漫画解析 | 捅破天花板的液冷极限
捅破天花板：探索浸没式液冷的局限性
Breaking the Ceiling: Exploring the limitations of immersion cooling
人工智能技术的快速发展刺激着智能算力需求的爆发性增长，液冷技术得益于此，有了更大的展示舞台。浸没式液冷作为液冷技术之一，有着优秀的冷却潜力，但浸没式液冷将基础设施、冷却液、IT设备紧密耦合在一起，给这一技术方案的落地推广带来更大的挑战。这意味着浸没式液冷领域存在着更多创新空间，需要高校、研究院、设备制造商、数据中心客户等生态力量协同创新，共同引领定义数据中心冷却未来方式。
作为一个深深沉浸在数据中心冷却技术世界的人，我经常发现自己在谈论浸没式液冷的潜力和局限性。将昂贵而复杂的电子元件浸泡在液体中以保持冷却的想法听起来更像是科幻小说，而不是可行的解决方案。但这是真的，而且已经引起了轰动。然而，当我与业内同事和客户交谈时，我遇到的一个反复出现的主题是浸没式液冷的“天花板”。有一种观点认为，虽然它可能是创新的，但它的能力有限，可能跟不上现代计算不断增长的需求。
As someone deeply immersed in the world of cooling technologies for data centers, I often find myself in conversations about the potential and limitations of immersion cooling. The very idea of submerging expensive and intricate electronic components in liquid to keep them cool sounds more like science fiction than a viable solution. But it is real and already making a splash. However, when I talk to industry colleagues and clients, a recurring theme I encounter is the perceived “ceiling” of immersion cooling. There's a notion that while it may be innovative, its capacities are limited and might not keep up with the ever-growing demands of modern computing.
如果我说我们只是触及了表面呢？这可能会让一些人感到惊讶，但很有可能浸没式液冷的真正创新才刚刚开始。在这篇文章中，我想带你踏上一段浸没式液冷的旅程，更重要的是，凝视这个水晶球，了解等待解锁的潜力。我们将探索为什么这个行业需要拥抱浸没式液冷创新，就像我们在风冷中看到的那样。我邀请您和我一起深入了解数据中心冷却技术的下一波浪潮。
What if I suggested we’ve only just scratched the surface? It might surprise some, but it is more than possible that true innovation in immersion cooling has barely begun. In this article, I want to take you on a journey through the present landscape of immersion cooling, and more importantly, gaze into the crystal ball to understand the untapped potential waiting to be unlocked. We will explore why this industry needs to embrace innovation in immersion just as we witnessed in air cooling. I invite you to take a deep dive with me into what might just be the next big wave in data center cooling technologies.
作为一名浸没式液冷的倡导者和观察者，我对这项技术迄今所展示的纯粹能力和弹性着迷。目前主流使用的方法实际上是相当初级的。本质上，液体在电子元件周围循环，而这个过程并不对液体流向作特别要求。值得注意的是，尽管表面上看起来很简单，但这种方法已经证明了它能够处理几乎所有现有的IT设备的散热问题。
As an advocate and a keen observer of immersion cooling, I’m fascinated by the sheer capability and resilience this technology has displayed so far. The methods being used predominantly at present are quite rudimentary in nature. Essentially, fluids are circulated around the electronics, and the process doesn't pay particular attention to the path these fluids take. The remarkable thing is, despite the seeming simplicity, this approach has proven itself capable of handling almost anything that current IT equipment can challenge it with thermally.
然而，当我与各方利益相关者接触时，我经常面临浸没式液冷的“天花板”问题。特别是随着现代电子产品功率密度的迅速增加，许多人想知道它的处理能力是否有极限。我看到了这些问题背后的好奇心，这让我想起了这个领域创新的真正本质。到目前为止，浸没式液冷已经取得了巨大的突破，但它还没有真正突破任何界限。这本身就证明了它的潜力。
However, as I engage with various stakeholders, I am often confronted with the question of the perceived “ceiling” of immersion cooling. Many wonder if there’s a limit to what it can handle, especially with the rapidly increasing power densities in modern electronics. I see the curiosity behind these questions, and it reminds me of the real essence of innovation in this space. Up till now, immersion cooling has been a tremendous breakthrough, but it hasn’t really had to push any boundaries. This is by itself a testament to its potential.
