014产品思考

为什么有人愿意买一台可以自己修的电脑和手机？

一次偶然的交流，让我开始研究 Framework 和 Fairphone：它们怎么设计、怎样赚钱、是否真的环保，以及中国为什么少见这样的产品。

2026/11/4约 14 分钟

可维修设计
模块化
可持续
产品观察
消费电子

从一位德国朋友的 Framework 和 Fairphone，想到模块化、可持续与产品设计。

01｜从两件不太熟悉的设备开始

前段时间认识了一位德国男生。交流中，我发现他使用的电脑是 Framework，手机是 Fairphone。

这两个品牌我以前都不太熟悉。和我更常见的苹果、华为、小米、联想等品牌相比，它们最先吸引我的并不是处理器跑分或摄像头参数，而是一个听起来有点反常识的特点：用户可以自己拆开、维修，并更换一部分零件。

我开始找美国可以自行组装电脑的品牌，接着又查模块化手机，后来甚至查到了 Linux 手机。问题也慢慢从“这是什么牌子”，变成了“为什么有人会选择这样的设备”。

我还不知道他选择这两个品牌的具体理由，也不想替他回答。也许是喜欢折腾硬件，也许在意长期使用或隐私，也可能只是觉得产品好用。但他的选择让我发现：我们习以为常的消费电子设计，并不是唯一的选择。

02｜Framework 和 Fairphone，到底在解决什么问题？

Framework 是一家美国电脑公司。它的笔记本把电池、存储、内存、接口等部分做得更容易拆卸和替换。部分型号可以通过更换整块主板升级处理器平台；Framework Laptop 16 还设计了可更换的独立显卡模块。2026 年的 Laptop 13 Pro 使用 LPCAMM2 内存，在低功耗与可更换之间寻找平衡。

但它也不是“笔记本乐高”。CPU 仍然集成在主板上，显卡模块也受扩展舱规格和机型兼容性限制。Framework 已公开扩展舱的机械与电气规格，但模块化并不意味着所有显卡都能任意拼接。

Fairphone 是一家荷兰手机公司，关注点略有不同。Fairphone Gen 6+ 提供 12 类可自行更换的零件，包括电池、屏幕、摄像头和 USB-C 接口，并承诺持续提供软件支持至 2033 年。但你不能像给电脑加内存那样，给它任意升级 RAM 或处理器。

所以，我更愿意这样理解它们：Framework 把“以后还能升级什么”当成产品问题；Fairphone 把“坏了以后怎样继续使用”当成产品问题。 两者都关注维修，但并不是同一种模块化。

还有一个让我重新理解的概念是“开放”。可以更换电池，不等于可以自由安装系统；支持 Linux，也不等于公开所有芯片设计。可维修、可升级、软件开放、硬件开放，实际上是四件需要分别判断的事。

03｜模块化真的更环保吗？

这是我最想查证、也最不愿轻易下结论的一部分。

Framework 的思路是延长设备使用寿命：电池老了换电池，性能不够时视型号升级主板。更有意思的是，旧主板还能装进独立机箱，继续充当一台小电脑。它也委托德国 Fraunhofer IZM 做过笔记本生命周期评估。

Fairphone 的目标更广。除了维修，它还关注原材料来源、再生材料、工厂劳动条件和电子垃圾。根据企业发布的 2025 年影响力报告，Fairphone Gen 6 的产品碳足迹约为 29 kg CO₂e；公司估算，延长设备使用寿命在 2025 年帮助避免了约 2,083 吨 CO₂e 排放。

这些数字很值得看，但也需要认真看计算方法。例如，避免排放量是基于使用寿命与对照情景估算出来的，不是给每一台旧手机直接装上仪器测得的结果；不同品牌、不同型号的碳足迹也不能简单横向比大小。

