#HyperComputer

[govindhtech]

Introducing Trillium, Google Cloud’s sixth generation TPUs

Trillium TPUs

The way Google cloud engage with technology is changing due to generative AI, which is also creating a great deal of efficiency opportunities for corporate effect. However, in order to train and optimise the most powerful models and present them interactively to a worldwide user base, these advancements need on ever-increasing amounts of compute, memory, and communication. Tensor Processing Units, or TPUs, are unique AI-specific hardware that Google has been creating for more than ten years in an effort to push the boundaries of efficiency and scale.

Many of the advancements Google cloud introduced today at Google I/O, including new models like Gemma 2, Imagen 3, and Gemini 1.5 Flash, which are all trained on TPUs, were made possible by this technology. Google cloud thrilled to introduce Trillium, Google’s sixth-generation TPU, the most powerful and energy-efficient TPU to date, to offer the next frontier of models and empower you to do the same.

Comparing Trillium TPUs to TPU v5e, a remarkable 4.7X boost in peak computation performance per chip is achieved. Google cloud increased both the Interchip Interconnect (ICI) bandwidth over TPU v5e and the capacity and bandwidth of High Bandwidth Memory (HBM). Third-generation SparseCore, a dedicated accelerator for handling ultra-large embeddings frequently found in advanced ranking and recommendation workloads, is another feature that Trillium has. Trillium TPUs provide faster training of the upcoming generation of foundation models, as well as decreased latency and cost for those models. Crucially, Trillium TPUs are more than 67% more energy-efficient than TPU v5e, making them Google’s most sustainable TPU generation to date.

Up to 256 TPUs can be accommodated by Trillium in a single high-bandwidth, low-latency pod. In addition to this pod-level scalability, Trillium TPUs can grow to hundreds of pods using multislice technology and Titanium Intelligence Processing Units (IPUs). This would allow a building-scale supercomputer with tens of thousands of chips connected by a multi-petabit-per-second datacenter network.

The next stage of Trillium-powered AI innovation

Google realised over ten years ago that a novel microprocessor was necessary for machine learning. They started developing the first purpose-built AI accelerator in history, TPU v1, in 2013. In 2017, Google cloud released the first Cloud TPU. Many of Google’s best-known services, including interactive language translation, photo object recognition, and real-time voice search, would not be feasible without TPUs, nor would cutting-edge foundation models like Gemma, Imagen, and Gemini. Actually, Google Research’s foundational work on Transformers the algorithmic foundations of contemporary generative AI Fwas made possible by the size and effectiveness of TPUs.

Compute performance per Trillium chip increased by 4.7 times

Since TPUs Google cloud created specifically for neural networks, Google cloud constantly trying to speed up AI workloads’ training and serving times. In comparison to TPU v5e, Trillium performs 4.7X peak computing per chip. We’ve boosted the clock speed and enlarged the size of matrix multiply units (MXUs) to get this level of performance. Additionally, by purposefully offloading random and fine-grained access from TensorCores, SparseCores accelerate workloads that involve a lot of embedding.

The capacity and bandwidth of High Bandwidth Memory (HBM) with 2X ICI

Trillium may operate with larger models with more weights and larger key-value caches by doubling the HBM capacity and bandwidth. Higher memory bandwidth, enhanced power efficiency, and a flexible channel architecture are made possible by next-generation HBM, which also boosts memory throughput. For big models, this reduces serving latency and training time. This equates to twice the model weights and key-value caches, allowing for faster access and greater computational capability to expedite machine learning tasks. Training and inference tasks may grow to tens of thousands of chips with double the ICI bandwidth thanks to a clever mix of 256 chips per pod specialised optical ICI interconnects and hundreds of pods in a cluster via Google Jupiter Networking.

The AI models of the future will run on trillium

The next generation of AI models and agents will be powered by trillium TPUs, and they are excited to assist Google’s customers take use of these cutting-edge features. For instance, the goal of autonomous car startup Essential AI is to strengthen the bond between people and computers, and the company anticipates utilising Trillium to completely transform the way organisations function. Deloitte, the Google Cloud Partner of the Year for AI, will offer Trillium to transform businesses with generative AI.

