支付受众网络；受众持U，USD买黄金，提高BTC相对比价，拉BTC，扩大Crypto支付受众网络。支持黄金就是支持BTC。

4. 如果tether持续买黄金，是不是有一天usdt的backer比usd的backer还要硬气？2.稳定币有一天可不可以与美元没关系？3. 黄金越高，BTC比价越贵，U的势力范围越广。奥尼尔越具有统治力，姚明就相比历史定位更高，tether买BTC和姚明捧奥尼尔很类似。

在合规与去中心化张力下，构建一种“链上黄金”为锚的多极共存金融秩序，作为对传统布雷顿—后布雷顿体系的去中心化再想象。不是回归“国家间金本位”，也非“纯加密本位”，而是把“黄金的现实信用”与“链上的程序化结算与可组合性”绑定，形成一种跨域可清算、部分去信任的锚定机制。Trustless System和Trust system 的竞合是当前的一个历史时期，当然我们今天不谈创新，我们谈的是历史时期中的生态位。

我们所看到的趋势

1. 观点：比特币与黄金的相关性正在趋缓，相关性越来越高。BTC的价格轨迹越来越和黄金驱动。对于黄金的投资者，如何帮助他们得到像BTC一样的支付结算网络。

2. 观点：U是一个历史过度时期，如果Tether库存大量黄金，他是不是长得最像二战时期美联储的样子？ ![[Pasted image 20251014170017.png]]

1. 观点：除非美元循环结算质押清算完蛋，否则美元不会比别的稳定币先完蛋，美元绝不会是第一个完蛋的。2，使用西方相同的经济制度的货币会铁索连环，如果债务持续高增稳定币最终会不稳定，如果要直面现状解决问题，又难免会起地缘冲突。最终的情形是，所有铆钉法币的稳定币会变得不稳定，目前一部分以U代替法币使用的持有者也会发现，由于法币本身不稳定，U也变得不再稳定了。（图：2023年以来西方世界对于美债的增持，以英国最为典型。）

   

   1. 黄金与美元齐张齐跌。传统上金价与美元指数负相关，但在2023–2025年，多次出现“同涨”。这是驱动因子从“单一的美国实际利率”转向“全球安全与流动性冲击”的信号。CME 也专门总结了这种非常态阶段：地缘紧张+央行购金高位，让金价在美元坚挺时仍走强。CME Group

   2. 为何美元先被抢？ 因为美元是“结算语言”与“负债计量尺”：出口开票、跨境债券/贷款、银行资产负债表的主要负债，大头仍是美元（发票、储备、外币债发行的占比数据都显示美元居首）。这使得赚不到美元（经济差）或冲击时对美元的刚性需求需要放首位。 ![[Pasted image 20251014170323.png]]

      

2. Tether的预判与对策，广义黄金的内部循环。 如果tether持续买黄金，是不是有一天usdt的backer比usd的backer还要硬气？BTC是用U计价的，U的发行商正在大量购买黄金，这意味着编辑上产生的U都会有一部分储备去配置黄金，如果usdt的网络持续扩张，黄金储备量继续抬升，由于USDT网络的增长意味着链上多一份U，链下多一份USD，链上的U对应Tether的负债，而链下的USD对应Tether的资产。当tether拿一部分资产去购买黄金的时候，同时也就带动了BTC的估值定价，如果我们以相对粗糙的方式认为BTC只有20%黄金，我们就可以看到买BTC和黄金的增速螺旋。那么如果我们将区块链是为美国发展的边际中心，其网络的扩张将不断捕获BTC，那么U相对USD就会有更强的Backer，即使其本身声称其等于1usd。 ![[Pasted image 20251014170206.png]]

   

3. 那么随着时局的发展，由于U的相对储备会比USD更有价值，我们会发现U的网络会比USD更有价值，起码U的负债是可控的。那么我们是否有理由相信tether发行的xAUG也能成为一个option呢？我不能说服bitcoin-max，但是我们起码可以把它当作一个option。当然你也可以选择，中国政府发布的纸黄金，这同样也是一种option。取决于你的文化地域身份。 ![[Pasted image 20251014170650.png]]

   