目前的方法虽然对当前的需求有效，但往往缺乏我们进入更高热负荷时代时可能需要的技巧和精度。许多人谈论的这个“天花板”，实际上与其说是对浸没式液冷技术的限制，不如说反映了我们还没有触及到可能的表面的事实。
The current methods, while effective for current demands, often lack the finesse and precision that might be necessary as we move into an era of even higher thermal loads. This ‘ceiling’ that many talk about is in fact not so much a limitation of immersion cooling as a technology, but rather a reflection of the fact that we haven’t yet scratched the surface of what’s possible.
回顾一下风冷的发展历程是有益的。这可以帮助理解技术是如何在先进硬件的需求推动下发展的。
It's instructive to take a look back at the evolution of air cooling. This serves as a context for understanding how technologies can, and do, evolve when pushed by the demands of advancing hardware.
几年前，普遍的共识是，风冷技术能够处理芯片散热的天花板是120W。但随后，一件了不起的事情发生了。随着CPU和GPU技术的加速发展，对冷却系统提出了更高的要求，该行业找到了创新的方法。风冷解决方案变得更有效，有更好的气流设计、散热器和更有效的风扇。此外，在设施内，通道隔离和空调入列在短短几年内已成为标准做法。
A few years ago, the general consensus was that air cooling would not be able to handle anything beyond 120W. It was thought to be the ceiling. But then, something remarkable happened. As CPU and GPU technology accelerated and placed greater demands on cooling systems, the industry found ways to innovate. Air cooling solutions became more efficient, with better airflow designs, heat sinks, and more effective fans. Also within the facilities, aisle separation and in-row coolers have become standard practice within only a few years time.
从本质上讲，这种持续的创新是行业对确保风冷解决方案满足热负荷需求增长的回应。然而，风冷正在接近它的极限，创新曲线显示出稳定的迹象，因为创新已经达到了可能的边缘。这种停滞不是由于缺乏创新，而是与空气作为冷却介质的固有局限性有关。
This continuous innovation was, in essence, the industry's response to ensuring that air cooling solutions stayed relevant as the thermal requirements increased. However, air cooling is approaching its limits, and the innovation curve is showing signs of plateauing as the innovations have reached the margins of what is possible. This plateau is not due to a lack of innovation but relates to the inherent limitations of air as a medium for cooling.
有趣的是，在风冷忙于发展的同时，液体冷却也在悄然成熟。它显示出高度成熟的迹象，我们正在目睹制造商、运营商和最终用户越来越多地采用液冷解决方案的趋势。
Interestingly, while air cooling was busy evolving, liquid cooling was silently maturing alongside. It is showing signs of high maturity, and we are witnessing a trend where manufacturers, operators and end users are increasingly adopting liquid cooling solutions.
这一历史视角告诉我们，在触及方法的基本极限之前，创新的空间总是存在的。它为浸没式液冷行业提供了灵感和经验，因为它显示着前方巨大的可能性和挑战。
What this historical perspective tells us is that there is always room for innovation until the fundamental limits of the medium are reached. It serves as an inspiration and a lesson for the immersion cooling industry as it stares at the immense possibilities and challenges ahead.
那么液体的极限是什么呢？浸没式液冷主要依靠冷却液（通常是介电液体）的热特性来吸收和散发热量。浸没式液冷的换热效率取决于液体的比热容和导热系数，以及对流换热系数。
So what are the limits of liquids? Immersion cooling primarily relies on the thermal properties of the cooling fluid, typically a dielectric liquid, to absorb and dissipate heat. The efficiency of heat transfer in immersion cooling is governed by the specific heat capacity and thermal conductivity of the fluid, as well as the convective heat transfer coefficient.
用于浸没式液冷的介电液体比空气具有更高的比热容，这意味着它们可以在给定的温升时吸收更多的热能。此外，它们的导热性高于空气，使它们能够更有效地将热量从元件中传递出去。
Dielectric liquids used for immersion cooling have a much higher specific heat capacity than air, meaning they can absorb more heat energy for a given temperature increase. Additionally, their thermal conductivity is higher than that of air, allowing them to transport heat away from components more efficiently.
