Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Since announcing Jalapeño, our first custom inference chip, we’ve been testing it and the system around it. The results show a major advance: more intelligence from every watt and faster responses, delivering both higher throughput and lower latency in one architecture without sacrificing efficiency.

OpenAI's profile picture

OpenAI

5,494,005 subscribers

3,084,495 Aufrufe • vor 2 Monaten •via X (Twitter)

Wissenschaft & TechnologieBildungNachrichten & Politik
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Before: A system focused only on connecting two of our major cities. After: A system delivering faster, cheaper and more reliable transport system across the UK. Read more:
0:15

Before: A system focused only on connecting two of our major cities. After: A system delivering faster, cheaper and more reliable transport system across the UK. Read more:

UK Prime Minister

1,168,648 Aufrufe • vor 3 Jahren

First V4 cabinet Supercharger now live 0.5 MW, 3X power density, 2X stalls per cabinet Higher throughput, higher efficiency, lower cost, faster deployments
0:19

First V4 cabinet Supercharger now live 0.5 MW, 3X power density, 2X stalls per cabinet Higher throughput, higher efficiency, lower cost, faster deployments

Tesla Charging

2,224,623 Aufrufe • vor 1 Jahr

Our self-improving agents optimized the full @vLLM_project inference stack, with up to 16% more throughput and interactivity for DeepSeek v4 Pro and Z.ai GLM 5.2 on B200s (no MTP). Every change was verified and our agents got better and faster at it with each iteration.
0:16

Our self-improving agents optimized the full @vLLM_project inference stack, with up to 16% more throughput and interactivity for DeepSeek v4 Pro and Z.ai GLM 5.2 on B200s (no MTP). Every change was verified and our agents got better and faster at it with each iteration.

Asari AI

143,394 Aufrufe • vor 3 Monaten

We’ve seen great results using GPT-4 for content policy development and content moderation, enabling more consistent labeling, a faster feedback loop for policy refinement, and less involvement from human moderators. Built on top of the GPT-4 API:
0:16

We’ve seen great results using GPT-4 for content policy development and content moderation, enabling more consistent labeling, a faster feedback loop for policy refinement, and less involvement from human moderators. Built on top of the GPT-4 API:

OpenAI

902,641 Aufrufe • vor 3 Jahren

Since announcing my candidacy for governor last week, we’ve been on the road meeting with voters and hearing what’s most important to them. Learn more about our America First campaign at and join the team!
0:14

Since announcing my candidacy for governor last week, we’ve been on the road meeting with voters and hearing what’s most important to them. Learn more about our America First campaign at and join the team!

Dave Yost

15,321 Aufrufe • vor 1 Jahr

We got our hands on NVIDIA's new Vera CPU and put it to the test. We built a benchmark using real-world agentic workloads to see how it performs under contention. In early tests, Vera delivered up to 2x the throughput of comparable x86 hardware, maintaining low latency even as hundreds of concurrent workloads competed for resources. For Modal Sandboxes, that could mean running more agents in parallel without sacrificing latency.
0:15

We got our hands on NVIDIA's new Vera CPU and put it to the test. We built a benchmark using real-world agentic workloads to see how it performs under contention. In early tests, Vera delivered up to 2x the throughput of comparable x86 hardware, maintaining low latency even as hundreds of concurrent workloads competed for resources. For Modal Sandboxes, that could mean running more agents in parallel without sacrificing latency.

Modal

27,552 Aufrufe • vor 13 Tagen

Our newest AI accelerator Maia 200 is now online in Azure. Designed for industry-leading inference efficiency, it delivers 30% better performance per dollar than current systems. And with 10+ PFLOPS FP4 throughput, ~5 PFLOPS FP8, and 216GB HBM3e with 7TB/s of memory bandwidth it's optimized for large-scale AI workloads. It joins our broader portfolio of CPUs, GPUs, and custom accelerators, giving customers more options to run advanced AI workloads faster and more cost-effectively on Azure.
0:11

Our newest AI accelerator Maia 200 is now online in Azure. Designed for industry-leading inference efficiency, it delivers 30% better performance per dollar than current systems. And with 10+ PFLOPS FP4 throughput, ~5 PFLOPS FP8, and 216GB HBM3e with 7TB/s of memory bandwidth it's optimized for large-scale AI workloads. It joins our broader portfolio of CPUs, GPUs, and custom accelerators, giving customers more options to run advanced AI workloads faster and more cost-effectively on Azure.