对我来说，真正重要的问题不是“它有没有模块”，而是：用户会不会修？零件能买到多少年？维修比换新便宜多少？设备最终多用了几年？

可维修是一种可能性，真正的可持续还需要被使用行为和数据证明。

04｜这样做生意，能赚到钱吗？

我觉得这比“环保手机”四个字更值得研究。

Framework 创始人在公开的公司战略文章里，把长期方向分成三部分：产品、Marketplace，以及维修服务网络。先卖能够维修、升级的设备，再通过主板、接口、配件、第三方模块、翻新或二手零件，形成持续的交易与服务机会。

这里有一个很有意思的商业转变：用户不是买完电脑就结束了与品牌的关系。两年后换电池、四年后升级主板，甚至把换下来的主板再利用，也可能成为整个生态的一部分。不过，战略规划并不代表每一项维修服务都已经在所有地区成熟落地。

Fairphone 则尝试把耐用性、维修便利、长期软件支持和更负责任的供应链变成用户愿意购买的价值。它在 2025 年报告及官方摘要中披露，设备销量同比增长超过 40%，但销量增长也不等于已经证明长期盈利能力。

两家公司都要面对一个不太浪漫的现实：长期供应零件需要库存，跨代兼容需要研发，规模小还可能带来更高的单位成本。好的理念如果不能覆盖这些成本，最后只会变成一个精致的小众爱好。

所以，我觉得它们不是在证明“环保一定能赚钱”，而是在尝试回答：能不能让用户长期使用产品这件事，本身也成为一门可以持续经营的生意？

05｜中国没有类似的公司吗？

查到这里，我原本以为中国或许缺少这种技术，后来发现并不是。

中兴早在 2013 年就展示过 Eco-Mobius 模块化手机概念机，设想把屏幕、核心计算、电池和摄像头做成可替换模块。它很有想象力，但概念展示不等于成熟量产和长期零件供应。

联想 ThinkPad 则提供了更接近现实的案例。2026 年发布的 ThinkPad T14 Gen 7 和 T16 Gen 5 获得 iFixit 10/10 可维修性评分，强调可更换电池、键盘、部分机型的 LPCAMM2 内存以及维修资料。这说明中国企业已经在主流电脑上推进可维修设计，但商业路线和 Framework 的开放模块生态仍有区别。

还有 DeepComputing 为 Framework Laptop 13 开发的 RISC-V 主板。它特别有意思：一个中国背景的硬件开发团队，不必从头再做一台笔记本，也能参与另一个品牌的模块化生态。

小米在 2025 年展示过可拆卸影像模块概念，着重探索手机摄影能力的扩展。这同样是模块化，但目标是新增功能，而不是像 Fairphone 那样围绕维修建立整套体系。

所以，中国并不是没有模块化设计，而是这些探索分布在不同地方：有概念机，有商用维修设计，有开发板和第三方模块，也有功能扩展原型。我暂时没有找到一个能在中国大陆主流消费市场中，与 Fairphone 的长期维修生态完整对应的量产手机品牌。

真正值得追问的，也许不是“我们会不会做”，而是“为什么还没有很多企业选择把它做成长期的产品和服务”。

06｜为什么我们身边很少有人使用它们？

最直接的一点是购买和服务的可获得性。Framework 官方支持配送德国等国家和地区，但目前没有把中国大陆列入官方发货范围。一个零件号称容易更换，如果你所在的地区买不到、售后不覆盖，它的优势就会明显打折。

另外，我们购买手机时往往先比较性能、拍照、外观、价格、应用生态和售后。模块化需要更多接口和固定结构，可能带来成本、空间、防水或重量上的取舍。它不天然比一体化设计更适合所有人。

德国的情况也不能简单归结为“德国人都更环保”。我只认识了一个使用这两种产品的人，无法据此推断整个国家的消费偏好。不过，在欧洲，维修权确实有更明显的制度背景：欧盟从 2025 年 6 月 20 日起实施面向新上市智能手机和平板的生态设计与能效标签要求，涉及关键零件供应、维修、系统更新和可维修等级等。

