跳到主要內容

黃金級贊助商 - Skymizer 帶來 ONNC 通用的開放神經網路編譯器

「AI 創新需要開源生態系,而 ONNX 確保了 frameworks 之間的互通性,ONNC 的目標是將所有的 DLA ASIC 快速、簡單地與 ONNX 連結起來,確保所有的 DLA ASIC 都可以在 ONNX 上執行。」

由於 AI 的應用層面越來越廣,數百種新 AI 晶片將在不久的未來大量出現,目前市場上卻沒有一個能完整支援各家 DLA 的開源編譯器。根據統計,在 2018 年會有超過 34 家 IC 與 IP 廠商提供各式各樣的 AI 晶片與 deep learning accelerator (DLA) ASICs,因此急需一個開源的編譯器來支援各種不同的 AI 晶片。

成立於 2013 年的 Skymizer 一直都專注在 compiler 和 machine learning 領域,看到這個趨勢之後,建立了基於 ONNX 的編譯器 Open Nerual Network Compiler – ONNC,將所有的 AI 晶片與 ONNX 連結起來。

ONNX 是開放類神經網路交換格式,目前常見的 AI framework 如Caffe2、PyTorch…等等各有各的支持者,格式間互通性差,因此訓練出來的深度學習模型無法套用到別的 framework 上,但有了 ONNX,開發者可在不同格式之間輕鬆轉換。而透過 ONNX 支援多平台的特性,ONNC 可以支援各種不同的 AI frameworks,如 Caffe、Caffe2 與 PyTorch,幫助 DLA ASIC 廠商在短時間內就可以支援各種 AI frameworks,提升性能並縮短開發時間。  






想要更了解 ONNC,歡迎到 https://onnc.ai 追蹤最新的動態,
目前 ONNC 已在 GitHub https://repo.onnc.ai 釋出,而所有開發的相關討論,
將會在 mailing list 進行。


ONNC features:
Easy Backend Integration
ONNC is integrated with the LLVM bitcode runtime and compiler. If a DLA already supports the LLVM compiler, it can be connected to ONNC seamlessly. This helps most CPUs, GPUs, and DSPs ported to ONNC in a very short time. On the other hand, if a DLA has unique computation features and is not compatible to LLVM, ONNC also provides a modular framework to speed up the compiler development. DLA vendors can quickly customize an ONNC backend from a so called “vanilla” backend, which already provides some necessary optimization algorithms.


Reusable Compiler Optimizations
Two of ONNC’s contributions are dividing the AI compilation into several clear phases and giving the corresponding APIs for algorithm development.
There are five phases carefully defined, each of which is focused on a particular compiler problem: IR building, partitioning, scheduling, allocation, and code emitting.
ONNC also provides a series of optimization algorithms ready for use. They are general and reusable, including tensor selection, tensor liveness analysis, linear scan local memory allocation, etc. ONNC’s pass manager is flexible and similar to LLVM’s. AI researchers and engineers who are familiar with LLVM can intuitively contribute their general or target-specific optimization algorithms to ONNC.


Status and Future Work
Project ONNC is not mature yet.
Still, we have a lot to do and need the community together to make it better. In this preview release, we carefully designed the software architecture so as to simplify the future development and
maintenance.
The items we have done include:



  • Clear compilation phases and the corresponding APIs.
  • Well-defined and extensible intermediate representation (IR) of target-specific instructions.
  • A pass manager that supports automatic scheduling according to the dependency claimed by each pass.
  • A Sophon backend that supports BITMAIN AI ASIC.
  • Shim library such as ADT, Diagnostics, JSON, etc.

Project ONNC follows the “release early, release often” principle,
so the next release should come soon by the end of August.
We plan to finish the following items.

  • An x86 backend that enables the execution of AI models on an x86 machine. That backend should support both JIT and interpreter way of execution.
  • Improvements of memory allocation based on liveness analysis.

JOIN US
To keep up with the latest development or make suggestions, please join the ONNC mailing list. Also, please head over to the official website https://onnc.ai/ and GitHub https://repo.onnc.ai for more information.








留言

這個網誌中的熱門文章

2016 Unconference 議程公告!

嗨、各位親朋好友們
今年的 Unconference 名單出爐啦! 由於每一個 Talk 都相當精彩,因此我們想了辦法讓所有投稿的人都有舞台可以分享。 Unconference 演講場地為中研院活動中心的平面會議室。 對於其中的講題有興趣的朋友們請不要錯過囉 :-D
8/21 Unconference 活動中心的平面會議室時間 講題 講者 10:45
|
11:05 用開放原始碼技術翻轉傳統產業IT架構 Derek Hsu11:10
|
11:30 會自動飛行的"神奇寶貝"and飛行技巧 gavin 11:45
|
12:05 ProbeDroid: Crafting Your Own Dynamic Instrument Tool on Android for App Behavior Exploration | 沈宗賢 (AndyShen)12:10
|
12:30 蒙地卡羅模擬與志願運算 趙元13:30
|
13:50 Runtime PM for CPU idling in Linux kernelfreedom Koan-Sin Tan 13:55
|
14:15 Fuzzing and Mozilla GARY KWONG gkw14:30
|
14:50 無良房東一度電收你 5.5 元,你難道不生氣嗎?——從自幹智慧電表講起 Felix and Timo14:55
|
15:15 COSCUP + AppDevKit Jeff Lin C 備註:議程時間可能會調動。

COSCUP 全議程軌資訊! Information about the tracks this year (submit your talk now!)

Here, we list all the tracks (Community Rooms) this year with their detail, so you can check them in one page -- okay, no excuses now, submit your proposal to COSCUP 2018!我們把今年所有的議程軌徵稿資訊都放在這裡以方便您閱讀,那麼看完以後快快來投稿吧GNOME.Asia SummitAndroid 開源專案Arch Linux Taiwan & ArchersBlockchainData Science PlaygroundDrupal 的一生、網站的醫生FOSS Compliance - Complex Made SimpleInternet Governance and Digital Privacy / 網路治理與數位隱私Julia LanguageKernel & Coding Serfs & SystemLet’s read the source code / 帶您讀源碼Make Technology Policies Open / 科技社群如何參與科技政策規劃? 你真的需要搞一個政黨出來嗎?Open Source DatabaseOpen GISOpen Source Farm InnovationOpen Source Software Development & ManagementRubySDN x Cloud NativeAD 廣告時間Click here to vote for your favorite community rooms!
議程人氣投稿已經開跑囉~快來幫你喜歡的議程軌投票吧 點我! GNOME.Asia SummitGNOME.Asia Summit 2018 invites proposals for presentations at the conference.GNOME.Asia Summit is the featured annual GNOME conference in Asia. It focuses primarily on the GNOME desktop, but also covers applica…

Full programme of COSCUP x GNOME.Asia x openSUSE.Asia 2018 is here!

The following is an overview of this year’s programme.While the official website is still under construction (It's alive!), you can see the time slots, speaker and room assignments for each topic through this Spreadsheet. The full programme, include abstract and other information, will be announced on coscup.org after the site is done. As usual: this agenda may change at any time. Please stay tuned.Any question? You can reach the program team at program@coscup.org. Again, thanks to all the participating communities and contributors!Check the full schedule!