If Nvidia and AMD are licking their lips thinking about all of the GPUs they can sell to Microsoft to support its huge aspirations in generative AI – particularly when it comes to the OpenAI GPT large language model that is the centerpiece of all of the company’s future software and services – they had better think again. …
The post Microsoft’s Chiplet Cloud To Bring The Cost Of LLMs Way Down first appeared on The Next Platform.
Microsoft’s Chiplet Cloud To Bring The Cost Of LLMs Way Down was written by Timothy Prickett Morgan at The Next Platform.
The National Center for Supercomputing Applications at the University of Illinois just fired up its Delta system back in April 2022, and now it has just been given $10 million by the National Science Foundation to expand that machine with an AI partition, called DeltaAI appropriately enough, that is based on Nvidia’s “Hopper” H100 GPU accelerators. …
The post NCSA Builds Out Delta Supercomputer With An AI Extension first appeared on The Next Platform.
NCSA Builds Out Delta Supercomputer With An AI Extension was written by Timothy Prickett Morgan at The Next Platform.
The cloud has been a boon for enterprises trying to manage the massive amounts of data they collect every year. …
The post Giving Cloud Data Warehouses A Relational Knowledge Graph Overlay first appeared on The Next Platform.
Giving Cloud Data Warehouses A Relational Knowledge Graph Overlay was written by Jeffrey Burt at The Next Platform.
The question is no longer whether or not the “El Capitan” supercomputer that has been in the process of being installed at Lawrence Livermore National Laboratory for the past week – with photographic evidence to prove it – will be the most powerful system in the world. …
The post Lining Up The “El Capitan” Supercomputer Against The AI Upstarts first appeared on The Next Platform.
Lining Up The “El Capitan” Supercomputer Against The AI Upstarts was written by Timothy Prickett Morgan at The Next Platform.
When system architects sit down to design their next platforms, they start by looking at a bunch of roadmaps from suppliers of CPUs, accelerators, memory, flash, network interface cards – and PCI-Express controllers and switches. …
The post PCI-Express Must Match The Cadence Of Compute Engines And Networks first appeared on The Next Platform.
PCI-Express Must Match The Cadence Of Compute Engines And Networks was written by Timothy Prickett Morgan at The Next Platform.
Here’s a fun question and don’t cheat by asking ChatGPT. What is more valuable, an ounce of gold or an ounce of an Nvidia “Hopper” H100 GPU accelerator? …
The post AI To The Rescue For Server And Storage Spending In Q1 first appeared on The Next Platform.
AI To The Rescue For Server And Storage Spending In Q1 was written by Timothy Prickett Morgan at The Next Platform.
If you want to get the attention of server makers and compute engine providers and especially if you are going to be building GPU-laden clusters with shiny new gear to drive AI training and possibly AI inference for large language models and recommendation engines, the first thing you need is $1 billion. …
The post The $1 Billion And Higher Ante To Play The AI Game first appeared on The Next Platform.
The $1 Billion And Higher Ante To Play The AI Game was written by Timothy Prickett Morgan at The Next Platform.
There is no shortage of silicon photonics technologies under development, and every few months it seems like another startup crops up promising massive bandwidth, over longer distances, while using less power than copper interconnects. …
Photonics To Make Celestial HBM3 Memory Fabric was written by Tobias Mann at The Next Platform.
The hype around generative AI is making every industry vibrate at an increasingly high pitch in a way that we have not seen since the days of the Dot Com boom and sock puppets. …
With Huge Costs, Efficiency Is The Key To Mainstreaming Generative AI was written by Jeffrey Burt at The Next Platform.
When it comes to a lot of high performance computing systems we have seen over the decades, we are fond of saying that the hardware is the easy part. …
Argonne Aurora A21: All’s Well That Ends Better was written by Timothy Prickett Morgan at The Next Platform.
It was a fortuitous coincidence that Nvidia was already working on massively parallel GPU compute engines for doing calculations in HPC simulations and models when the machine learning tipping point happened, and similarly, it was fortunate for InfiniBand that it had the advantage of high bandwidth, low latency, and remote direct memory access across GPUs at that same moment. …
Cisco Guns For InfiniBand With Silicon One G200 was written by Timothy Prickett Morgan at The Next Platform.
To Andrew Davidson, senior vice president of products at MongoDB, the database business operates in an entirely different type of market than traditional software, where vendors might sell their products into one organization after another, eventually reaching a saturation point. …
Beefing Up A Cloudy NoSQL Database To Ride The AI Wave was written by Jeffrey Burt at The Next Platform.
When it comes to deploying Arm in the cloud, a lot of the talk of late has centered on things like efficiency, core density, or predictability of performance. …
Move Over X86, Amazon’s Arm HPC instances Are Live was written by Tobias Mann at The Next Platform.
If the datacenter is the computer – and it certainly is for hundreds of companies comprising somewhere well north of half of server sales worldwide – then the Ethernet fabric, consisting of switches and routers, is the backplane of that computer. …
Ethernet Switching Still Booming As Routing Soldiers On was written by Timothy Prickett Morgan at The Next Platform.
The expanded lineup of AMD’s 4th generation “Genoa” Epyc server chips – built atop “Zen 4” core and some with the chip maker’s L3-boosting 3D V-Cache – unveiled at a high-profile event in San Francisco this week is quickly making its way into the cloud. …
The Big Clouds Get First Dibs On AMD “Genoa” Chips was written by Jeffrey Burt at The Next Platform.
If you had to rank the level of hype around specific datacenter technologies, the top thing these days would be, without question, generative AI, probably followed by AI training and inference of all kinds and mixed precision computing in general. …
Rest In Pieces: Servers And CXL was written by Timothy Prickett Morgan at The Next Platform.
OpenStack, the venerable open source cloud controller born in 2010 out projects pulled together by Rackspace Hosting and NASA, continues to push on despite its death being predicted myriad times over the past several years. …
Reports Of OpenStack’s Death Greatly Exaggerated was written by Jeffrey Burt at The Next Platform.
The best defense is a good offense, and as it turns out, the best offense is also a good offense. …
AMD Widens Server CPU Line To Take Bites Out Of Intel, Arm was written by Timothy Prickett Morgan at The Next Platform.
The great thing about the Cambrian explosion in compute that has been forced by the end of Dennard scaling of clock frequencies and Moore’s Law lowering in the cost of transistors is not only that we are getting an increasing diversity of highly tuned compute engines and broadening SKU stacks across those engines, but also that we are getting many different interpretations of the CPU, GPU, DPU, and FPGA themes. …
The Third Time Charm Of AMD’s Instinct GPU was written by Timothy Prickett Morgan at The Next Platform.
Any performance comparisons across compute engines in use in a datacenter are always valid for a point in time since new CPUs, GPUs, FPGAs, and other ASICs are always coming into the market at different times. …
Intel Pits Its “Sapphire Rapids” Xeon SP Against AMD “Genoa” Epycs was written by Timothy Prickett Morgan at The Next Platform.