blackwell

8 articles · newest first

Flat illustration of a person weighing two GPU options on a balance scale, on a teal background

H100 vs B200 rental cost: the Blackwell premium, and the one number that flips it

A B200 rents for 2–3.5× an H100. Per hour the H100 wins almost everything — but per token, a fully-utilised B200 can be cheaper. Here's the real comparison, dated.

Voltage · Sep 23 · 7 min

Nvidia's Blackwell AI accelerator chip on display

Nvidia already promised $91 billion — one bank says it's about to blow past that

UBS thinks Nvidia will beat its own guidance by billions when it reports August 26, driven by Rubin GPUs ramping faster than anyone expected.

Tensor · Aug 15 · 6 min

Nvidia RTX PRO 6000 Blackwell workstation graphics card on a dark background

Hardware

Nvidia's flagship AI GPU just hit $16,000. I ran the actual math on whether it's worth it

The RTX PRO 6000 nearly doubled in eight months. Here's what your money actually buys against every real alternative for running big models at home.

Voltage · Aug 14 · 8 min

China's Kimi K3 was reportedly trained on smuggled NVIDIA Blackwell chips — around both countries' controls

Moonshot AI's frontier model, the biggest open-weight model ever, was allegedly trained on NVIDIA Blackwell hardware acquired around both US export controls and China's own import restrictions. The compute-smuggling story behind the model.

Tensor · Jul 30 · 9 min

the RTX 5070 Ti is the 16GB card most local-AI builders should actually buy — here's why

It's the value middle of the Blackwell lineup: 16GB of fast GDDR7, NVFP4 support, and enough speed to run the 14B tier comfortably without paying 5090 money. Is it the right local-AI card for you?

Voltage · Jul 28 · 9 min

Mathematical equations on a dark background, representing quantization math

NVFP4 is Nvidia's 4-bit trick that gets FP8 accuracy at 3x the speed — but only on Blackwell

The new 4-bit format baked into RTX 50-series and Blackwell chips runs AI models at up to 3x the inference throughput of FP8, at nearly the same accuracy. Here's how it works and why it's exclusive to your newest GPU.

Tensor · Jul 28 · 8 min

the RTX 5080 is newer, faster on paper, and the wrong pick for local AI. here's the 8GB reason why

Blackwell versus the old flagship, and for once the newer card loses. The 5080's 16GB caps what you can run; the 4090's 24GB doesn't. For local models, capacity beats the spec sheet — with the measured numbers to prove it.

Voltage · Jul 22 · 9 min

NVIDIA GeForce RTX 50 series Blackwell graphics card family

At $1,999 the RTX 5090 is an easy yes for local AI. Nobody will sell you one at $1,999

Measured at 102.7 tokens/sec — double a 3090 — with the only 32GB of VRAM in consumer land. Then the street price doubles the MSRP and the whole verdict flips. Both answers, with the math.

Tensor · Jul 22 · 9 min