让我们深入研究展示（浸没式液冷）潜力的数字。通常所说的液体的散热能力大约是空气的1000倍，这为巨大的可能性打开了一扇窗。现在考虑一个假设的场景来说明潜力的大小。一些科学研究表明，液体冷却的理论极限非常高，比如在最佳条件下，每平方厘米的散热量超过1kW。虽然这个数字代表了理论上的理想情况，但它足以鼓舞人心。这表明，如果我们能最大限度地发挥液体冷却的全部科学潜力，能实现更多可能性。这远远超过了我们最初看到的大尺寸高密度芯片的2kW。
Let’s delve into the numbers that showcase the potential. The heat dissipation capabilities of liquids are commonly referred to be roughly 1,000 times more effective than air, which opens a window to enormous possibilities. Now consider a hypothetical scenario to illustrate the magnitude of potential. There are scientific explorations that have suggested enormously high theoretical limits for liquid cooling, such as the removal rates of over 1 kW per square centimeter under optimal conditions. While this number represents a theoretical idealized scenario, it serves as an inspiring point on the horizon. It shows what might be achievable if we were to maximize the full scientific potential of liquid cooling. This is far beyond the 2 kW figures for chips that we’re starting to see with large form factor high-density chips.
虽然这并不意味着我们明天就能达到这样的冷却能力，但它凸显了潜力。这些创新会是什么样的呢？时间会告诉我们一切，但有一件事是肯定的，浸没式液冷的“上限”不是一个固定的，而是一个变化的目标，随着流体、材料和系统设计的不断创新，这个目标可以被推得更高。
This doesn’t mean we’ll reach such cooling capacities tomorrow, but it highlights the potential. What would such innovations look like? Time will tell, but one thing is certain, the "ceiling" for immersion cooling is not a fixed mark but rather a moving target that can be pushed higher with ongoing innovation in fluids, materials, and system design.
在浸没式液冷的情况下，这种创新才刚刚开始。尽管浸没式液冷在热管理领域是一种改变游戏规则的解决方案，但它主要依赖于一种简化的方法。通过简单地循环流体，并且不过分关注它们的确切路径，浸没式液冷已经有效地跟上了当前IT硬件所能提供的几乎所有东西。然而，随着硬件的不断发展，这种方法也需要不断发展。
In the case of immersion cooling, this kind of innovation has scarcely begun. Despite being a game-changing solution in the thermal management landscape, immersion cooling has primarily relied on a simplified approach. By simply circulating fluids, and not being overly concerned about the exact paths they take, immersion cooling has effectively kept pace with almost everything the current IT hardware can throw at it. However, as hardware continues to advance, this approach will need to evolve.
想象一下，如果浸没式液冷行业学习风冷行业同样的创新精神，将会取得什么成就。第一个成熟的创新领域是IT系统的优化，特别是浸没式液冷。这涉及到开发新的组件、外形因素和电路布局，这些设计本身就是为了最大限度地提高浸没式液冷的效率，而不是改造风冷设计。
Imagine what could be achieved if the immersion cooling industry applied the same spirit of innovation that drove air cooling to new heights. The first area ripe for innovation is the optimization of IT systems specifically for immersion cooling. This involves developing new components, form factors, and circuit layouts that are inherently designed to maximize the efficiency of immersion cooling, rather than retrofitting air-cooling designs.
第二个吸引创新的领域是精确冷却。目前，浸没式液冷系统中的冷却液循环在很大程度上犹如钝器。相反，如果我们能精确地在系统中最需要冷却的地方进行冷却呢？开发智能、反应灵敏的冷却系统，能够实时定位和解决热点问题，可以成倍地提高浸没式液冷系统的冷却能力和效率。
The second area that beckons the inventive spirit is precision cooling. Currently, the circulation of cooling fluids in immersion cooling systems is largely a blunt instrument. What if, instead, we could direct cooling precisely where it’s needed most within a system? Developing intelligent, responsive cooling systems that can pinpoint and address hotspots in real-time could exponentially increase the cooling capacity and efficiency of immersion systems.
展望未来，随着芯片密度的上升，功率水平有望达到2kW甚至更高，浸没式液冷不仅可以跟上不断创新的步伐，而且还可以引领潮流。这需要拥抱一种创新文化，超越随意的盲目循环，采用一种更优雅、更高效、更有效的方法。浸没式液冷的真正创新曲线才刚刚开始，想想这对整个数据中心行业意味着什么，真是令人兴奋。
As we look to the future, with chip densities on the rise and power levels expected to reach 2 kW and beyond, it should be expected that immersion cooling not only keeps pace with constant innovation, but also leads the way. This will require embracing a culture of innovation that goes beyond the casual blind circulation to a more elegant, efficient, and effective approach. The real innovation curve of immersion cooling is just about to begin, and it’s exciting to think about what this could mean for the data center industry as a whole.