广义黄金的配置光谱

目的展示：不同投资者对光谱黄金的兴趣。

1. 所以当我们回答未来秩序的时候，有几点可以肯定：

   1. 当我们谁都不信的时候，我们想要买黄金。

   2. 当然动荡之时配置黄金，是不是遭遇劫匪？

   3. 如果你所在的区域政治局势稳定，街头安全，你大可以买纸黄金。

   4. 如果你所在的区域中上层局势不稳，但是街头安全，你可以买真黄金。

   5. 如果你所在的区域局势不稳，且街头也不安全，你可以买BTC。

   6. 当然你也可以说，你相信BTC未来可以超过黄金，你是一个BTC-MAX，那么你不需要思考，BTC就是你的唯一选择。

   7. 当然你也可以，担忧你觉得BTC在量子威胁下，认为黄金经历了千年的考验，你仍然相信黄金。那么在这种组合下，你真实想要的其实就是历史上最辉煌时期的美金。也就是黄金Token。

2. 你所在的区域动荡，你希望找一个更加有实力的经济体去托管和赎回你的黄金，甚至其发行的黄金Token是可以用于支付和日常使用的。

链上黄金作为BTC的【补充】

我们所面对的历史周期是一个必须打碎且必须重建的过程，是一个Trustless和Trust共存的特殊时期，我们需要满足婴儿潮一代的新时期财富管理需求，这一代人需要重新寻找一个能铆钉黄金，被大家接受的，具有网络效应和信誉的黄金支付凭证。在极端情形行驶支付职能。

而黄金凭证(Token)具备矛盾的诡异特性：

1. 具有统治力的黄金Token不可信。

2. 可信的黄金Token不具备统治力。

如果今天xAUG占据了如同美联储一样的统治地位，那么xAUG就可能也幻化成牙买加协议一般的信用凭证；正是因为其不拥有至高无上的权力，所以不兑便不足以获得信任和扩张。权力拥有点石成金之能，而追求权力必须付出真金白银。也因此，没有市场统治力，却有市场竞争力的黄金Token是值得信任的。

历史专栏：戴高乐的航母

二战后，布雷顿森林体系以“35美元=1盎司黄金”为锚，美元处于体系核心，初期因美国掌握约75%官方黄金储备而运转顺畅。进入60年代，越战与“大社会”开支推高财政与经常项目赤字，美元超发，“特里芬难题”显现；各国央行集中要求以美元兑金，体系承压。尽管通过互换额度与“伦敦黄金池”等权宜之计维稳，黄金池在1968年崩溃，官方与自由市场形成双轨金价，信心持续下滑。到1971年，外国持有的美元债权远超美国黄金储备，美元金本位难以为继。法国最早发难。戴高乐公开抨击美元的“不合理特权”，主张回归黄金结算，并按官价持续将海外黄金赎回、开航母运回本土，形成示范效应，带动欧洲多国跟进，显著加剧美元兑金压力。这一连串冲击最终促成尼克松于1971年8月关闭黄金窗口，布雷顿森林体系宣告终结。

正是由于铆钉黄金的货币无法适应于现代经济的增长，而现代经济增长很大的比例来自国家间的信用，债务和杠杆。信用是虚的，但是派生的货币和经济和增长是真实的，因此反过来要求货币必须虚无才能应对真实经济的扩张，牙买加体系是在这种氛围内诞生的；回头观察我们当今的局势，债务是真实的，我们寄希望于进一步印钞宽松以虚无解决问题，但债务毕竟是真的，也因此全球化对应信用的极致也终于往回走，对应也要求货币有更加真实的支撑。

![[Pasted image 20251014170337.png]]

黄金Token的存在：一来在链上提供一个高信用、低波动、跨链可组合的价值锚，改善抵押与清算质量；二来缓解加密体系对美元/美债的单点依赖，提升多极储备与结算机制的韧性，拓展稳健收益与风险对冲工具。3）终极贡献？逐步形成一个“去信任度更高、抗制裁、更透明”的国际化清算与储备网络，使货币秩序从单中心走向多中心的“共存秩序”。

水贝盘口爆仓与早年cex跑路何其相似？

主流项目概览

Tether Gold（XAUT）总量已到≈15亿美元量级。以金价 2,300–2,600 美元/盎司估算，这大致对应 18–23 吨实物金（1 吨=32,150.7466 盎司）。全球官方黄金储备总量约 3.4–3.6 万吨，美国一国就 8,133 吨。所以——XAUT 的“买金体量”≈全球官方储备的 0.05–0.07%，相当于一个小型央行的库存，远小于主要经济体，但已可与一些小国的官方储备量级相提并论。一点对照（吨）- XAUT：≈18–23 吨

