tensorbook深度学习笔记本电脑

Lambda is a relatively new company in the hardware space, being founded in 2012, but they have already produced some great pieces of hardware. In this article I will review Tensorbook, a deep learning laptop.

Lambda是一家在硬件领域相对较新的公司，成立于2012年，但他们已经生产了一些很棒的硬件。 在本文中，我将回顾深度学习笔记本电脑Tensorbook 。

Tensorbook功能 (Tensorbook Features)

First of all, let’s start with the actual parameters of Tensorbook. From the machine learning perspective, crucial things are 8GB GPU, 8 cores processor, memory of 64 GB and 2TB storage. Tensorbook is pretty light relative to its size and capabilities.

首先，让我们从Tensorbook的实际参数开始 。 从机器学习的角度来看，关键的事情是8GB GPU，8核处理器，64 GB内存和2TB存储。 Tensorbook相对于其大小和功能而言相当轻巧。

Features:

特征：

• GPU: RTX 2080 Super Max-Q, 8 GB GDDR6, 3072 CUDA cores, 1080 MHz base clock
  GPU：RTX 2080 Super Max-Q，8 GB GDDR6、3072 CUDA内核，1080 MHz基本时钟
• Processor: 8 cores, 2.30 GHz, Intel Core i7–10875H with 16 threads, 5.10 GHz turbo, and 16 MB cache
  处理器：8核，2.30 GHz，具有16个线程的Intel Core i7-10875H，5.10 GHz Turbo和16 MB缓存
• Memory: 64 GB, of 2666 MHz DDR4 SO-DIMM
  内存：64 GB，2666 MHz DDR4 SO-DIMM
• Storage: 2 TB; Samsung 970 EVO with 1 TB, 3,500 MB/s seq. read, 2,500 MB/s seq. write; Samsung 860 EVO with 1 TB, 550 MB/s seq. read, 520 MB/s seq. write
  储存空间：2 TB； 1 TB，3,500 MB / s seq的Samsung 970 EVO 读取，2,500 MB / s seq。 写; 三星860 EVO，1 TB，550 MB / s seq。 读取，520 MB / s seq。 写
• Size & weight: Width: 14.07" (357 mm), Height: 0.78" (20 mm), Depth: 9.37" (238 mm), Weight: 4.39 lbs (2 kgs)
  尺寸和重量：宽度：14.07英寸(357毫米)，高度：0.78英寸(20毫米)，深度：9.37英寸(238毫米)，重量：4.39磅(2千克)
• Display: 15.6", 1920x1080 (Full HD), 144 Hz, Matte, 72% NTSC
  显示屏：15.6英寸，1920x1080(全高清)，144 Hz，无光泽，72％NTSC
• Audio ports: 1x 3.5mm 2-in-1 audio jack (headphone + microphone), 1x 3.5mm 2-in-1 audio jack (microphone + S/PDIF optical)
  音频端口：1个3.5毫米2合1音频插Kong(耳机+麦克风)，1个3.5毫米2合1音频插Kong(麦克风+ S / PDIF光纤)
• Video ports: 1x HDMI with HDCP, 1x Mini DisplayPort 1.4, 1x Thunderbolt 3
  视频端口：1个带HDCP的HDMI，1个Mini DisplayPort 1.4、1个Thunderbolt 3
• USB ports: 2x USB 3.2 Gen 1 Type A, 1x powered (AC/DC) USB 3.2 Gen 1 Type A
  USB端口：2个USB 3.2 Gen 1 Type A型USB端口，1个有源(AC / DC)USB 3.2 Gen 1 Type A型端口
• Networking: Ethernet: 1x 10/100/1000 Mbps RJ45 port, Wireless: Intel Dual Band WiFi 6 AX, Bluetooth: Supported
  网络：以太网：1个10/100/1000 Mbps RJ45端口，无线：英特尔双频WiFi 6 AX，蓝牙：支持

Tensorbook和Lambda堆栈 (Tensorbook and Lambda stack)

Lambda Stack provides an easy way to install popular Machine Learning frameworks. With Lambda Stack, you can use apt / aptitude to install TensorFlow, Keras, PyTorch, Caffe, Caffe 2, Theano, CUDA, cuDNN, and NVIDIA GPU drivers. If a new version of any framework is released, Lambda Stack manages the upgrade. You’ll never run into issues with your NVIDIA drivers again.

