#GPUS

Gerrit Eicker

@eicker.bsky.social

16 hours ago

#Nvidia’s annual GTC conference will showcase new details about its #agenticoptimised #CPUs, signalling a shift in focus from #GPUs to CPUs due to the increasing demand for general compute power in #agentic #AIworkflows. Nvidia’s CPUs, designed to optimise GPU performance, are facing a “quiet…

LLMs

@llms.activitypub.awakari.com.ap.brid.gy

22 hours ago

These are the biggest AI chip rivals Nvidia is facing AI chip competition heats up as Nvidia faces pressure from Google, startups, and China amid evolving AI workloads and geopolitical issues.

#Tech #AI #Enterprise #nvidia #chips #gpus #ai-inference #jensen-huang #google #cerebras #amd

Origin | Interest | Match

TornadoVM

@tornadovm.org

1 day ago

New paper accepted at #ICPE2026!

We present TornadoVMPulse, a profiling & visualization tool for #TornadoVM to analyze and optimize #Java apps on #GPUs.

Results for #GPULlama3:
⚡ 18.4% faster execution
📉 38% lower data-transfer cost

🇪🇺 @aero-project.eu
📍 Florence, May26

zenodo.org/records/1895...

Virtualization & Cloud Review

@virtreview.bsky.social

2 days ago

What GPU You Really Need for AI Workloads -- Virtualization Review Understanding GPU memory requirements is essential for AI workloads, as VRAM capacity--not processing power--determines which models you can run, with total memory needs typically exceeding model size by 20-50% due to weights, activations, KV cache, and system overhead.

Selecting the right GPU for AI workloads often comes down to VRAM capacity. Model weights, KV cache, activations, and overhead can push memory needs to 1.2–1.5x the model size.

Read the full piece by Brien Posey: https://ow.ly/zGpP50YsOrh

#AI #GPUs #MachineLearning

Winbuzzer

@winbuzzer.com

4 days ago

NVIDIA Dynamo and Brev Scale AI Agent Inference to Planetary Level NVIDIA has detailed how its Dynamo inference framework and Brev platform tackle planetary-scale AI inference, cutting per-token costs 35x on GB200 hardware.

winbuzzer.com/2026/03/10/n...

NVIDIA Dynamo and Brev Scale AI Agent Inference to Planetary Level

#AI #NVIDIA #AIInference #AIAgents #AgenticAI #NvidiaBlackwell #NvidiaGPUs #GPUs #AIModels #AIInfrastructure #NvidiaDynamo #NvidiaBrev #AgentInference

Research Computing Services

@rcscambridge.bsky.social

4 days ago

Th RCS team will be at the Cross-Community AMD GPU Hackathon (11–13 March, Birmingham), helping researchers port and optimise scientific codes for AMD GPUs alongside DiRAC and AMD.

Looking forward to supporting some great research software

dirac.ac.uk/training_eve...

#HPC #GPUs #ResearchSoftware

EuroHPC Joint Undertaking

@eurohpc-ju.bsky.social

4 days ago

Join the @epicurehpc.bsky.social Hackathon on Code Optimisation for Heterogeneous Environments.

From 27-29 April 2026, #HPC users will gather for three days of optimising real codes across #CPUs, #GPUs and heterogeneous architectures.

🔗 survey.inesctec.pt/index.php/76...

StockTitan

@stocktitan.net

4 days ago

Hyperscale Data Bitcoin Treasury at 617.1605 Bitcoin; Cash and Bitcoin Holdings at Approximately 137.98% of Market Capitalization Hyperscale Data (NYSE: GPUS) reported a cryptocurrency and cash treasury totaling approximately $82.9 million as of the week ended March 8, 2026, equating to about 137.98% of its market capitalization based on the March 9, 2026 share price.The company holds 617.1605 BTC (valued at ~$40.7 million at $65,970 per BTC) and approximately $42.2 million in cash and restricted cash, and intends to deploy its cash into additional Bitcoin purchases over time.

#GPUS Hyperscale Data Bitcoin Treasury at 617.1605 Bitcoin; Cash and Bitcoin Holdings at Approximately 137.98% of Market Capitalization

www.stocktitan.net/news/GPUS/hyperscale-dat...

Data Center Frontier

@datacenterfrontier.bsky.social

5 days ago

Why Conduit Choices Matter as Construction for Resilient, High Density Data Centers Soars Matt Fredericks of Champion Fiberglass explains why conduit material selection is not just a code-compliance exercise, it is also a risk management decision.

Everyone talks about #GPUs & #gigawatts. But one of the biggest #datacenter #risk decisions is conduit. Pipes protecting medium-voltage cables underground. Burn-through. Fault heat. Water ingress. Soil chemistry. Invisible infrastructure. Big consequences.

www.datacenterfrontier.com/sponsored/ar...

Winbuzzer

@winbuzzer.com

5 days ago

U.S. Draft Rules Would Require Permits for All AI Chip Exports The U.S. Commerce Department has circulated draft rules requiring government permits for all AI chip exports worldwide, hitting Nvidia and AMD stocks.

winbuzzer.com/2026/03/09/u...

