New AI Supercomputer Outperforms NVIDIA

[@AnastasiInTech]

Let me know what you think!

[@CircuitSageMatheus]

Have you ever thought of creating a community for hardware engineers?

[@InfinitelyCurious]

Can you dive into the superconducting elements added to these advanced technologies(ex.Niobium)

[@dchdch8290]

@@CircuitSageMatheusI believe this is one, and we are part of it ;)

[@kevinsho2601]

Please do a vid on the company out of dubai who is creating medical AGI and what kinda of technology they are using and what they plan to do. Medical AGI is very broad. Would love to see what that means.

[@mariosebok]

What is the cost per performance comparison?

[@CircuitSageMatheus]

Awesome content, nowadays is very difficult to find channels rich in information like yours! Cheers to you for a job well done! 👏

[@DihelsonMendonca]

Linus tech tips is all about it, with excellent up-to-date articles on top of the edge technology, and many other excellent channels, BTW. It's just a tip. Perhaps you are not aware of. Good luck. 🎉❤

[@univera1111]

Truly excellent content

[@andrewwong1146]

Don't waste your time with this video. The CCP propaganda got a Russian girl to spew all kinds of shit and nonsense!

[@tanzeelrahman7835]

Your content is always very special and informative. You tend to choose topics that are not commonly found on other channels. The most important thing is the way you explain complex concepts so easily; that's truly awesome.

[@andrewwong1146]

Don't waste your time with this video. The CCP propaganda got a Russian girl to spew all kinds of shit and nonsense!

[@jp7585]

It seems like an apples to oranges comparison. Put it against GH200 Superpod with 256 Grace Hopper Superchips. That is Nvidias latest offering. It's not only fast, but energy efficient.

[@635574]

12x the gains. How the hell is nobody talking about it?

[@W1ldTangent]

@@635574 VHS vs betamax, Bluray vs HD-DVD... the latter was better in both cases and still lost because they couldn't get adoption. Nvidia gave away a lot of very expensive silicon for nothing in some cases or a small pittance to get CUDA in the hands of research teams at universities, who standardized on it, and eventually started teaching it. I love the idea of a competitor but they won't have an easy road if they're not willing to give away a lot of compute, and unlike Nvidia they don't have the gamers and crypto addicts buying every graphics GPU they could get their hands on for double MSRP to bankroll it.

[@waterflowzz]

Nvidia fanboy spotted. Power efficiency is a nonissue when you’re talking about the most powerful compute power cuz most people won’t have access to this power until way later.

[@nicknorthcutt7680]

​@@waterflowzzexactly, power efficiency is not the main issue 🤦

[@Leptospirosi]

You don't get the point: as the complexity of the problems you gave to feed on the AI grows these two systems stop scaling together. 90% of the raw costs is on people working on the project, so having a system that does NOT requires more work at all when your workload increases by orders of magnitudes, is a not brainer. You can start training your AI system months before you will on any Nvidia systems. The only thing Nvidia has on its side right now is the shear mass of chips produced each month, so I gues you can build a GH200 ai much faster then you can on Cerebras: not cheaper, but faster, despite being way behind in practicality and raw results.

[@pbananaandfriends]

The sheer compute power of this chip are promising a new era in AI technology. I’m eager to see how this will be utilized in various applications. Kudos to the team behind this innovation!

[@christophermullins7163]

Aliens going to start taking our AI computers like theyve been taking the nukes to protect us?

[@perc-ai]

at this rate bitcoin will be susceptible to a 51% attack lol thats so much power

[@JohnSmith-ut5th]

Not really, but there are a lot of investors that are going to making a killing shorting Cerebras.

[@zool201975]

yeah dont go counting the benefits just yet.. that is a chip that draws in a 100 fucking megawat hour. that thing cant run for more then moments withouth having parts of it being vaporized into heat. with computing almost ALL of the enrgy goes into heat so that is a bloody 99 megawat heater the size of a chesboard you got there.... you litteraly need a powerplant to run this crazy thing.