• 美国：≈8,133 吨

• 德国：≈3,352 吨

• 意/法：各≈2,4xx 吨

• 俄/中：各≈2,3xx 吨（申报口径）

• 印度/土耳其：≈800 吨量级

• 新坡：≈200+ 吨（近年增持）

• 若干国（如塞尔维亚、柬埔寨等）：10–60 吨区间——XAUT大致落在这条线上

1. 量级：XAUT 目前只是“一个小央行”的体量；对全球官方储备结构几乎不构成数量级影响

2. 性质：官方储备是主权资产，受条约与央行会计框架约束；XAUT是商业代币+托管实金的组合，法律地位、可执行性、审计透明度与主权储备不可同日而语。

3. 功能外溢：若体量继续上升，XAUT 可能在场外结算/抵押、加密市场保证金、跨境财富配置中充当“影子储备”角色，但这更接近私人保管与票据化仓单，不是央行储备。

结论

1. 黄金网络的派生是可以期待的，会成为USDT网络的一个有效补充。

2. Crypto与黄金网络的再组合。

参考

1. 《独立审计师报告：Tether International财务数据和储备》（2025年7月31日发布） assets.ctfassets.net

交易所的业务来源

1. 现货。

2. 期货。

3. 质押。

以coinbase的收入来源来看，50%的零售用户收入，30%的质押利息收入，机构业务4%，10%的链上。从交易所过往的经历来看：期货最容易起量，核心的指标是MM深度；现货最难增长，但底盘相对稳固，业务有很强韧性，核心指标是区域市场的占有率。而质押市场相对于是连带的理财业务，其增长主要伴随的是链上资产的活跃。

1. 存量项目生命周期短，21年所谓龙头的Defi项目，目前符合“资产逻辑”可以存续上线的比例不足15%。

2. 增量项目逻辑弱，无论是长尾博弈市场诸如 pump.fun 和 Polymarket 均。现金流项目可持续性存疑。

3. 逻辑顺项目GTM较难，尤其以Depin类项目为主。

讨论一: 上币类型：证监会式上币，用户获取式上币，投资型上币的三种不同类型？ 哪一种更合适？

证监会式上币是自上而下分析各类资产，按照存量实时的最新数据来进行判断和筛选上币。其优点在于数据清晰可查，验证过程可复现，审核结果可靠；其缺点在于P2E大量的资金驱动型数据，形式合规，实质不合规。**用户获取式上币**，典型的考核方式是binance，其实际的算账逻辑是用存量用户换增量用户，保证其市场占有率始终在优势地位，是一种保守型打法。优点在于查缺补漏，防守特征明显。

投资型上币，典型的诸如meme token or meme期货 优先上的战略。敞口可控的情况下，提前捕获交易量。用代币敞口换交易量。整个投资行为不需要项目方做市，自己主动来做。风险高，收益大，快速决策快速捕捉。

目前市场环境，已经不是早年万类霜天竞自由的市场，因此19-21年的经验已经不适用了。当时的环境是，任何概念都能获得散户的关注，散户是概念敏感的，今年以来，主流币的市场占有率到了历史极值，BTC+ETH占比80%，是一个概念不敏感的市场，且新概念冒头慢，半衰期短。我们不讨论叙事经济学是不是失效了，我们认为到了一个回归本源的一个状态，叙事是需要可行性检验的。这个状态下，上币和投资之间风险溢价是大大降低了，有数据和没数据的风险是一样的。（他没运营的时候，看不出啥；他运营的时候，就是一波营销，也看不出啥。）

讨论二：如何利用合规的市场影响力尽可能打开局面？

从质押这个点入手是比较容易打开局面的，原因有三：其一，质押的动作相对面临的合规压力较小，没有纷繁复杂的资产认可度文图；其二，质押动作可以放大合规的品牌作用，尽可能撬动外围金融机构的分销力量；其三，通过质押业务撬动外围分销力量，再撬动BTC相关生态，并且从一部分合作深度强，业务数据可验证的BTC机构这里寻找待上币资产，合作形式以联合上币为主，捕获头部交易所用户。

为什么现货不宜成为核心点？现货面临合规压力大，沟通成本高，决策窗口期短。部分机构面临是**获客压力**和**合规压力**的tradeoff，在目前新资产热度期短的背景下，要快速上币非常难，很难从这头获取竞争优势。但反过来说，现货梳理肯定是有意义和价值的，但短期作用不明显，起效链条太长。

