標題: [中央處理器] 量子處理器 [打印本頁]

---

作者: salesman    時間: 2017-7-26 17:26     標題: 量子處理器

本帖最後由 salesman 於 2017-7-26 17:39 編輯

將來真係推出的話, 現時的多核心架構係咪多餘?
https://zh.wikipedia.org/wiki/%E ... 1%E7%AE%97%E6%9C%BA
基本概念

---

作者: P14    時間: 2017-7-26 17:30

提示: 作者被禁止或刪除 內容自動屏蔽

---

作者: crackong    時間: 2017-7-26 17:36

提示: 作者被禁止或刪除 內容自動屏蔽

---

作者: alan216hk    時間: 2017-7-26 19:27



我淨係知道暫時好難變成家用囉

---

作者: minki    時間: 2017-7-26 20:23

量子計算只可以計某O的野...
同"通用電腦"有好大分別...
就正如GPU點快你都要用CPU一樣

---

作者: bhkk00    時間: 2017-7-26 20:27



   
量子電腦應該係可以完全代替現時既電腦
而唔係得特定用途

via HKEPC Ionic Reader v1.7.0 - Android

---

作者: minki    時間: 2017-7-26 20:47


"量子計算"同"經典計算"兩樣野黎

---

作者: macis    時間: 2017-7-26 20:48

有幾快??

---

作者: bhkk00    時間: 2017-7-26 21:23


大家用唔同方式運作
但係唔=量子電腦唔可以取代傳統電腦
用其他方式達到相同結果就己經可以取代

---

作者: minki    時間: 2017-7-26 21:28


你咁講GPU都可以取代CPU...

---

作者: P14    時間: 2017-7-26 21:29

提示: 作者被禁止或刪除 內容自動屏蔽

---

作者: minki    時間: 2017-7-26 22:08


佢係某幾樣野好快...唔係咩都計到

---

作者: 性情中人@yahoo    時間: 2017-7-26 22:14




咁如果多一層軟件去計返晒佢原生唔識計既野咁咪得咯
利申: 電腦白痴

---

作者: minki    時間: 2017-7-26 22:24


如同用ARM行X86 CODE一樣

---

作者: 性情中人@yahoo    時間: 2017-7-26 23:00

回覆 14# minki


咁如果個"量子電腦"好快... 係正常電腦既幾十倍咁快呢?
咁會唔會七除八扣後都仲可以諗?

---

作者: minki    時間: 2017-7-26 23:01



咁如果個GPU好快... 係正常CPU既幾十倍咁快呢?
咁會唔會七除八扣後都仲可以諗? ...

---

作者: 鬼(?)    時間: 2017-7-27 01:54




    你宜家一部pc可以完全simulate到ps2就知可以啦

---

作者: minki    時間: 2017-7-27 08:19


都話因為PC係"通用電腦"

---

作者: GetA_Life    時間: 2017-7-27 09:01

裝到Windows先(?

---

作者: KinChungE    時間: 2017-7-27 09:43



而家既GPU, 以Flops計, 絕對有CPU既幾十倍咁快
但係仍然未足以simulate/emulate x86

---

作者: Mtube    時間: 2017-7-27 10:06

電單車與泥頭車

---

作者: 燕飛    時間: 2017-7-27 10:07

回復 20 #KinChungE

GPU 一見到Branch ....    

via HKEPC Reader for Android

---

作者: s84292    時間: 2017-7-27 10:13

本帖最後由 s84292 於 2017-7-27 10:35 編輯


唔係唔可以,係依家仲未有人諗到一個高效的算法,
而且現時量子控制技術仍然唔成熟,
其實要量子電腦運算就必需推今所有現時計算機體系的傳統概念, 以現時的原始方式運算根本就係浪費

量子計算同經典計算雖然係兩個概念,但前者要取代後者是指日可待的事情, 因為量子計算的潛力太大了
IBM同中科院研發同試運行的量子電腦就係面對傳統通用計算

不過好多野都唔洗用到量子計算的能力,所以傳統計算應該都會共存

---

作者: superguyalan    時間: 2017-7-27 14:52

呢樣有排都未得, 就好似當年成功做到有控聚變
以為搵到無限能源, 果時到依家已經過咗60年啦
依家大家只不過仲係討論啲石化能源點減排
要做到可用而穩定嘅量子態, 同可發電聚變一樣咁困難

---

作者: BW    時間: 2017-7-27 15:04

Take a look here
"Last year IBM made a five-qubit quantum processor available to developers, researchers and programmers for experimentation via its cloud portal. The company has made significant progress since then, revealing in May that it has upgraded its cloud-based quantum computer to a 16-qubit processor—and created a more tightly engineered 17-qubit processor that could be the basis for commercial systems. Both are based on the wire-loop superconducting circuits, as is Google's 20-qubit processor, which the company announced at a conference in Munich, Germany, on June 22. Alan Ho, an engineer in Google's Quantum Artificial Intelligence Lab, told the conference his company expects to achieve quantum supremacy with a 49-qubit chip by the end of this year."
https://www.scientificamerican.c ... pete-for-supremacy/

---

作者: minki    時間: 2017-7-27 15:11


我就係咁GE意思...
起碼係可預見將來都無話邊個取代邊個....