我现在更能理解：一种产品是否被选择，不只取决于用户价值观，还取决于当地能不能买到、维修方不方便、法规怎样规定，以及品牌能否长期兑现承诺。

07｜如果我是产品经理，我应该学到什么？

这次研究让我记下了四个问题。

第一，产品到底在解决什么，而不是使用了什么技术？ Framework 没有把所有零件做成可以随意拼装的积木；Fairphone 也没有承诺用户能自由升级芯片。它们都在选择真正值得拆开的部分。

第二，交付之后的体验算不算产品的一部分？ 我以前容易把产品理解为上线、交付和使用，但电池老化、零件维修、系统更新、二手流转，可能才是一个长期产品关系真正的考验。

第三，价值观怎样变成可验证的机制？ “重视环保”是态度，提供可买到的零件、维修手册、系统支持年限和生命周期数据，才是用户能够检查的承诺。

第四，商业模式能否和用户利益站在一起？ 如果用户越能长久使用产品，企业反而越难经营，那么理念很容易和现实冲突。Framework 的模块市场和 Fairphone 的长期服务让我开始思考：有没有办法让企业因为帮助用户延长设备寿命，而不是仅仅促使他们换新，获得持续收入？

这些问题也可以拿来审视我正在接触的 AI 产品。一个 AI 工具也许可以很快做出 Demo，但用户以后怎么维护数据、迁移服务、理解系统限制，甚至在产品停止运营后怎么办？这不也是一种“产品能否长期属于用户”的问题吗？

08｜还有一些问题，我不想急着回答

我还没有实际使用过 Framework 和 Fairphone，也没有亲手拆开它们。本文的技术和环保判断主要来自公开资料，品牌承诺与实际体验之间，仍然需要更多独立验证。

我也想有机会问问那位德国朋友：他为什么选择这两个品牌？真的维修过吗？零件好买吗？这些便利有没有抵消价格或性能上的取舍？我不想提前替他给出答案。

这次最有意思的地方，或许是我原本只是在认识一个人，却意外认识了另一套产品设计逻辑。

买下一台设备，不应该只是在购买它今天的性能。我们也可以关心，五年之后它坏了怎么办，能不能继续使用，以及这些选择究竟掌握在谁手里。

我想把这次研究先留在 Notes。等真正找到可验证的具体问题，再把它做成 Lumi Lab 的拆解与维修交互实验。

参考资料

资料核对于 2026 年 11 月；本文区分官方披露、第三方报道与个人判断。

What a German friend’s Framework laptop and Fairphone taught me about modularity, sustainability, and product design.

01 | It started with two unfamiliar devices

Some time ago, I met a German guy. In conversation, I noticed his laptop was a Framework and his phone was a Fairphone.

I wasn’t familiar with either brand. Compared with what I usually see — Apple, Huawei, Xiaomi, Lenovo — the first thing that caught my attention wasn’t processor benchmarks or camera specs, but something that sounds almost counterintuitive: users can open them up, repair them, and replace parts themselves.

I started looking into American brands of user-assemblable computers, then modular phones, and eventually even Linux phones. The question slowly shifted from “what brand is this” to “why would anyone choose devices like these.”

I still don’t know his exact reasons for choosing these two brands, and I don’t want to answer for him. Maybe he enjoys tinkering with hardware; maybe he cares about longevity or privacy; or maybe he just thinks they’re good products. But his choice made me realize: the consumer-electronics design we take for granted is not the only option.

02 | What problems are Framework and Fairphone actually solving?

Framework is an American laptop company. Its laptops make the battery, storage, RAM, and ports easier to remove and replace. Some models let you upgrade the processor platform by swapping the entire mainboard; the Framework Laptop 16 also offers swappable discrete GPU modules. The 2026 Laptop 13 Pro uses LPCAMM2 memory, balancing low power with replaceability.

But it’s not “laptop LEGO” either. The CPU is still soldered to the mainboard, and GPU modules are constrained by Expansion Bay specifications and model compatibility. Framework publishes the bay’s mechanical and electrical specifications, but modularity doesn’t mean any GPU can be freely mixed and matched.