[@zool201975]

why would they need either? and with the power consumption of these things we do not need nukes to bloody glass the planet lol

[@RocktCityTim]

If you ignore politics and AI conspiracies, it's a great time to be alive! Thank you for sharing these positive breakthroughs.

[@JMeyer-qj1pv]

The bane of wafer scale computing has always been that some percentage of the wafer will have defects and be unusable. Does Cerebras has some way around that problem? There was a famous attempt at this back in the 80's and the company couldn't solve the problem and went bankrupt (Trilogy Systems).

[@prashanthb6521]

The final wafer processor always is quoted after taking into account the dysfunctional parts of that wafer. Meaning its always assumed to lose some parts to imperfections.

[@noobynoob138]

There's redundancy built into the chip, so a wafer defect only reduces performance instead of disabling the entire chip.

[@lllllMlllll]

Anastasi is such an amazing person

[@unkind6070]

Agree 😊❤

[@blackterminal]

You mean you have a crush.

[@hermanjohnson9180]

Indeed.

[@mefobills279]

Eugenics is good. Breed the superior specimens.

[@Human_01]

Lol.

[@Bassotronics]

Garfield was being nice to Odie when he was constantly trying to send him to Abu Dahbi.

[@meateaw]

lol, "we built a chip thats 50-60 times faster than a nvidia h100! its also 50-60 times bigger than a nvidia h100 ;) so maybe we should be comparing it to 50-60 H100's ? I also wonder how their chip yields work with manufacturing. With a smaller chip you can tolerate failures by throwing out the failued unit. But a single failure in your wafer scale chip wastes the entire wafer right? This is hardly the solution to availability issues. I don't doubt the value, but it hardly democratises this as your intro implied it could. ("Available right now!" - I'd love to buy a cerebras, but I don't think they'll be selling chips to me any more than Nvidia is selling me H100s)

[@aseeldee.1965]

This is very cool! Thank you for keeping us up to date with the AI evolution!

[@SocialMediaSoup]

I have no idea what she is talking about, but I keep watching her videos.

[@YouHaventSeenMeRight]

While this sounds incredible, I was a bit taken back by the unsupported claims that the CEO made. Is their supercomputer x50-x60 or even x200 times faster than a single A100? I want to see real comparable figures detailing the configuration of their solution because only then will we be able to determine the true level of performance that their solution brings. I've seen marketing material from Intel for their AI solutions where they made claims to be faster than NVIDIA, but when looking at their comparisons they were comparing their solutions against 2 generations older NVIDIA hardware. So I don't trust these sort of marketing stories. While building wafer scale processors seems like a great idea, the reality of process yields is that not all of these processors will come out of production functioning, which will drive up costs. The advantages are of course that you can pack more processing capability in a smaller form factor. Each A100 comes with external components that take up space, which also requires a system that is larger to house these multitudes of A100's. The latency issues that you mention that hamper the A100's will also hamper Cerebrus. If data needs to be moved between the wafer scale processors or even the geographically spread out Galaxy supercomputers they will incur latency penalties. There is no way around that. The advantage that NVIDIA has is that their A100 based supercomputers are not limited to AI work, they can also run general High Performance Compute workloads. No word on whether Cerebras's solution can do the same. Don't get me wrong, I see that this is quite an amazing achievement, but there are both technical and economical challenges associated with their solution, and I'm always skeptical when company CEO's start promoting their solutions. I've seen too many presentations where they push the most optimistic view of their product in order to wow the audience.

[@ZoOnTheYT]

Another awesome video Ana! Doing direct interviews is a great addition to your repertoire. I have an interest in AI as a social science person. A lot of videos either go way beyond my ability to comprehend, or are filled with superfluous information just to fill time. You consistently put out interesting and coherent information, that I also trust is valid, because of your background.