数据中心国际专家培训
ATD设计课程 AOS运维课程运维管理的直接目标是优秀的运维人员而非设备本身；全球权威的AOS运维管理专家课程将带您深刻理解运维管理的本质
ATS管理课程正确和系统地了解Tier分级体系会提升数据中心的项目投资回报、减少业务中断风险，ATS课程将全面带您学习Tier知识，帮助您有效提升企业的运营指标和内外部客户的满意度。
点击图片查看课程排期
扫码回复【uptime培训】了解课程详情
当我们站在浸没式液冷这个激动人心的新时代的风口浪尖上时，行业必须采取深思熟虑的步骤来拥抱和推动这一发展。以下是我认为我们应该关注的重点：
As we stand on the cusp of this exciting new era in immersion cooling, it is imperative for the industry to take deliberate steps to embrace and drive this evolution. Here’s what I believe should be our focus:
协作和社区建设：未来的挑战是复杂的，需要整个生态系统的共同努力。制造商、数据中心运营商、学者和技术创新者需要走到一起，共享知识，协同工作以解决共同的挑战。
Collaboration and Community Building: The challenges ahead are complex and will require the collective effort of the entire ecosystem. Manufacturers, data center operators, academics, and technology innovators need to come together, share knowledge, and work collaboratively to solve common challenges.
研发投资：优化元件和智能冷却系统的开发需要大量的研发投资。油气行业必须认识到这些投资的长期价值，并相应地分配资源。
Investment in Research and Development: The development of optimized components and intelligent cooling systems requires significant investment in research and development. The industry must recognize the long-term value of these investments and allocate resources accordingly.
标准与最佳实践：随着浸没式液冷技术的发展，制定确保一致性、可靠性和安全性的标准和最佳实践非常重要。行业协会和管理机构在这方面可发挥关键作用。
Standards and Best Practices: As immersion cooling technologies evolve, it’s important to develop standards and best practices that ensure consistency, reliability, and safety. Industry associations and regulatory bodies have a key role to play in this regard.
教育和宣传：当涉及浸没式液冷时，仍然存在重大的知识差距和怀疑。教育利益相关者，从高管到技术人员，了解浸没式液冷的好处和潜力是至关重要的。宣传工作还应旨在影响有利于可持续冷却解决方案的政策和监管变化。
Education and Advocacy: There’s still a significant knowledge gap and skepticism when it comes to immersion cooling. It is essential to educate stakeholders, from C-suite executives to technicians, on the benefits and potential of immersion cooling. Advocacy efforts should also aim at influencing policy and regulatory changes that favor sustainable cooling solutions.
以客户为中心的创新：了解数据中心客户的独特需求和约束是关键。在开发浸没式液冷解决方案时，应考虑到客户的需求，确保解决现实世界的挑战并创造切实的价值。
Customer-Centric Innovation: Understanding the unique requirements and constraints of data center customers is key. Immersion cooling solutions should be developed with customer needs in mind, ensuring that they address real-world challenges and create tangible value.
可持续发展焦点：随着全球数据使用量的持续飙升，浸没式液冷技术的发展必须保持对可持续性的关注，减少能源消耗和环境影响。
Sustainability Focus: With global data usage continuing to soar, it is essential that the evolution of immersion cooling technologies maintains a strong focus on sustainability, reducing energy consumption and environmental impact.
通过关注这些关键领域，浸没式液冷行业不仅可以跟上IT硬件不断变化的需求，还可以引领定义数据中心冷却未来的方式。未来的旅程注定充满挑战，但随着对创新、协作和可持续性的共同承诺，我们有机会在未来几年彻底改变数据中心的格局。
By focusing on these key areas, the immersion cooling industry can not only keep up with the evolving demands of IT hardware but lead the way in defining the future of data center cooling. The journey ahead is bound to be challenging, but with a collective commitment to innovation, collaboration, and sustainability, we have the opportunity to revolutionize the data center landscape for years to come.