Lambda Stack提供了一种简单的方法来安装流行的机器学习框架。 使用Lambda Stack，您可以使用apt / aptitude安装TensorFlow，Keras，PyTorch，Caffe，Caffe 2，Theano，CUDA，cuDNN和NVIDIA GPU驱动程序。 如果发布了任何框架的新版本，则Lambda Stack将管理升级。 您再也不会遇到NVIDIA驱动程序问题。

Great thing about TensorBook is that it comes with the Lambda stack pre-installed so you can update and install it with one command:

TensorBook的伟大之处在于它预装了Lambda堆栈，因此您可以使用以下命令进行更新和安装：

sudo apt-get install lambda-stack

You can read more about Lambda Stack here.

您可以在此处阅读有关Lambda Stack的更多信息 。

Tensorbook的设置，用法和第一印象 (Tensorbook set up, usage and first impressions)

Tensorbook comes nicely packaged in the following box:

Tensorbook很好地包装在以下框中：

Lambda Tensorbook Lambda Tensorbook

After you take it out of the box, you can start using it right away.

开箱即用后，您可以立即开始使用它。

Lambda Tensorbook Lambda Tensorbook

I was able to set up my first machine learning experiment within 30 minutes of taking the Tensorbook out of the box. And what a pleasure it was!

开箱即用后的30分钟内，我就能设置我的第一个机器学习实验。 真是太高兴了！

Tensorbook’s keyboard is great to use. Writing really comes easy and it feels natural. The touchpad is also working very well and you won’t need a mouse (unless you’re going to run action games).

Tensorbook的键盘非常好用。 写作真的很容易，而且感觉自然。 触摸板也可以很好地工作，并且您不需要鼠标(除非您要运行动作游戏)。

The screen gives you a crystal clear display. The only lacking parts are internal speakers and the laptop’s camera. Fortunately, headphones solve the first problem, while the internal camera can be replaced by a new webcam — thanks to the great number of ports you have. Nevertheless, listening to music nor recording videos are not the primary functions of the laptop.

屏幕显示清晰。 唯一缺少的部件是内置扬声器和笔记本电脑的摄像头。 幸运的是，耳机解决了第一个问题，而内置摄像头可以替换为新的网络摄像头-这要归功于您拥有大量的端口。 但是，听音乐或录制视频不是笔记本电脑的主要功能。

Tensorbook和机器学习 (Tensorbook and Machine Learning)

What’s crucial is that it simply works fast and lets you train machine learning models locally on your own laptop. Unless you run large models, Tensorbook will be often able to replace cloud computing for you, so instead of paying monthly to a cloud provider, you have to just pay upfront for the laptop. The prices start at $3,300 with Ubuntu and $3,500 with Windows pre-installed.

至关重要的是，它可以快速运行，并允许您在自己的笔记本电脑上本地训练机器学习模型。 除非您运行大型模型，否则Tensorbook通常将能够为您替换云计算，因此您不必每月向云提供商支付费用，而只需为笔记本电脑支付前期费用。 Ubuntu的起价为3,300美元，预装Windows的起价为3500美元。

For the comparison with cloud providers, AWS p3.2xlarge and p2.xlarge instances are the most similar and they cost respectively $3/hour and $1/hour. So if you need to run them constantly, it’s much cheaper to actually pay upfront for Tensorbook.

为了与云提供商进行比较，AWS p3.2xlarge和p2.xlarge实例最为相似，它们分别花费3美元/小时和1美元/小时。 因此，如果您需要不断运行它们，那么实际支付Tensorbook的费用要便宜得多。

All in all, if you’re looking for a mobile deep learning workstation, Tensorbook is definitely a great choice.

总而言之，如果您正在寻找移动深度学习工作站， Tensorbook 绝对是一个不错的选择。

If you want to have a look at how the laptop looks and feels like, watch this Tensorbook Video Review:

如果您想看看笔记本电脑的外观和感觉，请观看此Tensorbook Video Review ：

演示地址

Tensorbook, a deep learning laptop review Tensorbook，深度学习笔记本电脑评测

Tensorbook is a great laptop for deep learning!

Tensorbook是一款出色的深度学习笔记本电脑！

Lambda Tensorbook Lambda Tensorbook

翻译自: https://medium.com/data-science-rush/tensorbook-a-deep-learning-laptop-57e523046d5c