CEX看成“交易所”会面临极大的挑战，不如暂时视为CeDeFi，类似于一个为BTC服务的实体营业部。从确定性角度讲，虽然我们不知道BTC的占有率会不会持续提高，但是BTC相关的市值和交易量应该是稳步上升的。围绕着这个热点的用户，交易量，资产托管量，都会稳步提升。至于其他资产，目前生命周期，存在逻辑，用户和资产的捕获方式都面临很大挑战，作为一些零散的方向，择机再进入。小交易所的重心应该是BTC+MEME的杠铃策略。

围绕端侧上行流量痛点，构建可信数字底座模组生态

1. The Coming Future We Face

   问题一：端侧上行流量的成本应该有谁承担？如何降低用户门槛

   问题二：如何保证端侧上传数据的真实性，避免模型收到垃圾数据污染？

2. How to solve the question？

细数AI为物联网带来的变化，围绕蜂窝流量，我们相信以下几个变化:

1. 本地推理成为PlanB。端侧算力增强，算法优化，隐私顾虑都可能使得本地推理成为一种选择。想象一下，一台家用主机部署了大模型，你在办公室通过手机向家用主机发起请求，进行交互，得到了AI的支援。在这种情形下，家用算力需要有上行算力。

2. 流量结构更加多元，未必以手机为中心。目前的设备端，均以手机为核心的辐辏结构，我们的耳机音响通过蓝牙与手机连接，得到手机的信息。随着端侧AI的兴起，产品或独立与云端连接，或与家用主机连接。端侧上行流量增大，不同于传统手机流量。

3. 端侧传感器矿机。目前的AI大模型停留在概念和文字等抽象信息，未来会更多往视频，声音，去理解实际生活中的意义。边缘传感器矿机是否会从手机中拆分出一类，或未可知。端侧传感器的上行流量，是挖矿商业模式当中成本之一，拥有商业模式优化的空间。

4. 矿机所消费的数据必须以真实为前提，边缘端侧上行数据必须在安全可信的环境当中。于C端是安全不泄露，于B端是可信可炼丹。边缘侧的安全可信底座是整个Depin世界的基石。

5. AI硬件投资热将逼迫项目方采购上行流量，降低用户参与成本。

在我们考察整个业态的过程中，我们认为边缘AI+miner的商业模式中有两大问题非常关键。

未来各个AI模型将会向用户采购数据，那么谁来承担数据上传的成本呢？是APP还是用户呢？

AI的服务是典型上传和下载的交互，和传统的内容不同。传统内容偏向下载，而AI动作上传和下载都同样重要。那么边缘侧的上传流量谁来买单呢？如果说过去的流量结构是用户购买内容，是多对一下载的结构；那么未来的流量结构，会呈现多对一的上传结构。如果用户每一个设备都有大量内容需要上传，是不是意味着终端用户要承担天量的成本压力？

从收益结构来看:

• 用户侧出发：不上传就没服务，所以用户要负责上传。

• 应用侧出发：不获取数据就没法迭代，所以你要支持用户上传的成本。

• 竞争力越强的AI服务，用户要承担更多成本；竞争力稍弱但迭代意愿很强的AI服务商，应用侧将会承担更多成本。

目前市场都认为AI变现很困难，场景选择非常困难。从项目闭环角度，我们可以认为PMF（场景）> 数据收集> 模型进化。如果赛道上项目摸到了PMF，那么很快后续的竞争就会激烈的展开。在后续竞争中我们认为两个关键问题：

1. 如何降低数据采集成本？

2. 如果降低现金流支出？

随着竞争白热化，流量很大概率会由APP/应用端/网络侧付费，那么这时候你准备好了嘛？

以目前端侧数据上传的可信性是非常存疑的。不解决数据可信性的问题，所有的数据贡献挖矿都是自娱自乐，其训练得到的模型只能是Garbage in Garbage out. 

