Reports from the Heart of the Machine
Editors: Chen Ping, Du Wei
Good computer, but a little expensive.
Razer Inc is a Singapore-based game peripherals company, also known as the "Green Light Factory", with two headquarters in Singapore and San Diego, usa, and its products are mostly named after carnivores.
Razer is known for its gaming PCs and peripherals, but it's no secret that the company is also involved in other areas. On Tuesday, Razer partnered with deep learning-focused company Lambda to launch the new Lambda TensorBook, which it bills as "the world's most powerful laptop designed for deep learning." This pc allows users to use Linux and Lambda's deep learning software.
Founded in 2012, Lambda is a provider of deep learning infrastructure that is used by thousands of businesses and organizations, including the U.S. Department of Defense and 97 percent of the top U.S. research universities. Lambda's offerings include GPU clusters, servers, workstations, and cloud instances that can train neural networks for a variety of use cases, including self-driving cars, cancer detection, and drug discovery.
The laptop has logos from Lambda and Razer.
Laptops designed for deep learning
Equipped with the Nvidia RTX 3080 Max-Q (16GB), this Tensorbook is aimed at machine learning engineers, especially those who lack laptops with independent GPUs and therefore have to share remote machine resources. It also comes with the previous generation Intel i7-11800H, which has 8 cores and a base clock speed of 2.3 GHz, which can be increased up to 4.6 GHz.
The machine also uses 64GB of DDR4-3200 RAM and 2TB of PCIe 4.0 SSD storage with a 15.6-inch screen, a resolution of 2560x1440, and a 165Hz refresh rate.
In terms of port selection, tensorbook has two Thunderbolt 4 ports, three USB-A ports (3.2 Gen 2), HDMI 2.1, a headphone jack, and an SD card reader.
While we can choose a more powerful laptop, tensorbook stands out among the many products for its excellent software package and Ubuntu Linux 20.04 LTS. Lambda also said tensorbooks also support Windows dual boot.
Tensorbook makes it easy to install PyTorch, TensorFlow, Caffee, and Caffee 2 deep learning frameworks through Lambda's "single-wire and managed upgrade path," as well as GPU-focused applications CUDA, cuDNN, and their development tools.
In a statement, Lambda co-founder and CEO Stephen Balaban said: "When you can't connect to a remote server via SSH, you don't have any local data or code, and it's even difficult to present your model to colleagues, Tensorbook solves this problem, it comes pre-installed with PyTorch and TensorFlow, allowing you to quickly train and demo models from the local GUI interface without SSH."
The machine's CPU and GPU are not upgradeable.
Users can order on lambdalabs.com for prices ranging from $3,499 (about 22,288 RMB) to $4,999 (about 31,830 CNY).
Reference Links:
https://lambdalabs.com/deep-learning/laptops/tensorbook
https://arstechnica.com/gadgets/2022/04/razer-designed-linux-laptop-targets-ai-developers-with-deep-learning-emphasis/