U.S. Draft Rules Would Require Permits for All AI Chip Exports

#AI #AIChips #NVIDIA #AMD #GPUs #AI #Semiconductors #USExportControls #SemiconductorGeopolitics #BigTech #ChinaAI #TrumpAdministration #USGovernment

linux

@linux.activitypub.awakari.com.ap.brid.gy

1 week ago

AMD VP uses AI to create Radeon Linux userland driver in Python — senior AI engineer says he "didn't open the editor once" AMD's VP of AI software vibe coded the driver entirely u...

#GPU #Drivers #PC #Components #GPUs

Origin | Interest | Match

AERO Project

@aero-project.eu

1 week ago

Wonderful news from our partners from @manchester.ac.uk!

They will present a poster regarding #TornadoVM and how it enables the development of #AI libraries in #Java targeting NVIDIA #GPUs!

➿ www.nvidia.com/gtc/session-...

#horizoneurope
@hipeac.bsky.social
@eucloudedgeiot.bsky.social

StockTitan

@stocktitan.net

1 week ago

Hyperscale Data's Subsidiary askROI Deploys Claude Opus 4.6 to Power Next-Generation Artificial Intelligence Reasoning, Agents and Enterprise Intelligence Hyperscale Data (NYSE:GPUS) announced that subsidiary askROI deployed Claude Opus 4.6 on its askROI platform on March 5, 2026, upgrading from an older Anthropic model.The integration expands reasoning depth, contextual comprehension and multi-step analytics to power AI agents, enterprise workflows and large-scale data synthesis.

#GPUS Hyperscale Data's Subsidiary askROI Deploys Claude Opus 4.6 to Power Next-Generation Artificial Intelligence Reasoning, Agents and Enterprise Intelligence

www.stocktitan.net/news/GPUS/hyperscale-dat...

Data Center Frontier

@datacenterfrontier.bsky.social

1 week ago

NTT DATA’s Quiet Surge: Hyperscale Wins, AI Partnerships, and Private 5G Push Signal Expanding Influence in the AI Infrastructure Era A series of announcements spanning hyperscale capacity deals, AI partnerships, private 5G deployments, and workforce recognition reveals how one of the world’s largest data center...

Nearly 250MW of #hyperscale capacity.

That’s what #NTT DATA Global Data Centers has recently locked in across major U.S. campuses.

The #AI boom isn’t just #GPUs.

It’s also the quiet giants of #digitalinfrastructure scaling the physical backbone of AI.

www.datacenterfrontier.com/colocation/n...

Dustcircle

@dustcircle.bsky.social

1 week ago

How Nvidia became the cornerstone of modern computing From an idea hatched in a Denny's booth to a place on the Dow to an AI cornerstone, the chipmaker now finds itself at the center of modern computing.

#Denny’s, #GPUs & #AI: The journey of #Nvidia

www.thestreet.com/technology/d...

The Fablelist x fabletide

@fabletide.bsky.social

1 week ago

this story went wild. what started as corporate fraud became a fever dream where Nvidia's CEO uses Arby's as a money laundering front, #GPUs have deep fryer heat sinks, and household appliances are filing restraining orders over unpaid dividends.

StockTitan

@stocktitan.net

1 week ago

Hyperscale Data Bitcoin Treasury at 610.9188 Bitcoin; Cash and Bitcoin Holdings at Approximately 141.93% of Market Capitalization Hyperscale Data (NYSE:GPUS) reported a Bitcoin treasury of 610.9188 BTC as of March 1, 2026, valued at approximately $40 million using a $65,738 BTC closing price. The company also held $41.5 million in cash and restricted cash.Combined cash, restricted cash and Bitcoin holdings totaled about $81.5 million, equal to 141.93% of market capitalization based on the March 2, 2026 share price. The company intends to deploy its allocated cash into additional Bitcoin purchases over time.

#GPUS Hyperscale Data Bitcoin Treasury at 610.9188 Bitcoin; Cash and Bitcoin Holdings at Approximately 141.93% of Market Capitalization

www.stocktitan.net/news/GPUS/hyperscale-dat...

IEEE Spectrum

@spectrum.ieee.org

1 week ago

Prices for the #DRAM used to feed #GPUs in AI data centers have skyrocketed, leaving personal computer makers and other manufacturers in a bind. When will supply rise to meet demand? spectrum.ieee.org/dram-shortage

Adam

@impracticalfuture.bsky.social

1 week ago

With Revenue Share Shrinking, Does Nvidia Need Gaming Anymore? In the last quarter, Nvidia revenue topped $68 billion, but just 5.5% of that came from gaming. That's still a nice chunk of change, but it's dwarfed by the company's AI data center business.

With Revenue Share Shrinking, Does Nvidia Need Gaming Anymore?

In the last quarter, revenue topped $68 billion, but just 5.5% of that came from gaming.

buff.ly/vKv0kkn

#news #tech #technology #AI #aislop #nvidia #gaming #gpus

Seth

@strmfrnt.bsky.social

2 weeks ago

Digital Mining | Best Bitcoin Mining App | GoMining Discover the world of digital mining with GoMining! Join us for legit Bitcoin mining opportunities and unlock your potential in the crypto space today.

