OnePlus 15が予告:巨大7,300mAhバッテリー&120W急速充電、10月27日発売&新色パープル登場!

10月 27, 2025
OnePlus 15 teased: Mammoth 7,300mAh battery & 120W charging, Oct 27 launch & new purple color
  • 巨大なグレイシャーバッテリー: OnePlusは、次期OnePlus 15に7,300 mAhバッテリーを搭載することを確認しました[1]。これはOnePlus 13の6,000 mAhセルからの大幅な増加です。ほとんどのフラッグシップ機よりもはるかに大きく(Googleの5,200 mAh Pixel 10 Pro XLより約40%大きく、5,000 mAh Galaxy S25 Ultraより46%大きい)[2]。OnePlusはこれを「グレイシャーバッテリー」と名付けており、新しいシリコンカーボンセル技術を示唆しています。この容量により、アナリストは「2日間フルで使える可能性がある」と予想しています[3]
  • 超高速充電: 大容量バッテリーに加え、OnePlus 15は120W有線充電(「スーパーフラッシュチャージ」としてブランド化)および50Wワイヤレス充電に対応します[4][5]。これはOnePlus 13の100W有線充電を上回りますが、米国モデルは安全のため約80Wに制限される可能性があります[6]。OnePlusはまた、アクセサリーの安全な充電のためのバイパス充電機能も確認しています。あるティーザーでは、OnePlusはこの組み合わせを「究極のパワーとバッテリー寿命のコンビネーション」として自慢し、ゲーマーがより長くプレイできるよう設計されていると述べています[7]
  • ハイエンド仕様: OnePlus 15は内部的にも真のフラッグシップとなります。Qualcommの新しいSnapdragon 8 Elite Gen 5チップセットを搭載し、6.78インチOLEDディスプレイ165Hzリフレッシュレート[8]が組み合わされています。この超高リフレッシュレートの画面(OnePlus 13の120Hzから向上)は、ゲーミング向けのアップグレードです。OnePlusはバッテリーも「超高フレームレートゲーミング」体験を強化するためのものだと述べています[9]。このスマートフォンはOxygenOS 16(Android 16)をプリインストールして発売されます[10]
  • 発売スケジュール: OnePlusは公式に、OnePlus 15が中国で2025年10月27日午後7時(北京時間、PT午前3時/ET午前8時)に発表されるとアナウンスしました[11]。この中国での発表(新しいAce 6モデルと同時)はすべての仕様を明らかにしますが、グローバル発売は後日となります。業界のリークによると、国際版のリリースは約2週間後、11月中旬(11月13日頃)になる見込みです[12]。(参考までに、前回のOnePlusフラッグシップは2024年10月に中国で発売され、世界発売は2025年1月まで待たれました。)
  • デザイン&カラー: OnePlus 15は、サンドストーン風のスタイルを継承し、わずか1.15 mmのベゼル[13]とシンプルなカメラアレイを備えています。少なくとも3つの仕上げで発売されます:シグネチャーのSand(または「Sand Storm/Dune」)とAbsolute Black、さらに新しいパープルのアクセントカラーです。TechRadarはOnePlusの投稿でこれを「Misty Purple」と呼んでいるのを確認しましたが、他のメディアでは「Fog Purple」「Mist Purple」と訳しています[14][15]。(注:OnePlus 14は存在しません ― 迷信的な理由でその番号は飛ばされています[16]。)

OnePlusは、新しいバッテリーと充電機能がゲーマーやパワーユーザー向けであることを強調しています。Android Centralによると、OnePlusのマーケティングは7,300mAhセルを「Glacier Battery」技術の一部として、より長いゲームセッションを可能にするとアピールしています。同社はこのバッテリーパックが「『超高フレームレートゲーミング』体験を強化するために設計された」と述べています[17]。WeiboのティーザーでOnePlusは、OnePlus 15を7,300mAhセルと「120W Super Flash Charge」を搭載し、「超高フレームレートゲーミング向けに設計された」と説明しています[18]

要するに、OnePlus 15は新たな耐久性のベンチマークを打ち立てることを目指しています。Android AuthorityのHadlee Simonsは、7,300mAhへのジャンプをOnePlus 13の6,000mAhバッテリーに対する「大幅なアップグレード」、さらに多くの競合機種に対しては「巨大なアップグレード」と評しています[19]。120Wの有線急速充電と組み合わせることで、1回の充電で数日間の混合使用が可能になるかもしれません。もちろん、Android Authorityが指摘するように、米国モデルはこれまで約80Wの有線充電に制限されてきたため[20]、アメリカの購入者はやや遅い速度になる可能性があります。それでも、80Wでも充電速度はどの基準でも非常に高速です。

10月27日の中国での発売は大きな注目を集めています。当日午後7時(北京時間)からOnePlusはウェブキャストを予定しており、そこでOnePlus 15とミッドレンジのOnePlus Ace 6の両方の(中国市場向けの)詳細な仕様と価格が発表されます[21]。すでにいくつかの詳細は判明しています。前述の通り、Snapdragon 8 Elite Gen 5チップと165Hzの6.78インチOLEDディスプレイ[22]、さらに大容量バッテリーと急速充電が搭載されます。また、同社はこのモデルでHasselbladとのカメラ提携を終了し、OnePlus独自の「DetailMax Engine」をイメージングに使用することを確認しています。初期プレビューでは、すべて50MPのトリプルカメラ構成、8K/120fpsの動画撮影対応、1.15mmの極細ベゼルが注目されています[23][24]

国際的な顧客も中国での発売後、それほど長く待つ必要はありません。OnePlusはすでに「世界の他の地域」にもこのスマートフォンを投入すると発表しており、Android Centralによると、国際版の発売は11月13日頃になるというリークもあります[25]。もしこれが事実なら、中国イベントから約2週間後(前モデルの3か月待ちよりはるかに早い)となります。グローバル価格が決まれば、OnePlus 15は今秋の他のフラッグシップ機と直接競合することになります。

