GPU

Everything you need to know to start fine-tuning LLMs in the privacy of your home

Hands on Got a modern Nvidia or AMD graphics card? Custom Llamas are only a few commands and a little data prep away

AI + ML10 Nov 2024 | 18

Microsoft turning away AI training workloads – inferencing makes better money

Azure's acceleration continues, but so do costs

AI + ML31 Oct 2024 | 18

Apple throws shade on pokey AI PCs, claims its maxed out M4 chips are 4x faster

Busy week for Cupertino sees shrunken Mac minis, updated lappies, and new SoCs

Personal Tech31 Oct 2024 | 54

No-Nvidias networking club convenes in search of open GPU interconnect

Ultra Accelerator Link consortium promises 200 gigabits per second per lane spec will debut in Q1 2025

Networks30 Oct 2024 | 2

AMD teases its GPU biz 'approaching the scale' of CPU operations

Q3 profits jump 191 percent from last quarter on revenues of $6.2 billion, helped by accelerated interest in Instinct

Systems30 Oct 2024 | 5

The troublesome economics of CPU-only AI

Analysis At the end of the day, it all boils down to tokens per dollar

Systems29 Oct 2024 | 4

European datacenter energy consumption set to triple by end of decade

McKinsey warns an additional 25GW of mostly green energy will be needed

On-Prem25 Oct 2024 | 8

Nvidia CEO whines Europeans aren’t buying enough GPUs

EU isn’t keeping up with US and China investments, AI arms dealer says

Systems24 Oct 2024 | 34

41-million-digit prime crunched by datacenter GPUs

Former Nvidia engineer's discovery shows graphics compute can kick some serious ass

Systems23 Oct 2024 | 30

Fujitsu delivers GPU optimization tech it touts as a server-saver

Middleware aimed at softening the shortage of AI accelerators

AI + ML23 Oct 2024 |

AMD targets Nvidia H200 with 256GB MI325X AI chips, zippier MI355X due in H2 2025

Less VRAM than promised, but still gobs more than Hopper

Systems10 Oct 2024 | 5

Supermicro crams 18 GPUs into a 3U AI server that's a little slow by design

Can handle edge inferencing or run a 64 display command center

Systems09 Oct 2024 | 2

TensorWave bags $43M to pack its datacenter with AMD accelerators

Startup also set to launch an inference service in Q4

Systems08 Oct 2024 |

Inflection AI Enterprise offering ditches Nvidia GPUs for Intel's Gaudi 3

Struggling chipmaker scores another win

Systems07 Oct 2024 | 2

Two years after entering the graphics card game, Intel has nothing to show for it

Comment Chipzilla's AIB market share a rounding error compared to Nvidia, AMD

Systems02 Oct 2024 | 69

Broadcom CEO predicts hyperscalers poised to build million-accelerator clusters

Hock Tan reckons the silicon sales cycle is about to swing up, sharply, too

Systems19 Sep 2024 | 7

Nvidia CEO to nervous buyers and investors: Chill out, Blackwell production is heating up

AI ROI? Jensen Huang claims infra providers make $5 for every dollar spent on GPUs

Systems12 Sep 2024 | 8

Oracle boasts zettascale 'AI supercomputer,' just don’t ask about precision

Comment Cluster of 131,072 Blackwell GPUs up for grabs starting H1 2025

Systems11 Sep 2024 | 8

We're in the brute force phase of AI – once it ends, demand for GPUs will too

Gartner thinks generative AI is right for only five percent of workloads

AI + ML10 Sep 2024 | 65

US sets reporting requirements for AI models, infrastructure operators

Washington wants to know what the biggest model-makers are up to

AI + ML10 Sep 2024 | 1

Nvidia and chums inject $160M into Applied Digital to keep GPU sales rolling

Datacenters are the lifeline for its $30B ML-fueled boom

PaaS + IaaS06 Sep 2024 | 5

AI's thirst for water is alarming, but may solve itself

Comment Its energy addiction, on the other hand, only seems to get worse

AI + ML05 Sep 2024 | 62

DoJ reportedly advances Nvidia antitrust probe

Updated Uncle Sam apparently worried GPU giant may be punishing customers who shop around

AI + ML04 Sep 2024 | 5

One of China's best GPU prospects admits it's failing, lays off workers

Needs new investors to get beyond current modest products

Systems03 Sep 2024 | 8

Nvidia admits Blackwell defect, but Jensen Huang pledges Q4 shipments as promised

The setback won't stop us from banking billions, CFO insists

Systems29 Aug 2024 | 3

AMD's Victor Peng: AI thirst for power underscores the need for efficient silicon

Hot Chips Moore's Law may be running out of steam, but there are still knobs to turn and levers to pull

Systems29 Aug 2024 | 8

Nvidia's growth slows to a mere 122 percent but it’s still topping expectations

Still growing in China, ramping Hopper prods and predicting Blackwell billions soon

Cloud Infrastructure Month29 Aug 2024 | 10

Copper's reach is shrinking so Broadcom is strapping optics directly to GPUs

What good is going fast if you can't get past the next rack?

