#zstandard

[ follow ]
Data science
fromInfoQ
1 week ago

Google's TurboQuant Compression May Support Faster Inference, Same Accuracy on Less Capable Hardware

TurboQuant compresses language models' Key-Value caches by up to 6x with near-zero accuracy loss, enabling efficient use of modest hardware.
Software development
fromInfoQ
2 weeks ago

When Every Bit Counts: How Valkey Rebuilt Its Hashtable for Modern Hardware

Redis clones offer opportunities for optimization and learning, but often lack full implementation and reliability essential for caching.
#ai
fromTechCrunch
1 month ago
Data science

Google unveils TurboQuant, a lossless AI memory compression algorithm - and yes, the internet is calling it 'Pied Piper' | TechCrunch

Data science
fromTheregister
3 weeks ago

TurboQuant is a big deal, but it won't end the memory crunch

TurboQuant is an AI data compression technology that reduces memory usage for KV caches but may not significantly alleviate memory shortages.
Data science
fromTechCrunch
1 month ago

Google unveils TurboQuant, a lossless AI memory compression algorithm - and yes, the internet is calling it 'Pied Piper' | TechCrunch

Google's TurboQuant is an ultra-efficient AI memory compression algorithm that significantly reduces memory usage without quality loss.
fromArs Technica
1 month ago

Google's TurboQuant AI-compression algorithm can reduce LLM memory usage by 6x

PolarQuant is doing most of the compression, but the second step cleans up the rough spots. Google proposes smoothing that out with a technique called Quantized Johnson-Lindenstrauss (QJL).
Roam Research
fromTheregister
1 month ago

RAM is getting expensive, so squeeze the most from it

Both work with Linux's existing swapping mechanism. Swapping (called paging in Windows) is a way for the kernel to handle running low on available RAM. It chooses pages of memory that aren't in use right now and copies them to disk, then those blocks can be marked as free and reused for something else.
Software development
Privacy technologies
fromArs Technica
1 month ago

Google quantum-proofs HTTPS by squeezing 2.5kB of data into 64-byte space

Google implements quantum-resistant cryptography in certificate transparency logs by combining classical and post-quantum encryption algorithms to prevent future attacks from quantum computers.
#brotli
fromInfoQ
2 months ago

.NET 11 Preview 1 Arrives With Runtime Async, Zstandard Support, and C# 15 Features

The headline runtime feature in this preview is Runtime Async, described as a major change to how asynchronous methods work in .NET. According to the release notes, since C# 5 introduced async/await, the compiler has been solely responsible for rewriting async methods into state machine structs that track progress across suspension points. With Runtime Async, the runtime itself now understands async methods as a first-class concept and takes over responsibility for suspending and resuming methods.
Software development
Software development
fromInfoWorld
2 months ago

8 reasons developers love Go-and 8 reasons they don't

Go is a simple, C-like programming language created at Google to simplify building concurrent, networked, large-scale software with safety, performance, and easy learning.
[ Load more ]