[@solidreactor]

I am very interested in Cerebras and Tenstorrent, where they seem to be the most viable alternative to Nvidia, both being companies that makes AI chip that is very scalable. The interesting differentiation between Cerebras and Tenstorrent is that Cerebras started with big chips working their way down (in a sense with enabling PyTorch compatibility) while Tenstorrent works from small chips and evolutionary works their way up. It's interesting to see these different contrasting startup philosophies work in the same industry having basically the same main competitors. Hope to see you cover these two companies in future videos.

[@geekinasuit8333]

Actually the most viable alternative to Nvidia right now is AMD's MI series of processors. The MI300 series is due to be widely available in 2024, and it will probably beat the H100 in terms of performance and flexibility. The research I've done indicates that Cerebras and Tenstorrent are very distant alternatives at this point in time relative to both Nvidia and AMD. There's also Intel with their Gaudi series, where it fits in comparatively is probably along with Cerebras and Tens, the most worrisome aspect being the longevity of the roadmap, Intel has been cutting product lines over the last for years. As we know anything can change quickly since the AI sector is in very early stages, so it's worth looking at all the players including the current batch of underdogs.

[@BienestarMutuo]

@@geekinasuit8333 We agree, if cerebras can lower her prices by 10 can be in the competition if not AMD will be the best alternative. 1 cerebras power computation = 50 nvidia power computation, but for the price of 1 cerebras (2.000.000 $) = you can buy 10 nvidia DGX (200.000 $ , 8 x a100 (10.000 $) ), in price nvidia win. And take in consideration that nvidia is expensive, very expensive. cerebras need to lower her price 4x to be competitive, 10x if want to be competitor.

[@robertweekes5783]

I have a feeling an aggressive AI takeover will come out of left field, some small company using a good amount of state-of-the-art tech 🤖

[@norik1616]

What is cost per TFLOP? Power per TFLOP? Is it 64 wafers each 50x the power of A100 all taking 1.75 MW? If so, they'll be taking 10 % more power than the 500W NVIDIA A100 (64x50 A100s).

[@RalphDratman]

What exactly does Feldman mean by "gradients" in the context of what is transmitted between geographically remote clusters?

[@Raymond-rr5iv]

What an interesting channel and a fabulous, clear presentation on this groundbreaking million dollar AI hardware that will facilitate probably the unimaginable in the near future. Anastasi and her co-host deliver a vivid picture what the company of Nvidia created. It is very exciting ... seeing the future unfold in this enormous leap foward. It tickles me to think that gamers... with their need for the fastest speed ten or fifteen years ago who were willing to pay top dollar to get what they wanted would create a niche market to spawn the likes of this billion plus computer chip made by Ceberas is only the size of an average floor tile, but more powerful than anyting known. This feeling of excitement seems like to me what it must have been to see the Wright brothers flying across the New York City skies for the first time. The significance of this chip is as unknown yet greatly anticipated to become probably the biggest scientific tsunami that will change our civilized world as we know it. Amazing development to learn about and thank you for your excellent presentation.

[@junpengqiu4054]

wonderful video, did some research about cerebras innovative and found out they really have done different and valuable things. "wafer scale engine" is what cerebras been known for, unlike traditional GPU, it is produced on an entire wafer. Conventionally, multiple cpu or gpu are 'printed' by EV on a single wafer, and later processes will cut them off the wafer. Therefore, one reason cerebras is delivering much better performance is because its 'GPU' is bigger. But this also leads to one problem: its even harder to produce than NVDIA GPUs, wafer often comes with defects, individual defected chips from conventional manufacture technique can be discarded. However, cerabras wafer scale engine needs the whole wafer to have no defects. In addtion, heat dissipation, even powering across whole surface are big challenges. Right now, cerebras is cheaper because it's not yet that popular, once market sees advantages from their super computer, their price can go higher than h100 since they are really difficult to make under current tech level.

[@florianhofmann7553]

With a core that size aren't the yields extremely low or is it even possible as there is always an error on the whole wafer? Or do the cores have some sort of fault tolerance built in like deactivating the affected sections?