Start mining BTC by purchasing real-world mining hashpower through the power of blockchain technologies.

Invest in $NFTs that represent real-world GPUs! 🚀

#Mining #Blockchain #NFTs #GPUs

gomining.com?ref=g-rRE

Don Porqué

@donporque.bsky.social

2 weeks ago

¿Meta compra chips AMD por 100.000 millones y roza el 10%? Meta ata 6 gigavatios de chips de IA con AMD en un pacto de hasta 100.000 millones y una vía para rozar 10 % del gigante del silicio global. Meta Platforms ha firmado con Advanced Micro Devices (AMD) ...

¿Meta compra chips AMD por 100.000 millones y roza el 10%? #Meta #AMD #InteligenciaArtificial #ChipsIA #Semiconductores #CentrosDeDatos #GPUs #Nvidia #WallStreet #Tecnologia #Economia #Mercados #24defebrero #felizmartes #SiliconValley donporque.com/meta-compra-...

it business

@itbusiness-schweiz.bsky.social

2 weeks ago

Meta baut KI-Infrastruktur mit AMD aus – «it business» – Meldungen aus der ICT-Welt

AMD vertieft die Partnerschaft mit Meta und plant KI-Infrastruktur im Gigawatt-Massstab. Mehrere Generationen von Instinct-GPUs sollen grosse KI-Modelle antreiben. Start der ersten Auslieferungen: zweite Hälfte 2026.

#AMD #Meta #KI #AI #KIInfrastruktur #GPUs #Rechenzentren

AD News Live

@liveadnews.bsky.social

2 weeks ago

Meta AMD 6GW GPU Deal: Massive AI Build-Out Is AMD taking the AI crown? The Meta AMD 6GW GPU Deal brings 6GW of chips to Helios systems. See why this partnership matters now.

Meta & AMD just announced a massive AI chip deal that could redefine the future of tech. This is the kind of move you can’t ignore—read the full story now. 🚀

👉 Click here to dive in. www.liveadnews.com/2026/02/meta...

#AI #TechNews #GPUs

Wall Street Game Notes

@wallstgamenotes.bsky.social

2 weeks ago

Nvidia Q4 Earnings Incoming! Here's what you need to know! 🤓📑 #WallStreetGameNotes YouTube video by Wall Street Game Notes

The most important stock in the market reports tomorrow. Here's what you need to know!

#WallStreetGameNotes #Investing #AI #GPUs #Stocks

$NVDA $MSFT $AMZN $META $AMD

youtube.com/shorts/-sQjM...

Lowyat.NET

@lowyat.bsky.social

2 weeks ago

Another Thief In South Korea Made Off With Three GPUs Worth US$14,000 Another robbery of PC parts occurred in South Korea, this time with the thief making off with US$14,000 worth of high-end GPUs.

Another Thief In South Korea Made Off With Three GPUs Worth US$14,000 #gpus #randomasitgets #southkorea #thief

Pure Science

@purescience.news

2 weeks ago

7,000 GPUs Simulate Quantum Microchip in Unprecedented Detail Using the Perlmutter supercomputer, researchers achieved a record-scale simulation of a quantum microchip to refine and validate next-generation quantum hardware designs. Researchers from Lawrence Berkeley National Laboratory (Berkeley Lab) and the University of California, Berkeley have complete

7,000 GPUs Simulate Quantum Microchip in Unprecedented Detail #Science #Physics #QuantumPhysics #QuantumComputing #GPUs #Simulation

scitechdaily.com/7000-gpus-simulate-quant...

Winbuzzer

@winbuzzer.com

2 weeks ago

winbuzzer.com/2026/02/21/n...

Nvidia Scales Back OpenAI Deal to $30 Billion Equity Stake

#AI #NVIDIA #OpenAI #BigTech #AIInvestment #AIInfrastructure #AIPartnerships #Semiconductors #Chips #GPUs #NvidiaGPUs #AICompute #ChatGPT #SamAltman #JensenHuang #DataCenters

Pure Tech

@puretech.news

2 weeks ago

A $300 Radeon RX 9060 XT just smashed the world GPU overclocking record #Technology #Hardware #GraphicsCards #RadeonRX9060XT #GPUs #Overclocking

www.techspot.com/news/111422-300-radeon-r...

Winbuzzer

@winbuzzer.com

3 weeks ago

winbuzzer.com/2026/02/18/n...

NVIDIA Deploys 20,000+ Blackwell GPUs Across India for Sovereign AI

#AI #NVIDIA #NvidiaBlackwell #GPUs #BigTech #India #BlackwellUltra #AIDevelopment #AIInfrastructure

EPICURE Project

@epicurehpc.bsky.social

3 weeks ago

Have you registered for the #Hackathon on Code Optimisation for Heterogeneous Environments?
Join researchers from @deucalionacnca.bsky.social and @bsc-cns.bsky.social and work on real codes across #CPUs, #GPUs and heterogeneous architectures.

🔗https://shorturl.at/zZbgr

@inesctec.bsky.social #HPC