https://www.tomshardware.com/news/google-a3-supercomputer-h100-googleio Skip to main content (*) ( ) Open menu Close menu Tom's Hardware [ ] Search Search Tom's Hardware [ ] RSS (opens in new tab) (opens in new tab) (opens in new tab) (opens in new tab) (opens in new tab) US Edition flag of US flag of UK UK flag of US US flag of Australia Australia flag of Canada Canada * * Reviews * Best Picks * Raspberry Pi * CPUs * GPUs * Coupons * Newsletter * More + Laptops + SSDs + Motherboards + Cooling + Desktops + PC Builds + Monitors + RAM + PC Cases + Keyboards + Headsets + Mice + Power Supplies + 3D Printers + Gaming Chairs + Webcams + Microphones + VR Headsets + About Tom's Hardware Forums (opens in new tab) Trending * Ryzen 7 7800X3D * Raptor Lake * Ryzen 9 7950X3D * RTX 4090 Ti / Titan RTX Ada * RX 7900 XTX When you purchase through links on our site, we may earn an affiliate commission. Here's how it works. 1. Home 2. News Google Launches AI Supercomputer Powered by Nvidia H100 GPUs By Brandon Hill published 10 May 2023 Google's A3 supercomputer delivers up to 26 exaFlops of AI performance * (opens in new tab) * (opens in new tab) * * (opens in new tab) * (opens in new tab) * (opens in new tab) * Comments (6) Nvidia H100 (Image credit: Nvidia) Google kicked off Google I/O this afternoon by talking for more than an hour about its numerous advances in artificial intelligence. The company discussed its new PaLM 2 large language model (LLM) for generative AI, which powers the Bard chatbot tool. This is a foundational pillar for adding AI-infused features across Google's product portfolio, including Google Maps, Google Photos, and Gmail (among others). With that in mind, there is a need for some serious horsepower in the cloud to power models in the wild, as millions (and eventually billions) of users send requests for operations as mundane as removing a person lingering in the background of a picture to composing an entire email for you based on a short text prompt. That's where Google's new A3 GPU supercomputer comes into focus. Google says the new A3 supercomputers are "purpose-built to train and serve the most demanding AI models that power today's generative AI and large language model innovation" while delivering 26 exaFlops of AI performance. Each A3 supercomputer is packed with 4th generation Intel Xeon Scalable processors backed by 2TB of DDR5-4800 memory. But the real "brains" of the operation come from the eight Nvidia H100 "Hopper" GPUs, which have access to 3.6 TBps of bisectional bandwidth by leveraging NVLink 4.0 and NVSwitch. According to Google, A3 represents the first production-level deployment of its GPU-to-GPU data interface, which allows for sharing data at 200 Gbps while bypassing the host CPU. This interface, which Google calls the Infrastructure Processing Unit (IPU), results in a 10x uplift in available network bandwidth for A3 virtual machines (VM) compared to A2 VMs. "Google Cloud's A3 VMs, powered by next-generation NVIDIA H100 GPUs, will accelerate training and serving of generative AI applications," said Ian Buck, VP for hyperscale and high-performance computing at NVIDIA. "On the heels of Google Cloud's recently launched G2 instances, we're proud to continue our work with Google Cloud to help transform enterprises around the world with purpose-built AI infrastructure." If your business wants to leverage A3 virtual machines, the only way to gain access is by filling out Google's A3 Preview Interest Form to join the Early Access Program. But as Google clearly states, plugging in your information doesn't guarantee a spot in the program. Be In the Know Get instant access to breaking news, in-depth reviews and helpful tips. [ ][ ]Contact me with news and offers from other Future brands[ ]Receive email from us on behalf of our trusted partners or sponsors[Sign me up] By submitting your information you agree to the Terms & Conditions (opens in new tab) and Privacy Policy (opens in new tab) and are aged 16 or over. Brandon Hill Brandon Hill Social Links Navigation Brandon Hill is a senior editor at Tom's Hardware. He has written about PC and Mac tech since the late 1990s with bylines at AnandTech, DailyTech, and Hot Hardware. When he is not consuming copious amounts of tech news, he can be found enjoying the NC mountains or the beach with his wife and two sons. More about ai Google Labs AI experiments How to Get Access to Google's New AI Tools Robot AI thinking about something OpenAI Whisper Audio Transcription Benchmarked on 18 GPUs: Up to 3,000 WPM Latest MSI Monitor MSI 27-Inch 170 Hz Curved WQHD Monitor Now $220 at Newegg See more