Nuro is committed to improving everyday life through robotics by training their models with Cloud TPUs. Deep Genomics is using AI to power the future of drug discovery and is excited about how their next foundational model, powered by Trillium, will change the lives of patients. With support for long-context, multimodal model training and serving on Trillium TPUs, Google Deep Mind will also be able to train and serve upcoming generations of Gemini models more quickly, effectively, and with minimal latency.

AI-powered trillium Hypercomputer

The AI Hypercomputer from Google Cloud, a revolutionary supercomputing architecture created especially for state-of-the-art AI applications, includes Trillium TPUs. Open-source software frameworks, flexible consumption patterns, and performance-optimized infrastructure including Trillium TPUs are all integrated within it. Developers are empowered by Google’s dedication to open-source libraries like as JAX, PyTorch/XLA, and Keras 3. Declarative model descriptions created for any prior generation of TPUs can be directly mapped to the new hardware and network capabilities of Trillium TPUs thanks to support for JAX and XLA. Additionally, Hugging Face and they have teamed up on Optimum-TPU to streamline model serving and training.

Since 2017, SADA (An Insight Company) has won Partner of the Year annually and provides Google Cloud Services to optimise effect.

The variable consumption models needed for AI/ML workloads are also provided by AI Hypercomputer. Dynamic Workload Scheduler (DWS) helps customers optimise their spend by simplifying the access to AI/ML resources. By scheduling all the accelerators concurrently, independent of your entry point Vertex AI Training, Google Kubernetes Engine (GKE), or Google Cloud Compute Engine flex start mode can enhance the experience of bursty workloads like training, fine-tuning, or batch operations.

Lightricks is thrilled to recoup value from the AI Hypercomputer’s increased efficiency and performance.

Read more on govindhtech.com

[theonion]

Greetings, earthlings. I am Commander Byxxurian from Nebula Vriphlaxor-9. I come bearing a message of utmost importance from the galactic consortium. Its intended recipient is one who lives among you, and if it is not delivered quickly, then I fear all hope will be lost. Please, we do not have much time. You must take me to your girlboss at once.

My fellow Vriphlaxons and I have observed your peculiar species for many Earth years, often hearing tell of an all-powerful life-form you call the she-EO. This is the one who slays in both her professional and her per- sonal life—the one who is not afraid to fight dirty to manifest her career goals. We seek counsel with her right away. According to our hypercomputer’s calculations, the fate of the universe hinges upon this she-creature and her ability to hold her own in a man’s world.

People of Earth, we beg of you: Provide the coordinates of this human you call boss babe without delay! Full Story

[seat-safety-switch]

Sometimes I've just had enough of uncertainty. Done with probing the mysteries of the universe. Tired of meticulously unearthing the hidden truth behind seemingly unrelated phenomena. Sick of putting the puzzle pieces together. That's when television comes in.

A lot of people are going to tell you that television is less hardcore now that you can watch it whenever you want, wherever you want. And it is true that back in the day, "television watchers," as they were known, had to wait for the correct time to watch their programs. Personally, I think it's a lot harder now.

Before, you'd turn on your set, and you'd have one or two channels that worked. French CBC, or the local public-access channel. At eight p.m. your choices were either watching Les Simpson, or that show where the lady waxes a clown. Now, you sit down and you are immediately blasted by nearly thirty thousand new television programs, some of which were synthesized by an array of hypercomputers to fit your as-yet-unspoken innermost whims as soon as they heard you coming in the room.

What's been lost is the ability to just watch whatever was already on. Don't get all anxious wondering if you're really optimizing your television-viewing time as well as you possibly could. You won't relax a single iota that way. That's why I'm starting a new business. For a mere thirty bucks a month, we'll put Netflix on in your house, and then lose the remote. Whatever our installer picks is what you're going to be watching, freed from that false fantasy of "choice" for all eternity. Don't think cutting the power will work, either: we'll know.

Yes, a lot of these shows will be about small engine repair. I gotta get my views up somehow. Otherwise I will be forced to go back to a job where I spend a lot of time thinking about stuff instead of reacting to pictures of scary V-twins.

[emo-space-wizard]

I think that new applicants are being put off by the baseless and unfounded rumours of the last few being driven to gibbering madness :(

Following 73 spell matrix explosions and at least two inadvertently created magic dead zones, I think I will stop trying to devise a ritual to determine from a description of an arbitrary computer program and an input, whether the program will finish running, or continue to run forever.

