《投资者网》吴微
不出意外，中秋国庆假期前的9月，消费电子市面依旧热闹。
作为市面的焦点，苹果（AAPL）在9月发布了iPhone 18 Pro系列，并首次推出折叠屏iPhone Duo；华为则提前于9月7日推出Mate XT 2三折叠及HarmonyOS 7；vivo在9月21日发布X500系列，OPPO于9月22日推出Find X10系列，小米集团（01810.HK）也于9月23日发布小米18 Pro系列，荣耀则将Magic 9系列发布会定于9月28日。
从这些新品可以看到，行业竞争正在从过去的芯片、显示屏、快充等单项参数比拼，转向折叠形态、专业影像、端侧AI、自研或先进制程芯片以及电池续航能力的综合竞争。
但另一面，消费电子市面却并没有因为新品密集发布而同步进入景气上行周期。中国信通院数据表明，2026年上半年国内手机出货量为1.33亿部，同比下降5.5%；其中智能手机出货量1.26亿部，同比下降3.0%。6月单月国内手机出货量更降至1914.9万部，同比下降15.3%。
供给端不断推新，需要端却仍然承压。更值得关注的是，存储芯片售价上涨正在成为连接两者的一条关键纽带，它一维度推高手机、PC等终端成本，另一维度也迫使厂商寻找新的机型结构和工艺路径。
新机密集，手机竞争从“堆料”转向综合创新
今年秋季新品潮最直观的变化，是旗舰手机的工艺竞争正在变得更加“立体”。
苹果此次iPhone 18 Pro系列将重点放在A20 Pro、端侧AI、影像和电池续航上。苹果官方披露，A20 Pro使用2纳米制程，并增加神经网络相关计算能力；同时，新机引入可变光圈，并借助新的散热结构提高持续表现。
图片来源：苹果官网
首款折叠机型iPhone Duo则进一步补齐了苹果在折叠屏市面的机型形态，成为苹果近年来最大的工艺革新；iPhone Duo也因而成为苹果此次秋季发布会的宠儿，备受市面关注。
华为的路线则更加鲜明。9月7日发布的Mate XT 2继续押注三折叠，同时将HarmonyOS 7、大屏交互、芯片与平台协同等能力结合起来。新华网指出，折叠屏行业的关注点正从“能否做出新形态”转向结构可靠性、握持感受以及软件任务流等日常使用感受。
其他国机型牌的新品路线也展现出类似的趋势。
vivo X500系列继续突出专业影像、长焦和大电池；OPPO Find X10系列则把双2亿像素镜头、8000mAh电池和ColorOS 17等能力结合起来；小米18 Pro则将高规格影像、新一代旗舰芯片以及澎湃OS 4等作为核心卖点。
这意味着，过去“更大内存、更高像素、更快充电”的参数竞赛仍然存在，但已经不是唯一答案。
在新的市面环境中，手机厂商需要回答的难题正在改变。折叠屏能否真正提高使用效率？AI能否减少使用者操作步骤？影像能否形成可感知的差异？更高算力能否转化为电池续航和长期使用感受？这些难题都在考验着厂商。
随着各大厂商新品的陆续发布，这种综合能力的竞争就越显著。
内存涨价，正在改变消费电子的“量价关系”
新机发布越发热闹，市面则越关心内存成本上涨后，厂商如何处理成本与售价间的关系。
2025年以来，AI服务器建设持续增加对HBM、高容量DDR等存储机型的需要，上游厂商因而不断调整产能结构。TrendForce数据表明，2026年二季度LPDDR4X售价预计环比上涨70%至75%，LPDDR5X上涨78%至83%；进入三季度后，移动DRAM售价涨幅有所收窄，总体预计环比上涨约8%至13%，但售价仍处于高位。
因而，与其简单理解为“内存还在疯狂涨价”，不如说消费电子正在面对一轮持续时间较长的高成本压力。
这种压力已经传导至整机市面。
图片来源：小米官网
IDC在8月最新预测中，将2026年全球智能手机出货量预期下调至略高于10亿部，同比下降16.7%，为该机构有记录以来最大年度降幅；与此同时，全球智能手机平均售价预计上涨27.6%至581美元，手机市面总价值反而预计增长6.3%至6130亿美元。
由此形成一个颇为特殊的“量跌价升”的局面。
对于手机厂商而言，涨价并不是没有代价。TrendForce已经指出，随着存储售价处于高位，消费者端的售价承受能力接近极限，手机厂商因而趋向更谨慎的生产和采购策略。
这种变化在PC市面同样显著。IDC预计，2026年全球PC出货量将下降11.3%，四季度同比跌幅可能达到20%；与此同时，PC平均售价预计上涨18.3%。IDC认为，存储供应紧张至少要持续到2027年底，市面稳定可能要等到2028年。
这也解释了为何消费电子行业开始浮现“涨价与规格调整并存”的现象。
过去几年，厂商不断将大内存、超大存储向中端机型下放，如今随着存储成本抬升，继续单纯增加容量的经济性显著下降。机型规划由“规格升级”转向更加精细的成本控制，或已成为市面竞争的新常态。
AI成为主线，但硬件成本决定普及速度
如果说存储涨价是现实约束，那么AI则是手机乃至于整个消费电子行业寻找下一轮增长的关键工艺方向。
从今年新品可以看到，AI正在从独立能力逐步进入操作平台和底层芯片。
