DEV Community

#inference

Posts

👋 Sign in for the ability to sort posts by relevant, latest, or top.
Renting GPUs by the hour in the EU: how to size a card and read the bill

Renting GPUs by the hour in the EU: how to size a card and read the bill

Comments
5 min read
What are the best fal.ai alternatives for model inference?

What are the best fal.ai alternatives for model inference?

Comments
8 min read
Running a 2.8-trillion-parameter model without a graphics card is possible

Running a 2.8-trillion-parameter model without a graphics card is possible

Comments
3 min read
How Much GPU Memory Do I Need?

How Much GPU Memory Do I Need?

Comments
9 min read
Why a GPU is still essential, even when a CPU runs 2.8 trillion parameters

Why a GPU is still essential, even when a CPU runs 2.8 trillion parameters

Comments
3 min read
Benchmarking across some of our published families

Benchmarking across some of our published families

Comments
10 min read
What Happens When You Hit Enter?

What Happens When You Hit Enter?

Comments
5 min read
How Many Closed Findings Before You Trust a Routing Change

How Many Closed Findings Before You Trust a Routing Change

1
Comments
10 min read
Pose Estimation: MediaPipe or YOLOPose for Real-Time Pose Classification?

Pose Estimation: MediaPipe or YOLOPose for Real-Time Pose Classification?

Comments
7 min read
Test-Time Compute vs More Parameters: Where the Inference Budget Actually Goes

Test-Time Compute vs More Parameters: Where the Inference Budget Actually Goes

Comments
5 min read
수락 길이는 22.8% 늘었는데 처리량은 14.0%만 늘었습니다: DSpine 카드 수치와 공개 파일 헤더로 투기 디코딩 초안 모델을 읽는 법

수락 길이는 22.8% 늘었는데 처리량은 14.0%만 늘었습니다: DSpine 카드 수치와 공개 파일 헤더로 투기 디코딩 초안 모델을 읽는 법

Comments
3 min read
Prefix Caching vs Recomputed Prompts: The Hit-Rate Break-Even

Prefix Caching vs Recomputed Prompts: The Hit-Rate Break-Even

Comments
6 min read
The Ridge Dies at 228 Tokens on a 4090, and KV Precision Cannot Save It

The Ridge Dies at 228 Tokens on a 4090, and KV Precision Cannot Save It

Comments
6 min read
Mixture-of-Depths vs Uniform Compute: The Routing Tax on Every Token

Mixture-of-Depths vs Uniform Compute: The Routing Tax on Every Token

Comments
5 min read
12.8x faster for coding agents. At full precision.

12.8x faster for coding agents. At full precision.

1
Comments
2 min read
👋 Sign in for the ability to sort posts by relevant, latest, or top.