[@Deciheximal]

It's the fault tolerance thing, it's the only way they can make it work on waferscale with all the defects.

[@woolfel]

it's great to see so many people and companies working on AI hardware, but without a full software stack, it won't be a credible competitor to NVidia. As ML technology advances, they'll have to make sure their compiler handles the workload scheduling efficiently. That's not an easy task.

[@rilwanj]

What if they made their hardware compatible with the Nvidia software? I think in this video it was mentioned that existing tensorflow code for cuda can also work on their hardware.

[@kleanthisgroutides7100]

My issue with Cerebas was the moment they said thier wafers had zero defects... this irritated me more than anything since by design, logic and physics that can't be true. Also they have repeatedly refused to release any further information... which seems like a red flag to me personally. This in all gives me memories of the early Rambus.

[@anonymousjoe3576]

I'm not sure where you heard that, but that's not correct. By design, their Wafer Scale chips are able to isolate and bypass any defective areas. A defective area in other company's chips invalidates the entire chip.

[@kleanthisgroutides7100]

@anonymousjoe3576 that's still a defect, you've lost some area and the same technique is used everywhere else... its pretty simple to slice off areas of SRAM, interconnect, cores etc... those with many defects simply fall into lower product tiers. They were specific in the interview which could of been by Anandtech... hence why I remembered the quote... "100% yield", I lost interest at that point. Just because you haven't cut up the wafer doesn't mean your yield is 100%, they are bending the metric and not accounting for the dead area... (which they don't disclose as a % obviously).

[@anonymousjoe3576]

A scholarly article that I read stated that when the Wafer Scale Engine did have a defective area, it was typically less than 1-2%. Give me a day or two and I will try to find the article.

[@mathiasjacquelin2146]

I think what is meant by that is that every wafer will offer the same compute capabilities / processing power. Internally a wafer may have defects but the user will not suffer from them, meaning that the wafer is still "usable", hence the yield. This is not necessarily doable for smaller chips.

[@PalimpsestProd]

I'd love to see a breakdown and compare of this tech against Dojo. Code scaling, Watt's per output unit, data types, and flexibility.

[@Wirmish]

... and cost.

[@MegaWilderness]

Dojo is a dead duck

[@JohnSmith-ut5th]

You're comparing new Cerebras technology to old Nvidia technology? That hardly seems fair. I doubt Cerebras will be able to catch up to Nvidia, let alone surpass it. This is because Nvidia technology is actually optimal for ML.

[@ok373737]

Nvidia H100 and GH200 are more versatile and general-purpose systems that can handle a wide range of AI workloads. Cerebras AI supercomputer is not suitable for all types of AI workloads, such as inference or data analytics, which may require more flexibility and scalability.

[@markissboi3583]

Every year there's always some faster tech comming out around 2025 we'll see Home pc's get sorted bulky power hungry GPUSs aint the way

[@thedeadbatterydepot]

Smaller isn't always better. I theorized such a computer with the whole wafer, the whole complier part was out of my skills, parallel data bus would be the only way. They have achieved the removal of the 2 compiler stages to get to machine language, the single stage compiler with whole wafer design has Nvidia beat, for much cheaper for title of most powerful AI. Dude knows what he has, I will seek to buy one of his systems, for a upcoming product. Thank you great video!

[@unkind6070]

It's really crazy as much as it is amazing it's kinda scary 😅❤

[@617steve7]

Anastasia In Tech my engineering crush!( Not to be confused with my academic crush, Sabine Hossenfelder)Exceptional content! keep them coming!

[@zandrrlife]

I've been selfishly hoping this company would stay a hidden gem 😂😂. Superior compute in terms of training models and on-premise inference. SUPERIOR.

[@pensiveintrovert4318]

No one cares about your supercomputer. Give us a cheap, fast, large ram, low power GPU to plug it into our own computers and then I'll be impressed.