图片来源：豆包手机官网
苹果就将Apple Intelligence和Siri AI进一步嵌入平台；华为HarmonyOS 7强化平台级AI能力；OPPO、荣耀、小米等品牌也在推动AI Agent、跨应用协同以及端侧智能的发展；豆包手机也将发售。
其背后的工艺逻辑很清晰，过去使用者借助一个个App完成任务，而AI时代的终端希望理解使用者意图，再自动调用多个应用和平台能力完成任务。
这意味着手机厂商的竞争对象也在发生变化。
将来手机不只是“运行App的硬件”，而可能逐步成为个人智能中枢。处理器因而不再单纯比CPU、GPU表现，NPU算力、内存带宽、散热效率以及端云协同能力都会变得关键。
苹果A20 Pro的外观已经体现出这一趋势。其新一代神经引擎用于兼容端侧AI和计算摄影，同时借助芯片封装、散热和电池等环节改善持续表现。
但难题也随之浮现。AI越深入终端，对算力、内存和能耗的要求越高，而这些恰恰与当前的成本压力形成矛盾。
TrendForce近期指出，AI服务器需要正在持续挤压传统DRAM供给，而手机品牌则开始借助降低容量、优化机型结构等方式应对成本压力。
因而，AI手机短期内未必意味着所有手机都需要“更大的内存”。
相反，厂商可能更加依赖模型量化、端云协同、平台级调度和芯片架构优化，在有限硬件条件下提高AI感受。
这也是下一阶段工艺竞争的关键。厂商们不是简单把更大的模型塞进手机，而是让AI能力在成本、功耗和感受之间找到平衡。
从手机到AI眼镜，消费电子进入结构分化期
从更大的消费电子市面看，行业正在浮现一个越来越显著的变化。随着工艺进步陷入瓶颈，传统成熟品类继续承压，新形态机型开始寻找增量空间。
IDC预计，2026年全球平板电脑出货量将下降8.6%至约1.39亿台；PC市面同样进入调整周期。
但折叠屏仍然属于少数保持增长的手机细分领域。IDC预计，2026年全球折叠屏手机出货量将增长12.6%至2290万部，并预计2027年进一步增长18%。
这与今年厂商新品密集押注折叠屏的现象形成呼应。
更值得关注的是，AI正在推动消费电子从“显示屏中心”向更加自然的交互形态延伸。
图片来源：小米官网
IDC数据表明，2026年二季度全球智能眼镜及XR眼镜出货量同比增长35.3%，其中无表明智能眼镜占二季度出货量的70.3%；IDC则预计，2026年智能眼镜总体出货仍将保持较快增长。
Omdia的数据同样表明，2026年上半年全球AI眼镜出货量达到420万部，同比增长127%。其增长主要来自拍摄、开放式音频和AI助手等相对明确的应用情形。
这说明，消费电子的下一轮创新不一定来自手机本身。
手机当前仍然会是核心计算和生态入口，但AI眼镜、耳戴设备以及其他智能终端正在成为新的交互节点。随着AI逐步从“能力”变成“平台能力”，不同设备之间的连接能力也会成为新的竞争壁垒。
不过，行业也需要面对另一面。
当前消费电子市面的需要并未全面恢复，AI真正可以形成高频使用的“杀手级情形”仍在探索；与此同时，存储成本高企又抬高了AI终端的硬件门槛。
因而，将来两三年，消费电子市面或许很难再回到单纯依赖出货量增长的时代。
一端是高端手机、折叠屏以及具备显著工艺差异的新品，承担提高单机价值的任务；另一端是AI眼镜、耳戴设备等新形态终端寻找新增量；中低端市面则更多考验厂商的成本控制、供应链管理和机型取舍。
对于手机厂商而言，新品发布依旧关键，但“发布什么”正在比“发布多少”更加关键。
从折叠屏到端侧AI，从专业影像到大电池，再到AI眼镜等新形态设备，消费电子正在从过去的参数竞争进入工艺、生态与成本共同决定机型价值的新阶段。
而内存售价的变化，或许正是这轮行业转型中最现实的一道压力测试题。
您多久换一次手机？欢迎留言、评论、转发。（思维财经出品）■
消费电子
本文来源：https://c.m.163.com/news/a/L8ACDJTQ0519C3MG.html
若有版权问题，请来信告知：13951431913#139.com 或在右方提交评论。
Open NetEase News for more images
Open NetEase News for more images
'Investor Network' · Wu Wei
As expected, September — the month before the Mid-Autumn and National Day holiday — kept the consumer electronics market lively.
As the focus of the market, Apple (AAPL) released the iPhone 18 Pro series in September and for the first time launched the foldable iPhone Duo; Huawei brought forward the Mate XT 2 tri-fold and HarmonyOS 7 on September 7; vivo launched the X500 series on September 21, OPPO introduced the Find X10 series on September 22, and Xiaomi (01810.HK) also released the Xiaomi 18 Pro on September 23.