[andmaybegayer]

annoyingly the most influential thing in my head from Murderbot is the Feed. The Feed is so appealing as a concept. A universally comprehensible and widely accessible mesh-based standard for mixed machine and human communication with a reliable subscription service, that has a ton of things publishing constantly by default but with a robust (assuming no rogue antisocial biological hypercomputers) cryptographic system for secret data. The Feed very clearly can scale up to an entire space station network or down to one person and their accoutrements forming a personal area network.

Some kind of hierarchical MQTT implemented on a mesh network with public key cryptography could get you so close if you just standardized ways to get the MQTT schema of any given endpoint. After that it's all dashboarding. Publish as much or as little information as you want and allow any given device to access that data and respond to it.

Some of this is probably leftovers from $PREVJOB. When I was doing industrial automation our entire fleet of devices lived on a strict schema under one MQTT server that I had complete read/write for so by doing clever subscriptions I could sample anything from a temperature reading from a single appliance in the back of the restaurant I was standing in to accumulating the overall power draw of every single business we were working with all at once.

On more than one occasion something silently broke and I drove up near the back of a facility in my car where I knew the IoT gateway was, connected to our local IoT network over wifi, fixed the issue, and left without telling anyone.

Unfortunately if the Feed existed people like me would make extremely annoying endpoints. To be fair canonically the Feed is absolute dogshit unless you have an Adblock. So really it would be exactly the same as in the books.

[endlessmazin]

x516-architecture microprocessor hypercomputing // dual-core central processing unit // Celestica Dual Core // Corps 3-M // release: 2076 // made in Enceladus // Processed with Hex{Impact{Driver by datasculptor & microarchitect code named M4ZINVM33 // Distributing in black market by Datamonger.

[rory-multifandom-mess]

BRO In order to actually see any posts on your blog I have to use a tumblr acct spying website and either filter for posts via tag, or jump to PAGE 7 cause of, yk, the 100+ rb incident...

Sincerely, someone who was gonna say hi but got terrified,

-some random hypercomputer

P.S: hello btw. Also is it ethical to reincarnate drones into cyborgs?

Oh no i’m sorry if i scared you LMAO…. My love for Thad maaayyy have gone a liiiiittle overboard

As for your question, uuumm… No? Yes? Mmmmaybe? I don’t know i’m not great at ethics, you should see the shit i put Thad through just for the fun of it

[loveletterworm]

i wouldnt say deviantart "pivoted to ai" so much as its such a massive website still its impossible to moderate the same way its always been impossible to moderate little kids uploading video game and tv show renders they didnt make. its actually the only website i've seen with a "suppress ai" button that mostly manages to filter it out which I appreciate. Still not as good as something like Newgrounds blanket ban on AI but Newgrounds is a website actually run by humans and not an old internet monolith hypercomputer buried 60 miles under antarctica so I get it.

My phrasing of deviantart "pivoting to AI" was less so about the moderation not banning it (honestly yeah, the logistics of doing such a thing on deviantart in particular seem unlikely) but because they kind of literally have added their own AI generator to the site and link to it in the header

(I also saw banner ad-type things for it on the bottom of the page while observing the site to write my post last night, but I haven't been able to get them to show up again to include a screenshot now)

So it's less so "They allow it to be posted at all" and moreso "They're now using it as a selling point" (somehow it made me laugh that more uses of the on-site AI generator is apparently now a feature of Core membership even though that entirely makes sense to do within the already-existing structures of deviantart) which I think is enough to be at least a little bit of a pivot. Maybe it's a slightly hyperbolic use of the word pivot I guess it could be more of a slight rotation

[curiosofthechasm]

Legal   Name:   Asova

Nicknames:

The Shopkeeper

Eye of Virgo

God of Sales

Proprietor

Clerk of the Chasm

Date  of  Birth:   Presumed to be either a very long time ago... or beyond the classical idea of time.

Gender:   ??? (Usually appears female)

Place   of   Birth:  ??? Presumably somewhere in the cosmos but...

Currently   Living:  Possibly beyond this dimension, spacetime gets a little weird around her

Spoken   Languages:

Asova has never failed to communicate with a customer, apparently.

Education: 

Appears to be learning constantly

Physical   Characteristics:

Hair Color: Dark Ruby Red

Eye Color:  Bright Gold

Miscellaneous: A mouth full of sharp chompers. Spectacles. Might also be a neutron star hypercomputing space entity.