[@Number6_]

1.75 mega watts.. thats like £400 per hour to run. Just for the electric ! This will upset a lot of people.

[@gregbarber8166]

Hi Anastasiia The're girls half as pretty as you making millions a year with half of what you have naturally please your a smart girl check it out once you have money like that you can live and do anything you want Travel the world Girl. I only tell you this because you are so nice and your intellect is amazing the WORLD is your play ground my dear take care Gb

[@craighutchinson1087]

I guessed the company accurately before listening to video. Tech tech potato youtube channel had some good content on this waffersized chip Your video was very well presented

[@das2003]

This is all just noise. No matter how good they are or what promises they make, Nvidia and their customers are the only ones who will see it's benefit, as it will create competition for Nvidia and give alternatives to the market. Until this company goes public, it's just noise.

[@vast634]

The REAL bottleneck in chip supply is ASML. They make tools to build the chips.

[@danielmurogonzalez1911]

I am a simple man, video I see from Anastasi, video I like.

[@Shuttterbugg]

Sorry bur this is apple's and oranges ur comparing a company that makes gpus and a company that makes super computers..Nvidia yes is getting into the business but this is a horrible comparison

[@windmillfire]

Thanks for making these videos 😀

[@HonestyLies]

very interesting, I wonder how saleable they are for production, honestly seems like companies will be fighting for these limited quantity high speed chips, surprised ive never heard of them! Great vid

[@riaanvanjaarsveldt922]

Spelling mistake, "bootleneck" 😂 Maybe get your AI to check it

[@matias.bevilacqua]

You've been talking about Cerebras for ages. A really interesting video would be for you to explain why they seem to go nowhere.

[@Deciheximal]

I would imagine that it's more difficult to make it fault tolerant for waferscale than they let on, and it's also harder to program software that computes over such a wide area.

[@abrahamsatinger265]

Where's the memristor component of the tech? And, that waste heat, recapture it, and shunt it back into the system or store it? Already paid for the cpus, and no bonuses or support or synergy or adaptability with the memristor tech, no or yes?

[@MrWingman2009]

Maybe I'm not looking hard enough, but this is the only place I've found good, well summarized info on AI hardware progress. Thanks Anastasi! 😊

[@andrewwong1146]

Don't waste your time with this video. The CCP propaganda got a Russian girl to spew all kinds of shit and nonsense!

[@signupisannoying]

It's not cost-effective. If one tiny part of the wafer get damaged, the whole wafer is defective.

[@andreasschaetze2930]

I remember a time where my uncle as an engineer got a PC with 40MB storage and I wondered how he would ever fill that much space. Today I need that space for one single digital raw photo 😅 It’s amazing how fast and capable hardware and software (some not so much 😂) has become

[@federicomasetti8809]

1998, my first computer (well, the "family computer", because they were very expensive, for what they could do): Pentium II at 300mHz, 32megabytes of RAM and I think something like 500megabytes of hard drive, but I'm not sure about this. Of course with floppy disk and cd drives, in that distinctive "greyed white" of the time. I was 13 back then and it feels like another era in the history of humanity 😅😂

[@Bobby.Kristensen]

Great video! Thanks!

[@lehsu]

AI race is heating up. Competition is great for innovation.

[@happyandhealthy888]

Andreas + Nastja = Anna

[@D3adP00I]

I wan't a girl that smiles at me the way you smile at chips.

[@jonmichaelgalindo]

Data quality matters vastly more than parameter count though. Improving LLMs and Stable Diffusion right now is all about figuring out how to get better data.

[@Eugbreeze1]

Good info 👍 I was able to get some shares as ipo .......

[@wilgarcia1]

I was wondering if anyone was going to try to build a CHip from the whole wafer.

[@marktahu2932]

Very interesting and shows a broader view than just the Nvidia or AMD approach. Mind boggling how fast and how far this work is going.