latest > Topics AI See all comments (6) [ ] 6 Comments Comment from the forums * gg83 Here we go. The most advance text message answering service on the planet! I can't wait Reply * GoldyTwatus gg83 said: Here we go. The most advance text message answering service on the planet! I can't wait Damn you aren't easily impressed, you must be pretty cool. Reply * Ravestein NL Admin said: Google's A3 GPU supercomputer is the company's latest platform for powering LLMs. Google Launches AI Supercomputer Powered by Nvidia H100 GPUs : Read more I have my reservations about all the AI's growing in the wild. Who is checking what data the AI has access to? Who is checking if the answers the AI gives on certain questions is factually correct? In other words I'm afraid that this tech can easily be used to feed people only the data that's "allowed" and true facts can easily stay hidden. There is always someone pushing the buttons and is that someone to be trusted? Everybody knows that self regulation in big tech companies doesn't exist! Reply * gg83 GoldyTwatus said: Damn you aren't easily impressed, you must be pretty cool. Huh? I'm confused by this reply. Reply * Strider379 I call BS. Frontier is now the fastest computer in the world and just hit 1.7 exaflops. Nvidia says that this computer is 15 times faster? Reply * samopa Strider379 said: I call BS. Frontier is now the fastest computer in the world and just hit 1.7 exaflops. Nvidia says that this computer is 15 times faster? They say that A3 is "delivering 26 exaFlops of AI performance", so most likely it is not usual (normal) floating point operations. Normal Floating Point Operation usually take 80 bit data (or more) per operations, but in AI, such precision is not needed, hence they (usually) use TF32 or even FP8 data format for floating point operations. So, assuming they used FP8 to produce this result, its only need 1/10 performance (hence 10 time faster) than normal floating point operations. They also do not disclosed how many A3 are used to produce this result. Reply * View All 6 Comments Show more comments Most Popular [missing-im]Cryptominers Repurpose GPU Farms Amid AI Hardware Shortage By Mark TysonMay 11, 2023 [missing-im]DDR5-10708: Colorful's CVN Icicle Modules Almost Hit DDR5 Record By Anton ShilovMay 11, 2023 [missing-im]Nvidia CEO Gets $2.5 Million Pay Cut, Probably Won't Notice By Mark TysonMay 11, 2023 [missing-im]OpenAI Whisper Audio Transcription Benchmarked on 18 GPUs: Up to 3,000 WPM By Jarred WaltonMay 11, 2023 [missing-im]AMD Radeon RX 7600 Specs Seemingly Leaked By Francisco PiresMay 11, 2023 [missing-im]Raspberry Pi Weather Station Uses One Sensor, Displays Data via Web Based Dashboard By Ash HillMay 11, 2023 [missing-im]Xbox Series S Suffers from VRAM Limitations, Just Like 8GB GPUS By Aaron KlotzMay 10, 2023 [missing-im]Quantinuum Injects Topology Into Ion-Chain Quantum Entanglement to Solve Quantum Computing Error Correction By Francisco PiresMay 10, 2023 [missing-im]Google Announces New Bard Features, Drops Waitlist, Integrates With Workspace By Sarah Jacobsson PurewalMay 10, 2023 [missing-im]Chinese Chip Industry to Focus on Perfecting Mature Nodes: Report By Anton ShilovMay 10, 2023 [missing-im]Google Launches AI Supercomputer Powered by Nvidia H100 GPUs By Brandon HillMay 10, 2023 MOST POPULARMOST SHARED 1. MSI Monitor 1 MSI 27-Inch 170 Hz Curved WQHD Monitor Now $220 at Newegg 2. 2 MSI Latest SSD Pushes PCIe 4 to the Limit: Up to 7.4 GB/s and 1M IOPS 3. 3 Stanford Researchers Leverage Palladium for Disruptive Memristor Design 4. 4 Raspberry Pi Alternative Orange Pi 5 Plus Brings RK3588 to the SBC Game 5. 5 Crucial's PCIe 5.0 SSD Throttles to HDD Speeds Without Cooler, but Avoids Thermal Shutdown 1. MSI Monitor 1 MSI 27-Inch 170 Hz Curved WQHD Monitor Now $220 at Newegg 2. 2 MSI Latest SSD Pushes PCIe 4 to the Limit: Up to 7.4 GB/s and 1M IOPS 3. 3 Stanford Researchers Leverage Palladium for Disruptive Memristor Design 4. 4 Raspberry Pi Alternative Orange Pi 5 Plus Brings RK3588 to the SBC Game 5. 5 Crucial's PCIe 5.0 SSD Throttles to HDD Speeds Without Cooler, but Avoids Thermal Shutdown Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher. Visit our corporate site (opens in new tab). * Terms and conditions (opens in new tab) * Contact Future's experts (opens in new tab) * Privacy policy (opens in new tab) * Cookies policy (opens in new tab) * Accessibility Statement (opens in new tab) * Advertise (opens in new tab) * About us (opens in new tab) * Coupons (opens in new tab) * Careers (opens in new tab) (c) Future US, Inc. Full 7th Floor, 130 West 42nd Street, New York, NY 10036.