Satya Nadella

851,488 Aufrufe • vor 9 Monaten

Our first preprint as a lab is out! This one is short and sweet: We thought there had to be a faster/higher throughput way of making transgenic fish, so we set out to test if lipofection works, and it does!
0:11

Our first preprint as a lab is out! This one is short and sweet: We thought there had to be a faster/higher throughput way of making transgenic fish, so we set out to test if lipofection works, and it does!

Adrian Jacobo

27,858 Aufrufe • vor 2 Jahren

Nano Banana 2 Lite is now live on Fotor! Create high-quality AI images with faster generation, lower latency, and greater efficiency. Whether you're exploring ideas, creating social content, or producing visuals at scale, Nano Banana 2 Lite helps you go from prompt to image in seconds. Try it now via the link in our bio.
0:14

Nano Banana 2 Lite is now live on Fotor! Create high-quality AI images with faster generation, lower latency, and greater efficiency. Whether you're exploring ideas, creating social content, or producing visuals at scale, Nano Banana 2 Lite helps you go from prompt to image in seconds. Try it now via the link in our bio.

Fotor

10,717 Aufrufe • vor 4 Monaten

Ultralytics YOLO11 vs Ultralytics YOLO26: Which one runs faster? ⚡ Evaluate inference speed and performance across both models in the ONNX format to understand how YOLO26 improves efficiency for real-time computer vision applications. Read more ➡️ #Ultralytics #YOLO26 #AI #Research #ComputerVision
0:12

Ultralytics YOLO11 vs Ultralytics YOLO26: Which one runs faster? ⚡ Evaluate inference speed and performance across both models in the ONNX format to understand how YOLO26 improves efficiency for real-time computer vision applications. Read more ➡️ #Ultralytics #YOLO26 #AI #Research #ComputerVision

Ultralytics

22,821 Aufrufe • vor 6 Monaten

Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue Across 100+ use cases in our product, keeping all up to date with the right model is a challenge - Either we're losing out on intelligence for the dollars we spend, or we're spending too much money for the intelligence the feature needs Ramp Router lets you benefit immediately without rewriting your application. We cut our LLM costs by 30%, while also making our features smarter and faster. We built it for Ramp. Now we’re opening it up to everyone. get access here:
0:12

Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue Across 100+ use cases in our product, keeping all up to date with the right model is a challenge - Either we're losing out on intelligence for the dollars we spend, or we're spending too much money for the intelligence the feature needs Ramp Router lets you benefit immediately without rewriting your application. We cut our LLM costs by 30%, while also making our features smarter and faster. We built it for Ramp. Now we’re opening it up to everyone. get access here:

rahul

962,791 Aufrufe • vor 3 Monaten

THIS JUST IN !! "One Eyed Bastard" (Live In Amsterdam) has arrived ⚡️ Thank you for making the song our 4th #1 at Alt radio from Saviors. Ya know, it's the first album to pull that off since '03, and we couldn't have done it without you. Now turn it upppp Nearly two years of Saviors and what a ride it's been. One more month of shows to send it off right. THANK YOU for being part of it!
0:13

THIS JUST IN !! "One Eyed Bastard" (Live In Amsterdam) has arrived ⚡️ Thank you for making the song our 4th #1 at Alt radio from Saviors. Ya know, it's the first album to pull that off since '03, and we couldn't have done it without you. Now turn it upppp Nearly two years of Saviors and what a ride it's been. One more month of shows to send it off right. THANK YOU for being part of it!

Green Day

42,974 Aufrufe • vor 1 Jahr

Consistency has been hyped for years, yet usable solutions remain scarce. Kling O1 is the first system I’ve used that delivers on reference alignment and cross-scene continuity. Lighter workflow, lower cost, and production-grade results all in one. This isn’t incremental progress; it reshapes the boundaries of practical content creation.
0:10

Consistency has been hyped for years, yet usable solutions remain scarce. Kling O1 is the first system I’ve used that delivers on reference alignment and cross-scene continuity. Lighter workflow, lower cost, and production-grade results all in one. This isn’t incremental progress; it reshapes the boundaries of practical content creation.