[@geekinasuit8333]

There's not a lot of information about Cerebras, so thanks for making this video. I'd like to know how flexible a machine like this is with experimenting with different models? Will you be limited to only a few kinds of models, if so then what exactly are those limitations? One known issue that Cerebras acknowledges as an intentional trade off, is that a machine like this is limited with floating point accuracy and will not be suitable for models that require higher 64bit precision. It appears the machine is optimized for 16bit precision only. I expect there will be other limitations besides the FB accuracy and a summary of what those limitations and what the tradeoffs are (pros and cons) will be nice to know about.

[@OriginalRaveParty]

Fascinating

[@teekanne15]

I am curious how this will play out. At first glance it sound like those "new battery 10x more volumetric capacity" click baits. While these numbers can be claimed under certain controlled environment it is questionable if these can be reproduced in real world application. Also the world of technology is operating within the world of business where longterm contracts and other strings make switching supplier a rarity,

[@goodtothinkwith]

How does it behaving like a single GPU compare to NVidia’s GH200?

[@panama-canada]

Everyone is chasing smaller and smaller chips. And size doesn't matter in supercomputers. We can go large. No miniaturization required. It's not a laptop LOLs.

[@Steamrick]

I can't get any real use out of the 'waferscale vs A100' comparison... sure, the Cerebras superchip is many times faster, but a server rack full of nvidia GPUs is going to have quite a number of chips inside. As such, a 'single rack' comparison would really interest me in terms of compute power, investment cost and energy consumption.

[@mathiasjacquelin2146]

One thing to keep in mind when assembling a server rack full of GPUs is that you will be far from having perfect strong scaling. In other words, if the Cerebras CS-2 is about 200x faster than a single A100, 200 A100 will still be slower as they will not get perfect strong scalability.

[@VaibhavPatil-rx7pc]

Excellent information

[@dchdch8290]

Wow , nice summary. I was actually wondering how they utilise all those wafer scale engines. Now it is clear. Thank you !

[@billpage7438]

OMFG...WHAT A GODDESS YOU ARE !!!

[@amj2048]

36 Exaflop ... 😲

[@Rhomagus]

Cerebras: AI supercomputer networked across three of the same type Cerberus: Three headed hound that guards the gate to Hades ... just in case you may have been confused. Don't be.

[@HMexperience]

Just one thing. A100 is 312 tflops and H100 is 4000 tflops so about 13x faster not 2x as you say in video. Otherwise great video. Thanks 🙏

[@dannyboy4940]

I am astonished to see that beauty and science can coexist

[@Arthur-ue5vz]

Thank you for doing these videos and helping the rest of us to see what's going on in the world of AI and computing in general. I appreciate your efforts 😊

[@KingTubeAR]

Do you think this will be successful enough to fill the supply gap in the ai gpu market? It would be amazing because ai startups are starting to buy gaming GPUs which are not desined with enough VRam that it takes to work on ai

[@rewirestrike]

Wow

[@MichaelLloydMobile]

OMG... I had to watch this video because your introductory image is adorable!

[@chillcopyrightfreemusic]

Fantastic video I just subscribed. Mr. Feldman was speaking my mind when addressing the tokenization of the arabic language. I don't speak arabic sadly but have been trying to find good models to handle it and found that only gpt4 and bloom were decent. I think his company is on to something forging connections to the gulf. Great video thank you!

[@georgelionon9050]

its not the size that matters...

[@donaldedward4329]

This video is one year old. Nothing new since. Absolute Vaporware. NVIDIA RULES!

[@Patrick1985McMahon]

I could see Nvidia making a dedicated AI PCIE X16 card. In the future you would be not only upgrading a GPU but an AGI Card too. Thankfully Many boards have multiple PCIE X16 slots.

[@heberje]

Compelling offerings

[@donaldedward4329]

A 12" by 12" chip ! WOW! Wikipedia: The etymology of Cerberus' name is uncertain. Farce.

[@MarienFournier]

You de a wonderful job. Thank you very much for your outstanding content

[@mariussefu3]

is this a reupload?