まとめると、OnePlusはOnePlus 15の目玉機能としてバッテリー持ちと充電性能を積極的にアピールしています。業界関係者によれば、現行のAndroidスマートフォンで7,300mAhバッテリーと120W充電の組み合わせは他にありません。9to5Googleによれば、OnePlusはこのモンスターバッテリーとともに「今年唯一の『120Wスーパーフラッシュチャージ』」を約束しているとのことです[26]。長持ちするスマートフォンを求めるファンにとって、これは10月の発表会で最も大きなニュースとなりそうです――10月27日の公式発表直前に。

情報源: 確認された仕様と日付は、OnePlusがWeiboで発表し、Android Authority、Android Central、TechRadar、9to5Googleなどのメディアで報道されました。[27] [28] [29] [30]などです。上記の各ポイントは、これらの情報源によって確認されています。

OnePlus Nord CE5 | Battery, unleashed.

References

1. www.androidauthority.com, 2. 9to5google.com, 3. www.androidauthority.com, 4. www.androidcentral.com, 5. 9to5google.com, 6. www.androidauthority.com, 7. 9to5google.com, 8. www.techradar.com, 9. www.androidcentral.com, 10. www.livemint.com, 11. www.techradar.com, 12. www.androidcentral.com, 13. www.livemint.com, 14. www.techradar.com, 15. www.androidcentral.com, 16. www.techradar.com, 17. www.androidcentral.com, 18. 9to5google.com, 19. www.androidauthority.com, 20. www.androidauthority.com, 21. www.techradar.com, 22. www.techradar.com, 23. www.livemint.com, 24. www.livemint.com, 25. www.androidcentral.com, 26. 9to5google.com, 27. www.androidauthority.com, 28. 9to5google.com, 29. www.techradar.com, 30. www.androidcentral.com

Technology News

  • The new language of AI tech workers: Silicon Valley slang, 996ers and NPCs | The Indicator
    October 27, 2025, 3:22 AM EDT. Planet Money's The Indicator explores a fresh workplace dialect born from the AI boom. Jasmine Sun guides listeners through Silicon Valley slang-from 996ers to NPCs-and asks what this language reveals about job anxiety, power dynamics, and the broader tech industry culture. The episode profiles a labor market reshaped by ongoing AI buildout and considers how this worldview influences negotiations over roles, rewards, and the future of work. A revealing window into how the language of AI tech workers mirrors ambition and unease as the field evolves.
  • Quantum computing and AI unite to probe particle physics with LHC data
    October 27, 2025, 3:06 AM EDT. In this Physics World Weekly podcast, experts discuss how quantum computing and artificial intelligence can help physicists search for rare interactions in data from an upgraded Large Hadron Collider. Guest Javier Toledo-Marín, based at the Perimeter Institute and connected to TRIUMF, explains the team's work on a paper titled "Conditioned quantum-assisted deep generative surrogate for particle-calorimeter interactions." The episode is recorded at the Perimeter Institute in Waterloo, Canada, and highlights how gate-based quantum computing scales with support from Delft Circuits, which provides IO solutions for quantum hardware.
  • Why Rigetti Computing Stock Fell 13.9% This Week
    October 27, 2025, 2:54 AM EDT. Rigetti Computing's shares slipped as rumors swirled around its CEO Subodh K. Kulkarni selling his stake, though he still holds around 0.8% via stock options. The stock later rebounded briefly on reports the Trump administration might take an equity stake, then tumbled again after a Commerce Department denial that talks were underway with quantum companies. The week underscored the volatility of pure-play quantum plays: JPMorgan's investment enthusiasm helped earlier gains, while unverifiable reports about government involvement cooled momentum. Analysts caution that the stock is currently inflated relative to fundamentals, with a speculative backdrop for quantum names. Investors may want to see valuations normalize before piling into Rigetti or other pure-play quantum plays.
  • Polish tops complex AI tasks in multilingual LLM benchmark, study finds
    October 27, 2025, 2:52 AM EDT. An international study by Microsoft, the University of Maryland, and the University of Massachusetts Amherst benchmarked six LLMs (OpenAI, Gemini, Qwen, Llama) across 26 languages using a new multilingual long-context benchmark. The results were surprising: in the most complex tasks, Polish achieved an average accuracy of 88%, finishing first, while English ranked sixth at 83.9% and Chinese around 62.1%. The second-best was Russian, followed by French, Italian, and Spanish; the worst included Sethoto, Tamil, and Swahili. The study highlights that language families (Slavic, Romance, Germanic) and scripts (Latin, Cyrillic) influence performance, especially at 64K-128K context lengths. Researchers note that dominant pretraining data in English/Chinese doesn't always predict top performance in long contexts.
  • AI-written code accelerates development, but security bottlenecks loom, per OX Security report
    October 27, 2025, 2:50 AM EDT. OX Security's Army of Juniors: The AI Code Security Crisis finds AI-generated code often looks clean but hides flaws that can become systemic risks. The study analyzed 300+ repositories, including 50 using AI tools like GitHub Copilot, Cursor, or Claude. It found AI code is not more vulnerable per line; the issue is velocity. Speed erodes code review, debugging, and oversight, letting vulnerable code reach production in days instead of months. Security teams remain overwhelmed, handling hundreds of thousands of alerts, and AI-assisted coding worsens the bottlenecks. The report lists ten anti-patterns common in AI code-from excessive comments, avoidance of refactors, over-specification, to by-the-book thinking-that breed technical debt and monolithic architectures. The danger lies in user ignorance about security, not AI's raw skill.

Don't Miss