Name: 出現新CPU製造商了！ (There’s a new CPU maker.)
Uploaded: 2026-03-28T21:36:31.000Z
Duration: 7 min 18 s
Description: 【看影片學英語】數萬部 YouTube 影片，搭配英漢字典即點即查，輕鬆掌握單字發音與用法，長久累積看電影不必再看字幕。

情態副詞

If you're watching this right now, you're probably using at least one ARM CPU to do it.

如果你現在正在看這段影片，你很有可能至少就用到了一顆 ARM CPU 在做這件事。

轉承詞

Or well, not an ARM CPU, because ARM doesn't actually make CPUs.

嚴格來說，不是 ARM CPU，因為 ARM 本身其實沒有在製造 CPU。

That's the big news that they sponsored us down here to their ARM Everywhere event to announce.

這就是他們贊助我們來到 ARM Everywhere 活動要宣佈的大新聞。

Behind me and in my hand is the ARM AGI CPU, built for performance, scale, and, as always, efficiency.

我身後和我手裡拿著的就是 ARM AGI CPU，它專為效能、擴展性，以及一如既往的效率而打造。

Up to 136 ARM NeoVerse v3 cores with 2 megabytes of level 2 cache each, built on TSMC's 3-nanometer process node,

高達 136 個 ARM NeoVerse v3 核心，每個核心有 2MB 的 L2 快取，採用 TSMC 的 3 奈米製程節點打造，

and it can run at up to 3.6 gigahertz, which, right out of the gate, raises some questions, doesn't it?

而且它的時脈速度最高可達 3.6 吉赫茲，這點一開始就讓人產生一些疑問，對吧？

Can it, like, dynamically boost a single core way higher or something?

它有沒有辦法動態提升單一核心的時脈速度到更高，或是做些什麼呢？

And according to ARM, that's actually a key feature, not a bug.

而且根據 ARM 的說法，這其實是個特色，不是 bug。

By issuing SMT multi-threading and the highly variable power consumption that's associated with constantly fluctuating clock speeds, not to mention designing a 12-channel DDR5 memory controller that can feed every individual core with a consistent 6 gigabytes per second of bandwidth, ARM is ensuring that every core in this CPU will perform its best at all times and keep power consumption more consistent, which will allow data centers to design to how much power their racks will consistently consume rather than having to build in a buffer for how much they might consume at peak.

透過發布 SMT 多執行緒以及與時脈速度不斷波動相關的高度可變功耗，更不用說設計一個 12 通道的 DDR5 記憶體控制器，每秒能為每個獨立核心提供 6GB 的穩定頻寬，ARM 確保這顆 CPU 中的每個核心都能隨時發揮最佳效能，並保持功耗更穩定，這將讓資料中心能夠根據機架的穩定功耗來設計，而不是必須為尖峰功耗預留緩衝。

And that's huge, considering that cooling, and

而且這點非常重要，考量到在一個快速擴展資料中心基礎設施的世界裡，散熱，

especially power, are just about the hottest commodities in a world that is rapidly scaling data center infrastructure.

尤其是電力，幾乎是最熱門的資源了。

Each AGI CPU has 96 lanes of PCI Express Gen 6, with support for CXL 3.0 for deploying massive shared memory pools over PCIe,

每個 AGI CPU 都有 96 個 PCI Express Gen 6 通道，支援 CXL 3.0，用於透過 PCIe 部署大規模的共享記憶體池，

and ARM showed off node designs with their hardware partners that deployed up to two of these CPUs on a single motherboard.

ARM 也和他們的硬體夥伴展示了節點設計，在單一主機板上部署了兩顆這樣的 CPU。

Super cool, but not exactly world-changing.

超酷的，但還稱不上是劃時代的。

To see the vision that led ARM to spend the last few years bringing this to life, you gotta zoom out and look beyond the individual node to the rack level.

要了解 ARM 花了幾年時間實現這個願景，你得拉遠一點，跳脫單一節點，看看機架層級。

這個機架包含 32 個節點 1P 伺服器。

So for those keeping count at home, that's 8,160 CPU cores.

所以在家計算的各位，這總共有 8,160 個 CPU 核心。

I mean, dense CPU racks are already a thing.

我的意思是，高密度的 CPU 機架本來就已經存在了。

This sick error message hoodie is now available from lttstore.com.

這件超酷的錯誤訊息連帽衫現在可以在 lttstore.com 買到。

I mean, it is, but that's not the big reveal.

我的意思是，確實有，但那不是重頭戲。

The big reveal is that everything that I just told you fits in a standard OCP 36 kilowatt air-cooled rack.

重頭戲是，我剛剛告訴你們的一切，都裝在一個標準的 OCP 36 千瓦風冷機架裡。

Each AGI CPU draws just 300 watts, a significant reduction compared to flagship x86 CPUs.

每個 ARM AGI CPU 只消耗 300 瓦，相較於旗艦級 x86 CPU，功耗大幅降低。

So when you throw liquid cooling at them, the numbers get, frankly, kind of ridiculous.

所以當你為它們加上液體散熱時，數字會變得，老實說，有點驚人。

In an OCP 200 kilowatt rack, ARM figures they can pack 42 8-node 1P systems for a grand total of 45,696 cores and over a petabyte of RAM, all while consuming only about half of that total available power budget.

在一個 OCP 200 千瓦的機架裡，ARM 估計可以塞進 42 個 8 節點的 1P 系統，總共會有 45,696 個核心和超過一個 PB 的 RAM，而且功耗只用了總可用功率預算的一半左右。

They are pegging the bottom line performance per watt in the neighborhood of double compared to x86.