从终局来看，所有端侧设备都必须在端侧配一个可信模组，保证其记录和上传数据的真实性。问题是如果说服具体端侧项目方去配合改造？

第一个问题：端侧上行流量的成本应该有谁承担？如何降低用户门槛？ 数字底座出一部分，将硬件设备的推广成本转换成流量池，去赠送和补贴AI项目。
第二个问题：如何保证端侧上传数据的真实性，避免模型收到垃圾数据污染？ 随着流量卡的分发和设备端可信节点网络的设立，我们可以利用运营商基站提供的LBS数据，以及可信节点数据去服务更多的硬件设备，做更多验证和治理的工作，诸如：

1. 节点的行为监管。节点如果行为不端，可以暂时让节点脱网，Slash其贡献。相比传统质押Slash的治理模式，这种治理模式是事中监管，且资金沉淀少，反应速度快。

2. C端小米之家的管理平台。多节点端侧矿机的C端管理，一键治理多个端侧矿机。

Author：Yolo Shen@Foresight Ventures

Traditional computers consist of five parts: Computer, Memory, Controller, Bus, I/O. Looking at the development of the blockchain, the calculator and memory is relatively complete. If the entire distributed system is compared to a human being, then the human brain and memory system are already complete, but the perception and sensory system are still in a very primitive state. At this stage, Depin is undoubtedly the hottest buzzword, how to achieve it? The entire system needs to evolve a “trustworthy touch”, and we all know feelings and reflexes rely heavily on the spine and nervous system.

If the blockchain system is the consciousness built on the iceberg, then the Infra network of depin is the subconsciousness under the iceberg. Now the challenge is here, who is the spine and nerve of the distributed system? How do we build a spine and nerve? In this article, we will start with the small lessons from the history of IoT, to help buidler make depin world quicker.

In this article, we will start with the small lessons from the development of IoT，to help buidler clear mind and make something real happen.

1. Depin should not be unitized by devices, as devices do not have the capability to scale horizontally; the core of Depin is Pin, and the core of pin is the authorization code. We regard the device as a collection of sensor modules, and the pin code of each sensor module is the permit for data to go online and the authentication permit for PoPW. The device with the permission to go online, contributes the recognized device, can be called a mining machine. Therefore, the core of the entire Depin track is how to make the contribution of edge devices measurable, how to make the contribution of different devices and the same sensors have a consistent measure.

2. According to the different information transfer of traditional computers, it can be divided into three categories: Data Bus that transfers various data information; Address Bus that transfers various address information; Control Bus that transfers various control signals. The DePin bus also exists similarly: as an identity certificate for device access. (Address Bus) as a PoPW certificate for data verification. (Data Bus) as a means of device management. (Control Bus)

a. Address BUS: Dephy

b. Data BUS: Virtual Communication Layer + Sensor Network

c. Control BUS: Cellular Management Module

3. The Depin project has some RW attributes, exists in the physical world, and is related to real economic life. Therefore, more real-time management methods are needed to achieve autonomous risk control. The main implementation channels are two: one is to govern the traffic of cellular operators, once the device violates the rules, it can lose the PoPW mining right from the traffic end, which is a more Real Time management method than PoS like slash. The second is to buy out upstream resources through the miners + resource pool method. For example, a dealer has 100 units resources, when 30 are at risk, there may be a warning of revoking the license. now we can mix the 30 resources with the resources of other dealers, buy out the real resources (RWR) through the miners, and control the resources through the mixed segment method, to obtain as many resources as possible under the premise of guaranteeing the risk of upstream dealers. Liquity model can be copied to all kinds of RW resources.

Looking back at the development of the Internet of Things(IoT) since 2015, there were two main challenges at the time: one was the single input and output of hardware devices. The second is that the product characteristics have not been enhanced after the device is connected to the network, and the device does not have Scale characteristics.

During this period, the core question was: what kind of changes will occur after the hardware device is connected to the network? the plain answer is，it make possible for hardware devices to upload and download. The next question is, why does the hardware need to upload and download? Can uploading and downloading increase product competitiveness? At that time, we saw a batch of products such as smart curtains, smart air conditioners, etc. Because the hardware has a relatively determined I/O dynamic line at the beginning of the design, the space for software development is relatively limited. Therefore, after being connected to the network, the product features only increased the feature of mobile terminal control, similar to “remote air conditioning, remote curtain pulling”, and most of the functions are remote + traditional controllers. For C-end users, this design is useless or even strange. Another core question is, does IoT devices have the ability to Scale after they are connected to the network? We mentioned earlier that hardware networking has more uploads and downloads. If the download is equivalent to the upgrade and expansion of the function, then the upload is the aggregation and integration of data. The value of the latter data lake in the early IoT era was very heavy, and there was a huge contradiction between the exponentially rising storage costs and the hard-to-explore data sales. In summary, IoT devices cannot improve product power and service dimensions in both download and upload modes.