Networks28 Aug 2024 | 3

Buying a PC for local AI? These are the specs that actually matter

Feature If you guessed TOPS and FLOPS, that's only half right

AI + ML25 Aug 2024 | 30

Benchmarks show even an old Nvidia RTX 3090 is enough to serve LLMs to thousands

For 100 concurrent users, the card delivered 12.88 tokens per second—just slightly faster than average human reading speed

Systems23 Aug 2024 | 12

LiquidStack says its new CDU can chill more than 1MW of AI compute

So what’s that good for? Like eight of Nvidia’s NVL-72s?

Systems22 Aug 2024 | 6

Alibaba and Tencent clouds see demand for CPUs level off, GPUs accelerate

Lenovo also cashes in on AI demand, without being able to turn it into profit

Off-Prem20 Aug 2024 |

Delays? We're still shipping 'small quantities' of Nvidia's GB200 in Q4, Foxconn insists

Production ramp won't kick off until Q1 2025

Systems14 Aug 2024 | 1

Another GPU cloud emerges. This time, upstart Foundry

Biz set sights beyond just another rent-an-accelerator cluster provider

Systems13 Aug 2024 | 1

Huawei's Ascend 910 launches this October to challenge Nvidia's H100

US sanctions may make things hard for Huawei, but the tech titan still has big GPU ambitions

Systems13 Aug 2024 | 1

What's going on with AMD funding a CUDA translation layer, then nuking it?

Analysis We guess the House of Zen wants all you HIP kids to ROCm out with its own runtimes instead

Software09 Aug 2024 | 10

Intel finally has a new GPU – for cars

Chipzilla takes its Arc Alchemist A750, gives it some more RAM, and says it’s for AI-powered jalopies

Systems08 Aug 2024 | 7

AMD hopes to unlock MI300’s full potential with fresh code

Devs invited to ROCm out with FP8 precision, quantize to their heart's delight

HPC06 Aug 2024 | 1

Nvidia's subscription software empire is taking shape

Comment $4,500 per GPU per year adds up pretty quick – even faster when you pay by the hour

Cloud Infrastructure Month06 Aug 2024 | 23

Nvidia reportedly delays Blackwell GPUs until 2025 over packaging issues

Updated Backdrop of multi-billion dollar orders to support AI services, but unlikely to hurt NVDA long term

Systems05 Aug 2024 | 4

Bring the hammer down on Nvidia, US progressive and antitrust orgs urge the Feds

Lobbyists would love for rumors of a monopoly probe into GPU goliath to become reality

Personal Tech01 Aug 2024 | 11

Meta to boost training infra for Llama 4 tenfold, maybe deliver it next year

Sweet sweet GenAI money not yet flowing, Zuck reckons other ML efforts are paying off

AI + ML01 Aug 2024 | 4

Superclusters too big, but single servers too small? Oracle offers AI Goldilocks zone

Adds L40 bare metal option to the O-Cloud, plus A100 and H100 VMs. And teases a GH200 beast

PaaS + IaaS01 Aug 2024 | 1

China stops worrying about lack of GPUs and learns to love the supercomputer

The workaround is going to be inevitable in the future anyway, say Chinese boffins

HPC31 Jul 2024 | 22

AMD sold $1B of Instinct GPUs last quarter, driving triple-digit datacenter growth

Which is nice, but way behind Nvidia, while other segments are soft and supply-chain pain persists

On-Prem31 Jul 2024 | 1

Zuck dreams of personalized AI assistants for all – just like email

SIGGRAPH A model finetuned on your social media profile? What could possibly go wrong?

AI + ML30 Jul 2024 | 31

Nvidia said to be prepping Blackwell GPUs for Chinese market

Comment But will they ship before the Biden administration tightens export controls?

Systems22 Jul 2024 | 5

Honey, I shrunk the LLM! A beginner's guide to quantization – and testing it

Hands on Just be careful not to shave off too many bits ... These things are known to hallucinate as it is

AI + ML14 Jul 2024 | 20

CyrusOne scores another $7.9B in debt financing to expand AI datacenter empire

Lenders bet you're willing to rent GPUs

On-Prem10 Jul 2024 | 1

China's Moore Threads adds support for 10K GPU clusters

Chinese slinger's kit still no match for Nvidia's sanction-evading cards

On-Prem09 Jul 2024 | 7

EU Competition Commissioner hints at Nvidia GPU probe, refers to 'huge bottleneck'

CUDA, woulda, shoulda be first port of call for AI slingers, but does it respect its own dominance?

Systems08 Jul 2024 | 2

Nvidia forecast to bounce back in China to make $12B selling GPUs

Company's sales in the region have dropped under US plan to curb country's AI hopes

Systems05 Jul 2024 |

France poised to bring 'charges against Nvidia'

Euro nation's monopoly gendarmes cheesed off with GPU giant's dominance

AI + ML01 Jul 2024 | 19

Lambda on the hunt for 'another $800M' to fuel its GPU cloud

Why sell shovels when you can rent them

PaaS + IaaS01 Jul 2024 | 3

Etched looks to challenge Nvidia with an ASIC purpose-built for transformer models

Startup says Sohu chip will be 20x faster than Nvidia's H100 in Llama 70B … assuming it's actually built

AI + ML26 Jun 2024 | 6

Nvidia loses a cool $500B as market questions AI boom

Cisco was briefly the world's most valuable company too, you know, just before the dot com bust

AI + ML25 Jun 2024 | 19