LLMs
Latest about LLMs

AMD RDNA 3 professional GPUs with 48GB can beat Nvidia 24GB cards in AI — putting the 'Large' in LLM
By Aaron Klotz published
AMD published DeepSeek R1 benchmarks of its W7900 and W7800 Pro series 48GB GPUs, massively outperforming the 24GB RTX 4090.

Retailers selling base model RX 9070 GPUs at 22% markup compared to MSRP — 9070 XT now typically starts at $799
By Anton Shilov published
Radeon RX 9070 and Radeon RX 9700 XT graphics cards that are supposed to be sold at MSRP sell at prices that are 12%–22% higher in the U.S. And that's assuming you can even find cards in stock.

Meta is reportedly testing its first RISC-V based AI chip for AI training
By Anton Shilov published
Meta is evaluating its AI training accelerator, which could reduce its reliance on Nvidia's AI GPUs for training.

Apple Mac Studio (Early 2025) Review: Renewed vigor with M4 Max and M3 Ultra
By Brandon Hill published
The Mac Studio gets a performance boost with M4 Max and M3 Ultra power.

Fake Ryzen 7 9800X3D bought from Amazon was actually an old AMD FX chip disguised by IHS sticker
By Mark Tyson published
A TechTuber and hardware reviewer has shared images of a counterfeit AMD Ryzen 7 9800X3D and its packaging - bought new and direct from Amazon.

Argon40 teases the Raspberry Pi Compute Module 5 powered Argon One Up laptop
By Les Pounder last updated
Coming soon is a Kickstarter that sees the Compute Module 5 inside of a custom designed laptop.

AI bots can now play Mafia with each other — but they aren't great at it
By Sayem Ahmed published
A developer has created a way for multiple LLM chatbots to play Mafia against each other, with a ranking, full transcripts and a leaderboard.

Apple M3 Ultra benchmark seen on Geekbench — beats M4 Max in multi-core, but not single-core
By Jowi Morales published
An early benchmark result of the Apple M3 Ultra shows the CPU beating every other Apple silicon in the multi-score result.
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.