Time always changes，Ai bring the new feature and possibilities，3 things happen：

1. Anthropomorphism of all things, independent upload and download demands. If the edge side cannot infer large models, then the terminal side needs to be independently networked to the cloud. This will transform the past radial structure with the mobile and the device into a structure with independent networking of devices.

2. Device sovereignty. From mere product sales to a dual-drive of user purchase + data sales. The device is responsible to the user not only as a product, but also as a sensor collection, responsible to the data merchant.

3. “Data is trustworthy, privacy is reliable” are the prerequisite conditions for ordinary devices to transform into mining machines. If the data is not trustworthy, then logically opening multiple virtual machines can hack the entire incentive system; if privacy is not reliable, then in the long run, the user’s interaction willingness will be suppressed.

With the development of AI, we see that Depin may have some different possibilities:

1. The emergence of AI has increased the necessity for AI hardware to connect to the internet autonomously, and the cost of device networking is likely to decrease rapidly in the next 3 years. Combined with the decline in storage and computing costs, the cost of deploying edge computing/sensors is also likely to decrease. Once many devices have been deployed, transforming them into mining machines to collect sensor data will reach a tipping point.

2. After solving the problem of devices independently connecting to the cloud, there will be more scenarios for interconnection between devices. How to use NFC and other low-cost hardware for interactive gameplay will also become a potential innovation point.

3. Making the collected perception data a bulk commodity is the core bottleneck of device mining. Setting the standard for abstract information products is a major challenge.

Based on the past 5 years of IoT development experience and changes in AI new features, we believe there are three major investment themes:

• Hardware infrastructure centered on cellular modules.

• Abstract communication layer services with communication information commodities.

• General miners as a type of dealer service.

A module integrates baseband chips, memory, power amplifier devices, etc., onto a circuit board and provides a standard interface functional module. Various terminals can realize communication functions with the help of wireless modules. With the development of the entire computing network, the definition of modules is continually enriched, forming a cellular networking + computing power + end-side application ecosystem:

• Traditional Cellular IoT Modules: Basic connection modules, the main function is to implement cellular communication. These modules only contain chipsets that support this connection, without the need for additional functions.

• Intelligent Cellular IoT Modules: In addition to providing connection functionality like traditional modules, they also integrate additional computing hardware in the form of central processing units and graphics processing units (CPUs and GPUs).

• Artificial Intelligence Cellular IoT Modules: In addition to providing functions similar to intelligent cellular IoT modules, they also include dedicated chipsets for artificial intelligence acceleration, such as connection modules for neural, tensor, or parallel processing units (NPUs, TPUs, or PPUs).

Looking at the entire industry chain, upstream chip and downstream device manufacturers occupy most of the value chain. The module layer in the middle is characterized by high market concentration and low gross profit. Traditional service devices mainly include: PCs, mobile phones, POS machines. Due to its huge concentration, once a module middle layer with wide consensus is deployed, it naturally migrates various existing devices into mining machines. If traditional Web3 users are on a per-person basis, then the middle layer represented by modules will allow a large number of smart devices to enter Web3, and the Tx between these devices will generate a large amount of demand on the chain.

Looking back at the early competition between Nvidia and Intel, we gained a lot of historical experience: The early computer chip market was dominated by the Intel CPU X86 system. In some marginal markets such as graphics acceleration, there was competition between the accelerator card ecosystem led by Intel and Nvidia’s GPU; and in a broader market (areas with uncertain demand), Intel’s CPU and Nvidia’s GPU had cooperation. The two companies coexisted and prospered for a while. The turning point appeared in Crypto and AI, where a large number of computing tasks are characterized by small tasks being parallelized on a large scale, which is compatible with the computing characteristics of the GPU. When the wave came, Nvidia made several dimensions of preparation:

1. Cuda’s parallel computing instruction set. It helps developers better utilize GPU hardware.

2. Rapid iteration capability. The iterative speed beyond Moore’s Law has won it survival space.

3. The pattern of competition and cooperation with the CPU. It effectively leveraged and utilized Intel’s existing resources, and quickly seized market opportunities in some decision-sensitive areas.

Returning to the module market, there are several similarities with the competition pattern of GPUs and CPUs back then:

1. High industry concentration, and the leading group has quite strong pricing power for the entire industry.

2. Development depends on new scenarios. Communication modules + intelligent chips + standard protocols are very likely to lay a moat at the device end.