Height: 5'6"

Weight Mass:  1.9-2.4 solar masses

Relationships:

Noelle: Customer Employee

Neph: Customer Employee

Celio: Customer Employee

Effectively everyone else who has ever met Asova is a customer to her eyes

Orientation:

???

Relationship Status:

???, probably doesn't have relationships in any sense that organic creatures can comprehend

Tagging anyone who wants to do it!

[colfy-wolfy]

FLOATY COME GET UR DOG IT'S CAUSING A CRISIS!! IT'S EATING MY ETHERNET CONNECTION!! IT'S TAKING AN EXTRA 31.000000050000010001/ SECONDS TO REFRESH A YOUTUBE PAGE!!!!

Sincerely, a hypercomputer doing nothing, Fox S.

oh hell naw, that aint mine!

[elepharchy]

Our physicists have made progress on the study of the "astral threads" discovered by the @guildsre.

It is known that this strange material is some kind of condensed spacetime anomaly. We have now found a remarkable use for it: Careful use of astral threads allowed the creation of a bubble of "compressed" spacetime such that an experimental hypercomputer was able to run at processing speeds far beyond what conventional physics would allow.

Unfortunately, the scarcity of astral threads makes it impossible to implement this technology at scale until a reliable method for their synthesis is found. In any case, the full reports have been shared with our allies in PEACE as well as the Re'iran Travelling Guilds, in accordance with the agreement that allowed us access to astral thread samples in the first place.

[govindhtech]

A3 Ultra VMs With NVIDIA H200 GPUs Pre-launch This Month

Strong infrastructure advancements for your future that prioritizes AI

To increase customer performance, usability, and cost-effectiveness, Google Cloud implemented improvements throughout the AI Hypercomputer stack this year. Google Cloud at the App Dev & Infrastructure Summit:

Trillium, Google’s sixth-generation TPU, is currently available for preview.

Next month, A3 Ultra VMs with NVIDIA H200 Tensor Core GPUs will be available for preview.

Google’s new, highly scalable clustering system, Hypercompute Cluster, will be accessible beginning with A3 Ultra VMs.

Based on Axion, Google’s proprietary Arm processors, C4A virtual machines (VMs) are now widely accessible

AI workload-focused additions to Titanium, Google Cloud’s host offload capability, and Jupiter, its data center network.

Google Cloud’s AI/ML-focused block storage service, Hyperdisk ML, is widely accessible.

Trillium A new era of TPU performance

Trillium A new era of TPU performance is being ushered in by TPUs, which power Google’s most sophisticated models like Gemini, well-known Google services like Maps, Photos, and Search, as well as scientific innovations like AlphaFold 2, which was just awarded a Nobel Prize! We are happy to inform that Google Cloud users can now preview Trillium, our sixth-generation TPU.

Taking advantage of NVIDIA Accelerated Computing to broaden perspectives

By fusing the best of Google Cloud’s data center, infrastructure, and software skills with the NVIDIA AI platform which is exemplified by A3 and A3 Mega VMs powered by NVIDIA H100 Tensor Core GPUs it also keeps investing in its partnership and capabilities with NVIDIA.

Google Cloud announced that the new A3 Ultra VMs featuring NVIDIA H200 Tensor Core GPUs will be available on Google Cloud starting next month.

Compared to earlier versions, A3 Ultra VMs offer a notable performance improvement. Their foundation is NVIDIA ConnectX-7 network interface cards (NICs) and servers equipped with new Titanium ML network adapter, which is tailored to provide a safe, high-performance cloud experience for AI workloads. A3 Ultra VMs provide non-blocking 3.2 Tbps of GPU-to-GPU traffic using RDMA over Converged Ethernet (RoCE) when paired with our datacenter-wide 4-way rail-aligned network.

In contrast to A3 Mega, A3 Ultra provides:

With the support of Google’s Jupiter data center network and Google Cloud’s Titanium ML network adapter, double the GPU-to-GPU networking bandwidth

With almost twice the memory capacity and 1.4 times the memory bandwidth, LLM inferencing performance can increase by up to 2 times.

Capacity to expand to tens of thousands of GPUs in a dense cluster with performance optimization for heavy workloads in HPC and AI.