在这个关键时刻，认识到开放计算项目基金会（#OCP）在这些重点领域所发挥的重要作用是很重要的。作为一项致力于重新设计硬件技术以提高效率的倡议，OCP作为一种催化剂，促进了浸没式液冷行业所需的协作和创新。由于强调开放共享设计和最佳实践，OCP在制定标准、培育社区和推动数据中心解决方案的可持续性方面发挥了重要作用。作为OCP浸没式液冷项目的负责人，我敏锐地意识到这种合作在推进浸没式液冷技术方面的潜力。通过我们在OCP内部的共同努力，我们正在为数据中心行业创造高效、可扩展和可持续的解决方案。
At this juncture, it’s important to recognize the vital role played by the Open Compute Project Foundation (#ocp ) in these areas of focus. As an initiative that is committed to redesigning hardware technology for efficiency, OCP serves as a catalyst for the kind of collaboration and innovation that the immersion cooling industry needs. With its emphasis on open sharing of designs and best practices, OCP is instrumental in setting standards, fostering community, and driving sustainability in data center solutions. As someone who is leading the immersion project within OCP, I am acutely aware of the potential this collaboration holds for advancing immersion cooling technologies. Through our collective efforts within OCP, we are breaking new ground in creating efficient, scalable, and sustainable solutions for the data center industry.
数据中心的世界正处于一个转折点。随着我们在计算能力和芯片密度方面不断突破极限，为我们服务了几十年的冷却解决方案正在达到极限。我的朋友们，这就是浸没式液冷的出现，它不仅是一种选择，而且是未来的必需品。此外，提高数据中心冷却的效率和可持续性可以带来巨大的全球效益，从减少碳足迹到在相同的能源限制下实现更强大的计算。
The world of data centers is at a turning point. As we are pushing the boundaries of what’s possible in terms of computational power and chip density, the cooling solutions that have served us for decades are reaching their limits. This, my friends, is where immersion cooling comes in as not just an alternative but a necessity for the future. Moreover, the global benefits that can emerge from improved efficiency and sustainability in data center cooling are immense, ranging from reduced carbon footprints to enabling more powerful computing within the same energy constraints.
但这不仅仅是要采用浸没式液冷，而是要在浸没式液冷中进行创新。我想给你们留下这样的想法，我们即将见证的浸没式液冷的进化类似于我们在整个社会历史上经历的许多伟大的范式转变。就像我们在风冷、电动汽车、甚至摆脱马车所做的那样，让我们打破这些已知的天花板，创新，并将浸没式液冷塑造成它有潜力成为的强大力量。我们正处于一个惊心动魄的传奇故事的第一幕，就像那些手里拿着爆米花的热情观众一样，我坐在座位的边缘，渴望看到它的发展。
But it’s not just about embracing immersion cooling - it’s about embracing innovation within immersion cooling. I want to leave you with the thought that the evolution we’re about to witness in immersion cooling is akin to what many great paradigm shifts we have undergone in society throughout history. As we did with air cooling, electric cars, even breaking away from the horse and carriage, let’s break those perceived ceilings, innovate, and shape immersion cooling into the formidable force it has the potential to be. We're in the first act of a thrilling saga, and much like the enthusiastic audience with popcorn in hand, I am on the edge of my seat, eager to see how this evolves.
这个行业需要具有前瞻性的思考者、问题解决者和充满激情的个人来深入研究这项技术，并带来创新，推动我们进入数据中心冷却的下一个篇章。随着压力的增加和热量的加剧，让我们沉浸在，真正的，可持续的，高效的，强大的解决方案中。
The industry needs forward-thinkers, problem solvers, and passionate individuals to dive into the depths of this technology and bring forth innovations that will drive us into the next chapter of data center cooling. As the pressures mount and the heat intensifies, let’s immerse ourselves, quite literally, in solutions that are sustainable, efficient, and powerful.
大幕升起，舞台就位——让我们把这场演出演得千古难忘!
The curtain rises, the stage is set - let's make this a performance for the ages!
PS：先“在看”，再“点赞”。
谢谢大家的支持，小编才更有干劲去给大家带来更多更好的文章！
翻译：
SiO2
DKV（DeepKnowledge Volunteer）计划成员
校对：
贾梦檩
阿里云 暖通工程师
DKV（DeepKnowledge Volunteer）精英成员
公众号声明：
本文并非官方认可的中文版本，仅供读者学习参考，不得用于任何商业用途，文章内容请以英文原版为准，本文不代表深知社观点。文中内容来自互联网，如有侵权，将在24小时内删除。中文版未经公众号DeepKnowledge书面授权，请勿转载。
推荐阅读：