3. Rapid iteration has the opportunity to grasp new opportunities, traditional players have long decision cycles, emerging scenarios change quickly and have risks, the environment is suitable for the birth of new species.

In this competition, the Crypto Stack is undoubtedly the best technology stack for building protocols and ecosystems, and the migration of existing devices to cash flow mining machines will generate β-level opportunities. Among them, DePhy is an important player, by building an integrated module + Ledger + identity layer, to realize the distribution and management responsibilities for the entire Depin network.

What exactly is a mining machine? We believe any hardware/software that can generate specific information resources and intends to acquire token resources can be called a mining machine. In this understanding, there are several standards for mining machines:

1. Does it generate specific information resources?

2. Can it receive tokens? is that a real need？

Therefore, in the whole process, whether the device generates trustworthy specific information resources PoPW (proof of physical work) becomes very important. Therefore, we believe that every sensor capable of producing PoPW needs a trustworthy (TEE/SE) to ensure the credibility of edge data collection. In the field of sensors, many horizontally scalable networks can be generated. For example, the video resources collected by cameras of different devices will be measured uniformly in a network. Compared with independent collection by different devices, horizontally extended sensors + trustworthy modules can construct a larger PoPW resource market. The video materials collected can be better priced according to unified standards, which is more conducive to the formation of a large-scale market for information resources. This is not available in Device-Focus.

Since some Depin devices exist in the physical world and are related to traditional commercial society, and the Crypto world is characterized by PermissionLess, how to manage all kinds of participants in real-time without KYC becomes particularly important. We believe that the whole web3 world needs a communication abstraction layer that integrates cellular networks and public IP networks. Users/devices only need to pay Crypto currency to get the corresponding network services. Specific direction tracks include:

1. Integration of traffic. Connect the traffic resources of global operators, turn traffic into a kind of information bulk commodity, and price it with tokens.

2. Integration of number segments. Connect the number segment resources of the world, make numbers a kind of identity layer, trade and price with tokens, and use Blockchain as a governance system.

3. Integration of IP resources. Connect public IP resources, integrate public IP pools into a resource, make public IP a resource pool, support arbitrary access and jump, trade and price with tokens, and use Blockchain as a governance system.

1. Depin should not take devices as units, devices do not have horizontal scale capability; the core of Depin is Pin, and the core of Pin is the authorization code. We regard the device as a collection of sensor modules, and the pin code of each sensor module is the permission for data to enter the network and the authentication permission for PoPW. Devices with network access rights and contributions that are recognized can be called mining machines. Therefore, the core of the whole Depin track lies in how to make the contribution of edge devices quantifiable, and how to make the contributions of different devices with the same sensors have a consistent measurement standard.

2. According to the different ways of transmitting information in traditional computers, there are three categories: Data Bus for transmitting various data information; Address Bus for transmitting various address information; Control Bus for transmitting various control signals. Similarly, DePin bus also exists: As an identity certificate for device access (Address Bus), as a PoPW certificate for data verification (Data Bus), as a means of device management (Control Bus).

3. Depin project, due to its partial RWA properties, exists in the physical world and is related to real economic life. Therefore, more real-time management methods are needed to realize autonomous risk control. There are mainly two implementation channels: one is through the governance of cellular operator traffic. Once a device violation occurs, the device can lose the right to PoPW mining from the traffic end, which is a more real-time management method compared to Slash. The other is to buy up upstream resources through miners + resource pools. For example, a dealer has 100 number segment resources. When 30 are at risk, there may be warnings of penalties and license suspensions. Today, we mix the 30 resources with the resources of other dealers. Through miners buying up real resources (RWR) and mixing number segments to control the risk of resources, we can obtain as many resources as possible while ensuring the risk of upstream dealers. Replicate the Liquity model to all kinds of RW resources.

1. 《The rise of smart and AI-capable cellular IoT modules: Evolution and market outlook》，iot-analytics https://iot-analytics.com/rise-of-smart-ai-capable-cellular-iot-modules/

2. 《Depin Investment MEME 202505 Yolo》https://docs.google.com/presentation/d/1Zf-QbhuH_gYYc8adQ7ZQ8yGHWC62uDTvHVZ2GVLF8cQ/edit#slide=id.g2e09513e47b_0_137