---

作者: usei    時間: 2017-7-27 17:11



咁點用浮點數計整數

---

作者: 性情中人@yahoo    時間: 2017-7-27 21:51




我唔知.... 我堅唔識
我而家部電腦仲係行緊DDR2

---

作者: 性情中人@yahoo    時間: 2017-7-27 21:53

個TOPIC好深.... 堅唔明.....

---

作者: minki    時間: 2017-7-28 16:16


都話左兩樣野黎....

---

作者: Ryuta    時間: 2017-7-29 15:11

10年前夠話用石墨希代替硅啦

---

作者: kazenorin    時間: 2017-7-30 00:21

取唔取代到依加言之尚早,
但係量子電腦同"普通電腦"既分別, 絕對唔係 ARM vs x86 vs PowerPC 咁簡單...

分別比起 GPU 同 CPU 更大
似乎好多人重係唔明 GPU 同 CPU 分別其實有幾大

---

作者: lubbersteven    時間: 2017-7-30 00:35

有人就系仲意用電單車比泥頭車, 又要電單車甘靈活, 又要有泥頭車ge 負重.

---

作者: minki    時間: 2017-7-30 23:29

https://www.aps.org/publications/apsnews/201705/quantum.cfm
"At 49 qubits, this quantum computer will still be far from broadly useful. To make a "universal" quantum computer — something that might supplant your office computer — "you need millions of qubits," Martinis says."
"即使有了 49 個量子位元，這一台量子電腦仍然離廣泛的使用有相當距離。要製造一台能取代你研究室所用電腦的「通用的」量子電腦，馬丁尼斯說道：「你需要百萬個量子位元！」"

---

作者: Ray少    時間: 2017-7-31 01:20

軍用野整出黎都無可能話你知，係量子電腦面前現今所有加密系統都是垃圾

via HKEPC Reader for Android

---

作者: gakko    時間: 2017-7-31 01:34

唔係高科技就可以取代
分分鐘將依家指令集轉做量子指令集都要幾廿年

---

作者: 性情中人@yahoo    時間: 2017-7-31 22:25




所以我堅唔明...
但追完個PO, 即係唔係唔得既....
但量子電腦仲進步多二三十萬倍就得喇

---

作者: BW    時間: 2017-7-31 22:50

Practical Quantum Computers
Advances at Google, Intel, and several research groups indicate that computers with previously unimaginable power are finally within reach.

Availability: 4-5 years

"All the academic and corporate quantum researchers I spoke with agreed that somewhere between 30 and 100 qubits—particularly qubits stable enough to perform a wide range of computations for longer durations—is where quantum computers start to have commercial value. And as soon as two to five years from now, such systems are likely to be for sale. Eventually, expect 100,000-qubit systems, which will disrupt the materials, chemistry, and drug industries by making accurate molecular-scale models possible for the discovery of new materials and drugs. And a million-physical-qubit system, whose general computing applications are still difficult to even fathom? It's conceivable, says Neven, 「on the inside of 10 years.」"
https://www.technologyreview.com ... -quantum-computers/

---

作者: minki    時間: 2017-7-31 23:01


係比較容易攻擊RSA...
但對AES之流完全冇影響....

你要一樣野勁二三十萬倍先取代都一樣野之下都認為兩者一樣本身唔覺有問題?

---

作者: 性情中人@yahoo    時間: 2017-7-31 23:37




即係話理論上得
實際上唔會有人令佢發生
除非有突破性發展

---

作者: minki    時間: 2017-7-31 23:46


就正如SMART PHONE仍然未取代Feature phone一樣....
大家都係phone...不過好唔同.....

---

作者: 性情中人@yahoo    時間: 2017-8-1 00:19

回覆 41# minki


咁我就明喇
感謝你咁有心機

---

作者: minki    時間: 2017-8-1 17:21


太好啦

---

作者: erpanadol    時間: 2017-8-1 22:14

本帖最後由 erpanadol 於 2017-8-1 22:16 編輯

幾廿年? 淨系由量子位提取信息既算法就用左20年，E家大部分量子算法都可以追溯到70年代，再加上專利門檻，民用量子電腦比核聚變同星際移民更遙遠

---

作者: BW    時間: 2017-8-1 23:36

Quantum Computing Is Real - everyone can do it now:-

"quantum computers are extremely difficult to program today because they require highly specialized knowledge..."But we need more smart people thinking about applications, and another set thinking about software tools."...

Qbsolv is designed to help developers program D-Wave machines without needing a background in quantum physics...today the company released Qbsolv as open source, meaning anyone will be able to freely share and modify the software...

"Not everyone in the computer science community realizes the potential impact of quantum computing...Qbsolv offers a tool that can make this impact graphically visible, by getting researchers and practitioners involved in charting the future directions of quantum computing developments."

Of course to actually run the software you create with these tools, you'll need access to one of the very few existing D-Wave machines. In the meantime, you can download a D-Wave simulator that will let you test the software on your own computer. Obviously this won't be the same as running it on a piece of hardware that uses real quantum particles, but it's a start."
https://www.wired.com/2017/01/d- ... -quantum-computing/

---

作者: minki    時間: 2017-8-1 23:52


GOOGLE同IBM果個唔算民用?

---

作者: 16477    時間: 2017-8-2 14:34

咁 D-Wave  果部呢。