Razer x Lambda Linux laptop does machine learning

Lambda x Razer Tensorbook

Screenshot: Lambda

Razer latest laptop happened, but not what we expected. Not only this Razer Blade 15 Doppelganger is not intended for gaming, but it is not even sold by Razer. This unique new notebook is instead a collaboration between Razer and Lambda, a company that makes workstations, GPUs, and servers for machine learning. It’s marketed as a “deep learning” laptop, not a system aimed at gamers.

The laptop is called the Razer x Lambda Tensorbook, and it is almost the same as last year Razer Blade 15 Advanced. It shares the same organs: a 15.6-inch screen, 1440p at 165 Hz, an 11th generation Intel Core i7 processor, 64 GB of RAM, 2 TB of SSD storage and an Nvidia RTX 3080. (16GB vRAM) Laptop GPUs. And that’s it: there’s a single configuration that sells for $3,500.

Lambda makes bold claims about the laptop’s machine learning capabilities, claiming on its website that the system is up to 4x faster than a MacBook Pro 16 powered by Apple’s M1 Max chip during the execution of visual and linguistic training benchmarks.

“Most ML engineers don’t have a dedicated GPU laptop, which forces them to use shared resources on a remote machine, which slows down their development cycle.” wrote Stephen Balaban, the co-founder and CEO of Lambda, in a blog post. “When you’re stuck in SSH to a remote server, you don’t have any of your local data or code, and you even have a hard time demonstrating your model to your colleagues. The Razer x Lambda Tensorbook solves this It comes pre-installed with PyTorch and TensorFlow and lets you quickly train and test your models—all from a local GUI.”

The only customization options are for software and warranty: the $3,500 base choice comes with Ubuntu 20.04 with a Lambda stack (including drivers, PyTorch, TensorFlow, CUDA, and others) and a one-year warranty , a “standard” version costs $4,100 and gets you a two-year premium warranty, a $5,000 Enterprise edition on Windows 10, and a third year of premium warranty.

From the outside, the Tensorbook looks like a Razer Blade 15 with a custom paint job. It has the same dimensions and 4.4-pound weight, but swaps a black exterior for a silver exterior with matching white keys and replaces Razer’s famous green USB ports with purple-accented ones. If you ask me, the purple Lambda logo on the Tensorbook looks better than Razer’s three-headed green snake. Either way, the Tensorbook comes with the same robust selection of ports: two Thunderbolt 4 ports, three USB-A (3.2 Gen 2) ports, HDMI 2.1, a headphone jack, and an SD card reader.

The Tensorbook has no custom performance optimizations, so it won’t perform any better than other gaming laptops with similar or better specs. It’s not even Razer’s most powerful system now that the company has updated its own fleet to 12th generation Intel processors. What makes the Tensorbook unique is the included software package featuring one-line installation and management paths for PyTorch, TensorFlow, Caffee, and Caffee 2 deep learning frameworks, as well as CUDA and cuDNN applications. GPU-focused.

Min-Liang Tan, co-founder and CEO of Razer writes on facebook in 2017 that the company would create the “world’s best laptop supporting Linux”. It never happened. Hope the Tensorbook is a success, so Razer feels compelled to create its own Linux-based system – after all, Valve and Google are already using the operating system as a gaming platform.