TensorPool Blog

Insights, tutorials, and updates from the TensorPool team on GPU infrastructure, distributed computing, and AI workload optimization.

Introducing TensorPool Object Storage

Globally distributed S3-compatible object storage with intelligent caching that follows your GPUs. Up to 20x lower cross-region latency than Cloudflare R2, with up to 3.5x higher throughput.
Tycho Svoboda
Mar 30, 2026
5 min read
Read more

Introducing nb.dev: A Cracked Jupyter Notebook for ML

Jupyter notebooks are broken for ML. We built nb.dev — an ML-first notebook with GPU attach, full machine snapshots, and one-click branching for parallel experimentation.
Tycho Svoboda · Cofounder & CEO
Mar 11, 2026
5 min read
Read more

How ZeroEntropy Trained the zerank Model Family on TensorPool

How ZeroEntropy's reranker models achieved SOTA performance through innovative training methods and elastic GPU access.
Tycho Svoboda · Cofounder & CEO
Nov 24, 2025
4 min read
Read more

Introducing TensorPool Jobs: Git-Style GPU Workflows

Run your ML training jobs like you push code to GitHub. Pay only for runtime, get your results back automatically.
Tycho Svoboda · Cofounder & CEO
Oct 30, 2025
4 min read
Read more

Why B200s Are Cheaper for Large-Scale Training Than H100s

Despite 2x higher hourly costs, B200 GPUs deliver 10-20% lower total training costs for large models. Here's the math.
Tycho Svoboda · Cofounder & CEO
Oct 24, 2025
4 min read
Read more

Why Fast NFS Storage Matters for ML Training

How TensorPool's high-performance NFS storage eliminates I/O bottlenecks and accelerates your machine learning workflows.
Tycho Svoboda · Cofounder & CEO
Oct 8, 2025
3 min read
Read more

Getting Started with the TensorPool CLI

A comprehensive guide to managing your GPU infrastructure from the command line with our powerful CLI tool.
Tycho Svoboda · Cofounder & CEO
Sep 27, 2025
5 min read
Read more