From these new products, industry competition is visibly shifting from single-parameter contests in chips, displays and fast charging, toward comprehensive competition on foldable form factors, professional imaging, on-device AI, in-house or advanced-process chips, and battery endurance.
But on the other side, the consumer electronics market did not enter an upcycle in sync with the dense product launches. CAICT data shows that domestic phone shipments in the first half of 2026 were 133 million units, down 5.5% year on year; of those, smartphones were 126 million units, down 3.0%. June alone saw domestic phone shipments fall to 19.149 million units, down 15.3% year on year.
The supply side keeps launching new products while demand remains under pressure. More worth watching is that rising memory chip prices are becoming a key link between the two — pushing up the cost of phones, PCs and other devices on one hand, while forcing vendors to seek new model structures and process paths on the other.
Dense new launches mean phone competition is shifting from 'stacking specs' toward comprehensive innovation
The most intuitive change in this autumn's new product wave is that flagship phone process competition has become more 'three-dimensional'.
Apple's iPhone 18 Pro series this time focuses on the A20 Pro, on-device AI, imaging and battery life. Apple disclosed that the A20 Pro uses a 2nm process and adds neural-network-related compute capability; the new phone also introduces a variable aperture and uses a new thermal structure to improve sustained performance.
Open NetEase News for more images
Image source: Apple official website
The first foldable model, the iPhone Duo, further completes Apple's foldable form-factor lineup and becomes Apple's biggest process innovation in recent years; the iPhone Duo has therefore become the darling of Apple's autumn launch event, attracting wide market attention.