Atal

15,405 Aufrufe • vor 9 Monaten

Announcing our new streamlined product portfolio! 💻 🌐 We’re making it easier and faster to find the right PCs, accessories and services with a new unified naming structure. Learn more about the updates here:
0:15

Announcing our new streamlined product portfolio! 💻 🌐 We’re making it easier and faster to find the right PCs, accessories and services with a new unified naming structure. Learn more about the updates here:

Dell Technologies

24,527 Aufrufe • vor 1 Jahr

🔥 GLM-5.3-Flash Hits #1 on 🐮 GLM-5.3-Flash (Ox Alpha) is now the most-used and most popular model on with cumulative token throughput surpassing 2.41 trillion tokens. As the first native omni-modal model in the GLM-5 series, it packs 320B total parameters, with 18B active, and features a hybrid sparse and linear attention architecture with a 1M-token context window. The result: lightning-fast responses, powerful reasoning, and outstanding cost efficiency. 🎁 Still 100% FREE on From high-frequency API calls and coding to complex Agents and long-document processing, jump in and experience the #1 model on for yourself! 👉 Try it free now:
0:13

🔥 GLM-5.3-Flash Hits #1 on 🐮 GLM-5.3-Flash (Ox Alpha) is now the most-used and most popular model on with cumulative token throughput surpassing 2.41 trillion tokens. As the first native omni-modal model in the GLM-5 series, it packs 320B total parameters, with 18B active, and features a hybrid sparse and linear attention architecture with a 1M-token context window. The result: lightning-fast responses, powerful reasoning, and outstanding cost efficiency. 🎁 Still 100% FREE on From high-frequency API calls and coding to complex Agents and long-document processing, jump in and experience the #1 model on for yourself! 👉 Try it free now:

B.AI

222,739 Aufrufe • vor 1 Monat

Introducing Microsoft-Decision-1, our new model for fast decision-making. It delivers top performance on structured decision tasks, outperforming both LLMs and other decision models in latency and quality. We’re already testing it across Microsoft for everything from incident response and quality control to scientific discovery.
0:10

Introducing Microsoft-Decision-1, our new model for fast decision-making. It delivers top performance on structured decision tasks, outperforming both LLMs and other decision models in latency and quality. We’re already testing it across Microsoft for everything from incident response and quality control to scientific discovery.

Satya Nadella

1,706,566 Aufrufe • vor 23 Tagen

Grok 2 mini is now 2x faster than it was yesterday. In the last three days Lianmin Zheng and Saeed Maleki rewrote our inference stack from scratch using SGLang ( This has also allowed us to serve the big Grok 2 model, which requires multi-host inference, at a reasonable speed. Both models didn’t just get faster, but also slightly more accurate. Stay tuned for further speed improvements!
0:11

Grok 2 mini is now 2x faster than it was yesterday. In the last three days Lianmin Zheng and Saeed Maleki rewrote our inference stack from scratch using SGLang ( This has also allowed us to serve the big Grok 2 model, which requires multi-host inference, at a reasonable speed. Both models didn’t just get faster, but also slightly more accurate. Stay tuned for further speed improvements!

Igor Babuschkin

1,803,113 Aufrufe • vor 2 Jahren

It’s been one week since Kyber Network integrated NEAR Intents, and the results are already changing the game for cross-chain swaps. Users are now tapping into deeper liquidity, faster finality, and smoother UX, without bridging or slippage. 🌉🚫
0:18

It’s been one week since Kyber Network integrated NEAR Intents, and the results are already changing the game for cross-chain swaps. Users are now tapping into deeper liquidity, faster finality, and smoother UX, without bridging or slippage. 🌉🚫

NEAR Protocol

23,143 Aufrufe • vor 1 Jahr

Jensen announced on stage at #COMPUTEX2026 that we'll be among the first cloud providers to bring up NVIDIA Vera Rubin NVL72, to help accelerate the era of AI factories. As part of NVIDIA's full-stack AI factory platform, Vera Rubin NVL72 is designed to run frontier and open-source models with exceptional performance, efficiency, and scalability—delivering more intelligence per watt and lower token costs at scale. It's GO time. ⚡️
0:16

Jensen announced on stage at #COMPUTEX2026 that we'll be among the first cloud providers to bring up NVIDIA Vera Rubin NVL72, to help accelerate the era of AI factories. As part of NVIDIA's full-stack AI factory platform, Vera Rubin NVL72 is designed to run frontier and open-source models with exceptional performance, efficiency, and scalability—delivering more intelligence per watt and lower token costs at scale. It's GO time. ⚡️

CoreWeave

28,320 Aufrufe • vor 5 Monaten