他們預估每瓦的底線效能大約是 x86 的兩倍。

And this is largely thanks to carrying less legacy craft, but

這很大程度上要歸功於減少了舊有技術的包袱，但

also thanks to architectural choices like using fewer chiplets to keep memory latency down, along with ARM's traditional strength in instructions per clock, and

同時也歸功於像使用更少的 chiplet 來降低記憶體延遲的架構選擇，以及 ARM 在每時脈週期指令數上的傳統優勢，還有

taking just a no-silicon wasted approach to their design.

採取了不浪費任何矽材料的設計方法。

With the cost and scarcity of power, that's a number that is going to perk up a lot of ears.

考量到電力成本和稀缺性，這個數字肯定會讓很多人眼睛一亮。

Everybody knows that CPUs aren't good at AI compared to GPUs or application-specific neural processors.

大家都知道，跟 GPU 或專用神經處理器相比，CPU 在 AI 方面表現不佳。

所以，呃，這品牌定位是怎麼回事？

While GPUs and neural accelerators get all the attention, CPUs are still chugging along in the background coordinating tasks, with ARM estimating that a typical deployment today is going to have about 30 million cores per gigawatt.

雖然 GPU 和神經加速器備受矚目，但 CPU 仍在背景默默協調任務，ARM 估計現今典型部署每吉赫茲將有約 3000 萬個核心。

That's with humans handling most of the token requests.

這是在人類處理大部分 token 要求的情況下。

AI agents push requests much faster and, um, don't sleep, meaning that your expensive AI accelerators can end up sitting around because the CPU coordinators can't keep up with all of those requests.

AI 代理的請求速度更快，而且，嗯，不會休息，這意味著你昂貴的 AI 加速器可能會閒置，因為 CPU 協調器無法跟上所有這些要求。

So ARM figures that that 30 million cores per gigawatt number could go up a lot in the head node, next to the accelerator rack, as high as about four times as many.

所以 ARM 認為，在加速器機架旁邊的頭部節點，每吉赫茲 3000 萬個核心這個數字可能會大幅增加，高達四倍。

When these are doing all the actual AI work, nobody's going to want to spend more power budget on all of those CPUs.

當這些在執行所有實際的 AI 工作時，沒有人會想在所有這些 CPU 上花費更多電力預算。

Well, that's where ARM comes in, with their famously power-efficient designs.

這就是 ARM 的用武之地，憑藉其以省電效率聞名的設計。

Let's go to Nick from the lab to see this thing in action.

我們請實驗室的 Nick 來實際操作。

Many of the demos were focused on the ease of porting software to ARM and the support they're building for developers, which makes a lot of sense but isn't very visual, so let's check out this one instead, where they're encoding a 1080p video from H.264 to H.65 while running computer vision at the same time on the same CPU.

許多演示都著重於軟體移植到 ARM 的便利性以及他們為開發者建立的支援，這很有道理但不太直觀，所以讓我們來看看這個，他們正在同一顆 CPU 上同時進行 H.264 編碼成 H.65 的 1080p 影片，同時執行電腦視覺。

Let's go take a look at the man behind the curtain.

ARM actually had the stones to do it live, bringing an actual server running the actual hardware here to the show floor.

ARM 真的有種把它做成現場直播，把實際的伺服器和實際的硬體搬到展場來。

Doesn't all of this put ARM kind of in direct competition with their own customers?

這是不是讓 ARM 和他們的客戶直接競爭了？

You know, the ones who license their IP and their compute subsystems?

就是那些授權 ARM IP 和運算子系統的客戶嘛？

The guys who got them where they are today?

也是那些讓 ARM 有今天的人嘛？

But from ARM's perspective, this is actually something that many of their customers were asking for.

但從 ARM 的角度來看，這其實是他們很多客戶一直要求他們做的。

Expanding on that, ARM laid out how their roadmap and their policies account for how all three of their business models are going to go forward,

接著，ARM 說明瞭他們的藍圖和政策是如何規劃他們三種商業模式的未來發展的，

and they're positioning this as a choice between IP licensing, compute subsystems licensing, and

而且他們把這個定位成一個選擇，你可以選擇授權 IP、運算子系統授權，

physical CPUs, or, hey, why not some combination of all three?

或是實體的 CPU，或者，嘿，全部三種組合也行！

They'll gladly take your money any way you want to give it to them.

他們很樂意用任何你想要的方式收你的錢。

And it seems like that's the plan for the long haul.

而且看起來這就是長期的計畫了。

In a move that I don't think I've ever seen before, ARM stood up on stage and said the quiet part out loud.

ARM 做了一個我認為前所未見的舉動，他們站上舞臺，把心裡想的直接說出來。

The best is yet to come with our second CPU due next year.

最好的還在後頭，我們的第二顆 CPU 明年就會推出。

Like, obviously, given the timelines of silicon development,

就像，很明顯，考量到矽開發的時間表，

but you almost never hear that from a company who probably wants you to buy the hardware they have today, from partners like, you know, for example, Supermicro.

但你幾乎從來沒聽過一家公司這樣說，他們通常是希望你買他們今天就有的硬體，像是跟合作夥伴，你知道的，例如 Supermicro。

If you guys enjoyed this video, you might enjoy the one that we did at CES, also in partnership with ARM, highlighting some of the unexpected places that you can find ARM technology.

如果你們喜歡這支影片，你們可能會喜歡我們在 CES 做的另一支影片，也是跟 ARM 合作，介紹 ARM 技術出現在一些意想不到的地方。