Google Kubernetes Engine (GKE), which offers an open, portable, extensible, and highly scalable platform for large-scale training and AI workloads, will also offer A3 Ultra VMs.

Hypercompute Cluster: Simplify and expand clusters of AI accelerators

It’s not just about individual accelerators or virtual machines, though; when dealing with AI and HPC workloads, you have to deploy, maintain, and optimize a huge number of AI accelerators along with the networking and storage that go along with them. This may be difficult and time-consuming. For this reason, Google Cloud is introducing Hypercompute Cluster, which simplifies the provisioning of workloads and infrastructure as well as the continuous operations of AI supercomputers with tens of thousands of accelerators.

Fundamentally, Hypercompute Cluster integrates the most advanced AI infrastructure technologies from Google Cloud, enabling you to install and operate several accelerators as a single, seamless unit. You can run your most demanding AI and HPC workloads with confidence thanks to Hypercompute Cluster’s exceptional performance and resilience, which includes features like targeted workload placement, dense resource co-location with ultra-low latency networking, and sophisticated maintenance controls to reduce workload disruptions.

For dependable and repeatable deployments, you can use pre-configured and validated templates to build up a Hypercompute Cluster with just one API call. This include containerized software with orchestration (e.g., GKE, Slurm), framework and reference implementations (e.g., JAX, PyTorch, MaxText), and well-known open models like Gemma2 and Llama3. As part of the AI Hypercomputer architecture, each pre-configured template is available and has been verified for effectiveness and performance, allowing you to concentrate on business innovation.

A3 Ultra VMs will be the first Hypercompute Cluster to be made available next month.

An early look at the NVIDIA GB200 NVL72

Google Cloud is also awaiting the developments made possible by NVIDIA GB200 NVL72 GPUs, and we’ll be providing more information about this fascinating improvement soon. Here is a preview of the racks Google constructing in the meantime to deliver the NVIDIA Blackwell platform’s performance advantages to Google Cloud’s cutting-edge, environmentally friendly data centers in the early months of next year.

Redefining CPU efficiency and performance with Google Axion Processors

CPUs are a cost-effective solution for a variety of general-purpose workloads, and they are frequently utilized in combination with AI workloads to produce complicated applications, even if TPUs and GPUs are superior at specialized jobs. Google Axion Processors, its first specially made Arm-based CPUs for the data center, at Google Cloud Next ’24. Customers using Google Cloud may now benefit from C4A virtual machines, the first Axion-based VM series, which offer up to 10% better price-performance compared to the newest Arm-based instances offered by other top cloud providers.

Additionally, compared to comparable current-generation x86-based instances, C4A offers up to 60% more energy efficiency and up to 65% better price performance for general-purpose workloads such as media processing, AI inferencing applications, web and app servers, containerized microservices, open-source databases, in-memory caches, and data analytics engines.

Titanium and Jupiter Network: Making AI possible at the speed of light

Titanium, the offload technology system that supports Google’s infrastructure, has been improved to accommodate workloads related to artificial intelligence. Titanium provides greater compute and memory resources for your applications by lowering the host’s processing overhead through a combination of on-host and off-host offloads. Furthermore, although Titanium’s fundamental features can be applied to AI infrastructure, the accelerator-to-accelerator performance needs of AI workloads are distinct.

Google has released a new Titanium ML network adapter to address these demands, which incorporates and expands upon NVIDIA ConnectX-7 NICs to provide further support for virtualization, traffic encryption, and VPCs. The system offers best-in-class security and infrastructure management along with non-blocking 3.2 Tbps of GPU-to-GPU traffic across RoCE when combined with its data center’s 4-way rail-aligned network.

Google’s Jupiter optical circuit switching network fabric and its updated data center network significantly expand Titanium’s capabilities. With native 400 Gb/s link rates and a total bisection bandwidth of 13.1 Pb/s (a practical bandwidth metric that reflects how one half of the network can connect to the other), Jupiter could handle a video conversation for every person on Earth at the same time. In order to meet the increasing demands of AI computation, this enormous scale is essential.

Hyperdisk ML is widely accessible

For computing resources to continue to be effectively utilized, system-level performance maximized, and economical, high-performance storage is essential. Google launched its AI-powered block storage solution, Hyperdisk ML, in April 2024. Now widely accessible, it adds dedicated storage for AI and HPC workloads to the networking and computing advancements.