Huawei's approach is more distinctive. The Mate XT 2 released on September 7 continues to bet on the tri-fold, while combining HarmonyOS 7, large-screen interaction, chip and platform synergy. Xinhua noted that the foldable industry's focus is shifting from 'whether a new form factor can be built' toward structural reliability, grip feel and everyday usage such as software task flows.
New products from other domestic brands show similar trends.
The vivo X500 series continues to emphasise professional imaging, telephoto and large batteries; the OPPO Find X10 series combines dual 200-megapixel lenses, an 8000mAh battery and ColorOS 17; the Xiaomi 18 Pro centres on high-spec imaging, a new-generation flagship chip and HyperOS 4.
This means the old parameter race of 'more memory, higher resolution, faster charging' still exists, but is no longer the only answer.
In the new market environment, the questions phone vendors must answer are changing. Can foldables genuinely improve usage efficiency? Can AI reduce the number of steps users take? Can imaging produce a perceptible difference? Can higher compute translate into battery life and long-term experience? All of these are testing vendors.
As the various vendors roll out their new products, this competition in comprehensive capability becomes ever more pronounced.
Memory price rises are changing the 'volume-price relationship' in consumer electronics
With new launches growing ever livelier, the market cares more about how vendors handle the relationship between cost and pricing after memory costs rise.
Since 2025, continued AI server construction has driven growing demand for HBM and high-capacity DDR memory, so upstream vendors have constantly adjusted capacity structures. TrendForce data shows that in Q2 2026 LPDDR4X prices are expected to rise 70% to 75% quarter on quarter and LPDDR5X by 78% to 83%; entering Q3, mobile DRAM price gains narrowed somewhat, with an overall expected rise of about 8% to 13% quarter on quarter, but prices remain at high levels.
So rather than simply reading it as 'memory is still going crazy up', it is better to say consumer electronics faces a longer-lasting wave of high cost pressure.
This pressure has already passed through to the device market.
Open NetEase News for more images
Image source: Xiaomi official website
In IDC's latest August forecast, 2026 global smartphone shipment expectations were cut to slightly above 1 billion units, down 16.7% year on year — the largest annual decline in the agency's records; meanwhile, global smartphone average selling prices are expected to rise 27.6% to 581 US dollars, and the total market value is instead expected to grow 6.3% to 613 billion dollars.
This forms a rather special situation of 'volume down, price up'.
For phone vendors, raising prices is not without cost. TrendForce has pointed out that with storage prices at highs, consumer price tolerance is approaching its limit, so vendors are trending toward more cautious production and procurement strategies.
This change is just as notable in the PC market. IDC expects global PC shipments to fall 11.3% in 2026, with Q4 potentially down 20% year on year; meanwhile, PC average selling prices are expected to rise 18.3%. IDC believes tight storage supply will last at least until the end of 2027, with market stability possibly not arriving until 2028.
This also explains why the consumer electronics industry has begun to see 'price increases coexisting with spec adjustments'.
Over the past few years vendors have continually pushed large memory and huge storage down into mid-range models. Now that storage costs have risen, the economics of simply adding capacity have fallen significantly. Model planning shifting from 'spec upgrades' toward finer cost control may already have become the new norm in market competition.
AI is the main thread, but hardware cost determines how fast it spreads
If rising memory prices are the real constraint, then AI is the key process direction for phones and indeed the entire consumer electronics industry in search of the next round of growth.
From this year's new products, AI is gradually moving from a standalone capability into operating systems and underlying chips.
Open NetEase News for more images
Image source: Doubao phone official website
Apple embeds Apple Intelligence and Siri AI further into the system; Huawei's HarmonyOS 7 strengthens system-level AI capabilities; OPPO, Honor, Xiaomi and other brands are also advancing AI agents, cross-app collaboration and on-device intelligence; and the Doubao phone is also going on sale.