Fairphone is a Dutch phone company with a slightly different focus. The Fairphone Gen 6+ offers 12 categories of user-replaceable parts, including the battery, screen, cameras, and USB-C port, with promised software support through 2033. But you can’t upgrade its RAM or processor the way you’d add memory to a computer.

So here’s how I prefer to understand them: Framework treats “what can still be upgraded later” as the product question; Fairphone treats “how do you keep using it after it breaks” as the product question. Both care about repair, but they are not the same kind of modularity.

Another concept I had to rethink was “openness.” A replaceable battery doesn’t mean you’re free to install any OS; Linux support doesn’t mean all chip designs are public. Repairability, upgradability, open software, and open hardware are four separate judgments.

03 | Is modularity actually greener?

This is the part I most wanted to verify — and the part I’m least willing to conclude quickly.

Framework’s approach is extending device lifespan: swap the battery when it ages; upgrade the mainboard on supported models when performance falls short. More interestingly, an old mainboard can go into a standalone case and live on as a small desktop computer. Framework also commissioned a lifecycle assessment of its laptop from Germany’s Fraunhofer IZM.

Fairphone’s ambitions are broader. Beyond repair, it focuses on raw-material sourcing, recycled materials, factory labor conditions, and e-waste. According to its 2025 impact report, the Fairphone Gen 6’s product carbon footprint is roughly 29 kg CO₂e, and the company estimates that extending device lifespans helped avoid about 2,083 tonnes of CO₂e emissions in 2025.

These numbers are worth reading — and worth reading carefully, including the methodology. Avoided-emissions figures are estimates based on lifespan and counterfactual scenarios, not measurements taken by strapping instruments onto every old phone; and carbon footprints across brands and models can’t simply be ranked side by side.

For me, the questions that really matter aren’t “does it have modules” but: will users actually repair? How many years can parts be bought? How much cheaper is repairing versus replacing? How many extra years does the device actually live?

Repairability is a possibility; real sustainability still has to be proven by usage behavior and data.

04 | Can this kind of business actually make money?

I find this more worth studying than the phrase “eco-friendly phone.”

In a public strategy essay, Framework’s founder laid out the long-term direction in three parts: products, the Marketplace, and a repair service network. First sell devices that can be repaired and upgraded; then build ongoing transactions and services around mainboards, ports, accessories, third-party modules, and refurbished or used parts.

There’s an interesting commercial shift here: the customer’s relationship with the brand doesn’t end at purchase. Replacing the battery two years later, upgrading the mainboard four years later, even finding a second life for the old mainboard — all of it can become part of the ecosystem. That said, a strategy document doesn’t mean every repair service has already maturely landed in every region.

Fairphone, meanwhile, tries to turn durability, repairability, long-term software support, and a more responsible supply chain into value people will pay for. Its 2025 report and official summary disclosed year-on-year growth of over 40% in devices sold, but growth in device sales alone doesn’t prove long-term profitability.

Both companies face an unromantic reality: long-term parts supply requires inventory, cross-generation compatibility requires R&D, and small scale can mean higher unit costs. A good idea that can’t cover these costs ends up as nothing more than an exquisite niche hobby.

So I don’t think they’re proving “sustainability always pays.” They’re trying to answer: can keeping users on a product for the long run itself become a sustainably run business?

05 | Does China really have no similar companies?

Researching this, I first assumed China might lack the technology. It turns out that’s not the case.

ZTE showed off the Eco-Mobius modular phone concept as early as 2013, envisioning swappable modules for the screen, compute core, battery, and camera. It was imaginative — but a concept demo is not mature mass production with long-term parts supply.

Lenovo’s ThinkPad offers a more realistic case. The ThinkPad T14 Gen 7 and T16 Gen 5 released in 2026 received iFixit repairability scores of 10/10, emphasizing replaceable batteries, keyboards, LPCAMM2 memory on some models, and repair documentation. This shows Chinese companies already pushing repairable design in mainstream laptops — though the business route still differs from Framework’s open module ecosystem.