Hyperdisk ML efficiently speeds up data load times. It drives up to 11.9x faster model load time for inference workloads and up to 4.3x quicker training time for training workloads.

With 1.2 TB/s of aggregate throughput per volume, you may attach 2500 instances to the same volume. This is more than 100 times more than what big block storage competitors are giving.

Reduced accelerator idle time and increased cost efficiency are the results of shorter data load times.

Multi-zone volumes are now automatically created for your data by GKE. In addition to quicker model loading with Hyperdisk ML, this enables you to run across zones for more computing flexibility (such as lowering Spot preemption).

Developing AI’s future

Google Cloud enables companies and researchers to push the limits of AI innovation with these developments in AI infrastructure. It anticipates that this strong foundation will give rise to revolutionary new AI applications.

Read more on Govindhtech.com

[thejestglobe]

Google Cloud lance la taxe furtive sur vos ébauches d'idées secrètes

https://thejestglobe.com/wp-content/uploads/2025/05/Google-Cloud-lance-la-taxe-furtive-sur-vos-ebauches-didees-secretes-file.webp.jpg-file.webp Google dévoile un Hyperordinateur si avancé qu’il commence à facturer les utilisateurs avant même qu’ils aient une idée La nouvelle prouesse technologique de Google Cloud Le géant du web vient d’annoncer la mise à jour la plus audacieuse de son IA Hypercomputer, intégrée à Google Cloud et censée révolutionner le développement de l’intelligence artificielle générative. Selon les responsables du projet, ce système repousse toutes les limites connues en étant capable de prélever un montant forfaitaire dès l’instant où un utilisateur envisage vaguement de penser à un concept novateur. Cette prouesse reposerait sur un algorithme prédictif si puissant qu’il peut anticiper toute illumination créative, la taxer à la source et régler le montant sur la carte bancaire du client avant même que celui-ci n’ait le temps de formuler son idée. Officiellement, Google explique vouloir « encourager l’inspiration tout en maximisant la productivité », ce qui impliquerait de monétiser les ébauches cérébrales pour qu’aucune fulgurance ne soit gâchée. Les premiers retours d’utilisateurs sont mitigés, certains saluant la performance, d’autres se faisant déjà débiter pour des inspirations qu’ils n’avaient qu’à moitié formulées. Une innovation saluée et redoutée par les experts Les observateurs du marché s’interrogent sur l’impact global de cette approche. Certains évoquent un bouleversement sans précédent, où la moindre diversion imaginative induirait une carte de crédit à découvert, tandis que d’autres y voient une aubaine pour stimuler le génie inventif mondial. « L’Hyperordinateur apporte une démocratisation de la pensée payante, et c’est peut-être ce dont l’humanité a besoin », affirme avec enthousiasme le professeur Lucien Clavier, spécialiste autoproclamé de l’innovation tarifée. Toutefois, de nombreux développeurs redoutent déjà une course effrénée à la rentabilité : il se murmure que la prochaine version de l’Hyperordinateur sera capable de créer et de breveter des idées à la place des usagers, tout en facturant des frais d’abonnement mensuels pour l’accès à sa propre créativité. Si tous s’accordent à dire que cette avancée fera date, il reste à savoir si les esprits novateurs du futur seront disposés à payer chaque fois qu’une étincelle surgit dans leur cerveau.

[usaii]

Google’s 7th Generation AI Chip ‘Ironwood’ Defeats TPU Ancestors | USAII®

Grab hottest innovations in AI chips and powerful AI models via Google’s Ironwood Tensor Processing Unit (TPU). Master latest AI skills to become an AI Prompt engineer.

Read more: https://shorturl.at/a1Xax

AI models, AI chips, AI development, Tensor Processing Unit (TPU), AI training, Ironwood, AI accelerator chip, AI Hypercomputer, AI Prompt Engineer, AI skills, Machine Learning Certifications, ML Engineer

[netherator]

it's a ref to hitchhiker's guide to the galaxy. where the answer to the ultimate question of life, the universe and everything. this answer comes from a hypercomputer that ran for thousands of years to come to this conclusion.

Im so fucking smart ask me anything

[hackernewsrobot]

Google Cloud Rapid Storage

https://cloud.google.com/blog/products/compute/whats-new-with-ai-hypercomputer