The process logic behind this is very clear: in the past users completed tasks through individual apps one by one, whereas AI-era devices aim to understand user intent and then automatically invoke multiple apps and platform capabilities to complete tasks.
This means the thing phone vendors compete against is also changing.
In future the phone will be not just 'hardware that runs apps' but may gradually become a personal intelligent hub. Processors are therefore no longer compared on CPU and GPU performance alone — NPU compute, memory bandwidth, thermal efficiency and device-cloud coordination all become critical.
The exterior of Apple's A20 Pro already reflects this trend. Its new-generation Neural Engine is built for on-device AI and computational photography, while chip packaging, thermals and battery work together to improve sustained performance.
But difficulties follow. The deeper AI goes into the device, the higher the demands on compute, memory and energy consumption — and these precisely conflict with current cost pressure.
TrendForce recently noted that AI server demand continues to squeeze traditional DRAM supply, while phone brands are beginning to respond to cost pressure by reducing capacity and optimising model structure.
Therefore, AI phones in the short term may not mean every phone needs 'more memory'.
Instead, vendors may rely more on model quantisation, device-cloud coordination, system-level scheduling and chip architecture optimisation to improve the AI experience within limited hardware constraints.
This is also the key to the next stage of process competition. Vendors are not simply stuffing bigger models into phones, but finding a balance for AI capability across cost, power consumption and experience.
From phones to AI glasses, consumer electronics enters a period of structural divergence
Looking at the broader consumer electronics market, the industry is showing an increasingly notable shift. As process progress hits bottlenecks, traditional mature categories remain under pressure while new-form-factor products search for incremental growth.
IDC expects global tablet shipments to fall 8.6% to about 139 million units in 2026; the PC market likewise enters an adjustment cycle.
But foldables remain one of the few phone segments still growing. IDC expects global foldable phone shipments to grow 12.6% to 22.9 million units in 2026, and a further 18% growth in 2027.
This echoes this year's dense vendor bets on foldables.
More worth watching is that AI is pushing consumer electronics from a 'display-centric' model toward more natural interaction forms.
Open NetEase News for more images
Image source: Xiaomi official website
IDC data shows global smart glasses and XR glasses shipments grew 35.3% year on year in Q2 2026, with display-less smart glasses accounting for 70.3% of Q2 shipments; IDC expects overall smart glasses shipments to keep growing rapidly in 2026.
Omdia data likewise shows global AI glasses shipments reached 4.2 million units in the first half of 2026, up 127% year on year. Growth came mainly from relatively clear use cases such as shooting, open-ear audio and AI assistants.
This suggests the next round of consumer electronics innovation may not come from phones themselves.
Phones will remain the core computing and ecosystem entry point for now, but AI glasses, ear-worn devices and other smart terminals are becoming new interaction nodes. As AI gradually shifts from a 'capability' to a 'platform capability', the ability to connect between devices also becomes a new competitive barrier.
That said, the industry also has to face the other side.
Demand in the current consumer electronics market has not fully recovered, the 'killer use cases' where AI genuinely forms high-frequency use are still being explored, and meanwhile high storage costs raise the hardware barrier for AI terminals.
Therefore, over the next two or three years, the consumer electronics market may find it hard to return to an era relying purely on shipment growth.
At one end, high-end phones, foldables and new products with significant process differences carry the task of raising per-device value; at the other, new-form-factor terminals such as AI glasses and ear-worn devices search for new growth; and in the mid-to-low-end market, vendors' cost control, supply chain management and model trade-offs are put to the test.
For phone vendors, new product launches remain essential, but 'what you launch' is becoming more critical than 'how much you launch'.
From foldables to on-device AI, from professional imaging to big batteries, to new-form-factor devices such as AI glasses, consumer electronics is moving from past parameter competition into a new stage where process, ecosystem and cost together determine product value.
And the change in memory prices is perhaps the most realistic pressure test of this industry transition.
How often do you change your phone? Feel free to leave a comment, share and repost. (Produced by Thinking Finance) ■
Consumer Electronics