There’s also the RISC-V mainboard that DeepComputing built for the Framework Laptop 13. What makes it especially interesting: a Chinese hardware team can participate in another brand’s modular ecosystem without building a whole laptop from scratch.

And Xiaomi showed a detachable camera-module concept in 2025, exploring extensions to mobile photography. That’s modularity too, but aimed at adding capabilities — not at building an entire system around repair the way Fairphone does.

So China doesn’t lack modular design; these explorations are just scattered across different places: concept phones, commercial repairable designs, dev boards and third-party modules, and feature-extension prototypes. What I haven’t yet found is a mass-market phone brand in mainland China that fully parallels Fairphone’s long-term repair ecosystem.

The question really worth asking may not be “can we build this” but “why haven’t more companies chosen to turn it into a long-term product and service.”

06 | Why do so few people around us use them?

The most direct factor is availability of purchase and service. Framework officially ships to countries and regions including Germany, but mainland China is not in its official shipping range. A part may be advertised as easy to replace — but if you can’t buy it where you live and after-sales doesn’t cover you, the advantage wears thin.

Besides, when we buy phones we tend to compare performance, cameras, looks, price, app ecosystem, and after-sales first. Modularity requires more connectors and mounting structures, with trade-offs in cost, thickness, water resistance, or weight. It isn’t inherently better for everyone than integrated design.

Nor can the German situation be boiled down to “Germans are all more eco-conscious.” I’ve met exactly one person who uses these products; I can’t infer a nation’s preferences from that. Still, in Europe the right to repair does have a clearer institutional backdrop: since June 20, 2025, the EU has applied ecodesign and energy-label requirements to newly launched smartphones and tablets, covering spare-part availability, repair, system updates, and repairability ratings.

What I understand better now: whether a product gets chosen depends not just on user values, but on whether you can buy it locally, how convenient repair is, what the regulations say, and whether the brand can keep its promises for years.

07 | If I were the product manager, what should I learn?

This research left me with four questions.

First, what is the product actually solving — not what tech does it use? Framework didn’t turn every part into freely combinable bricks; Fairphone never promised you could freely upgrade the chip. Both chose the parts genuinely worth making removable.

Second, does the post-delivery experience count as part of the product? I used to think of a product as launch, delivery, and usage. But battery aging, parts repair, system updates, and second-hand circulation may be the real tests of a long-term product relationship.

Third, how do values become verifiable mechanisms? “We care about the environment” is an attitude. Parts you can actually buy, repair manuals, years of OS support, and lifecycle data are promises users can check.

Fourth, can the business model align with user interests? If the longer users keep a product, the harder it is for the company to survive, the idea will keep colliding with reality. Framework’s module marketplace and Fairphone’s long-term services got me thinking: is there a way for a company to earn recurring revenue precisely by helping users extend their devices’ lives — not just by pushing them to upgrade?

These questions also apply to the AI products I’m learning about. An AI tool can ship a demo quickly, but how will users maintain their data, migrate services, understand system limits — even deal with the product shutting down? Isn’t that also a question of “whether a product can truly belong to its users over time”?

08 | Questions I don’t want to rush to answer

I have never actually used a Framework or Fairphone, never opened one up. The technical and environmental judgments in this article come mainly from public sources; between brand promises and real-world experience, more independent verification is still needed.

I’d also like the chance to ask that German friend: why did he choose these two brands? Has he actually repaired them? Are the parts easy to buy? Did the convenience outweigh the price or performance trade-offs? I don’t want to answer for him in advance.

Perhaps the most interesting part of this episode is that I set out to get to know a person, and ended up getting to know a whole different logic of product design.

Buying a device shouldn’t just be buying its performance today. We can also care about what happens when it breaks in five years, whether it can keep working — and whose hands those choices actually rest in.

I’ll leave this research in Notes for now. Once I find a verifiable, concrete question, I’ll turn it into a teardown-and-repair interactive experiment in Lumi Lab.

References

Sources checked in November 2026; this article distinguishes official disclosures, third-party reporting, and personal judgment.

不是每件事都需要一个标签。

探索线索