[@johnbollenbacher6715]

The success of this product requires that a large amount of AI be performed in the cloud. Is this how we expect it to go AI progressively at the edge. Is this how we expect things to go? I’m not sure.

[@zeroonetime]

Collective intelligence Ci I.S. marching on, redeeming humanity's. Ni ~ Ai = Ci ~ 010

[@horikatanifuji5038]

I was born too soon... it will probably take roughly 20 years before these chips become consumer grade priced... I'm... sad...

[@slob12]

Cya nvidia!😊......nah...but great to see competition!!!....we must not forget that amd is also increasingly concentrating on AI 😊

[@HenryCalderonJr]

Love that everything you post you have a great explanation and always back up your information with real facts as a document video! Thank you 😊 your awesome! Your brilliance in awesome

[@ingemar_von_zweigbergk]

if cerebras force nvidia to make bigger gpu chips for supercomputers then would that make gaming gpu and cpu chips bigger too

[@u9Nails]

Whoa! This is awesome! Always brilliant content. Love this channel! Learning new words, like Wafer-scale, is eye opening!

[@willykang1293]

Thank you for your deeper introduction on Cereras!!! I won’t know this despite I stayed around Fremont and Santa Clara last month if I didn’t get into it much deeper…😄

[@512Squared]

bootleneck = bottleneck

[@GeinponemYT]

Never heard of this channel before, until it just popped up on my homepage. And I'm glad it did: great, clear information, with appropriate graphics (when needed), very in depth, but still understandable. One minor piece of constructive feedback: maybe tweak your audio settings a bit to decrease the harsh 's' sounds. I'm using heaphones, and your 's'-es are a bit uncomfortable. Otherwise: great video!

[@AnastasiInTech]

Thank you! Noted

[@brianrich8974]

How much power is used? Wow? 1.75MW. Have any nuclear plants near by to power that?

[@layt01]

Just to summerize, 56x larger chip (full wafer) is 200x faster than a single "small" Nvidia.

[@StarGateSG7]

20 YottaFLOPS at 128-bits wide! Runs multiple instances of real-time electro-chemical physics-based simulations of human neural tissue to get working WBE (Whole Brain Emulation)-based AGI systems! Yaaay! Canada Wins! V

[@g.s.3389]

1 cerebras is 1.3 million USD, 1 H100 = 20kUSD you can buy 65 H100 for 1 chip where you have to add all the rest.... the 100x performance is not yet justified... on Nvidia you have a standard, competence on the network infrastructure infiniband and so on... on the other side you have something interesting, but too new. we will see in next months how it will evolve...

[@MrJGxx]

"The first question that comes to mind is, latency".....Uh...no.. the first question that comes to mind is, Cerebras vs skynet. Which is better?

[@360TainanTw]

OK, so this ai chip is just Copy but slower than Tesla's DOJO AI chip xD OK, nice, good for this new arriving chip "maker"...

[@federicomasetti8809]

I just wonder why Nvidia keeps calling his professional products GPUs, if what they do is nothing "graphic". Isn't it misleading? Or is it part of a marketing strategy, to enflate the price of consumers' products? Just wondering, no "hate" intended 🤔 P.S. (and off topic): has "CIAO" become international? 😅

[@Sven_Dongle]

First off, CUDA isnt just for "parallel GPU computation". It's an entire software ecosystem that includes deep learning, linear algebra, sparse matrix, random number, image processing, FFT, profilers, etc. Secondly; PyTorch isnt CUDA. It's a higher level of abstraction that may use CUDA to run on NVidia GPU's. Thirdly; who is your audience? most of us could scrape together 7 grand for an A100 if you're really serious about ML, but how many are going to cough up 2 million or so for a proprietary system then plug it into your 1.75 MW outlet to run it? No, we are going to use NVidia card(s) on COTS hardware or scale a bunch with NVLink and not be too concerned how cool it is to put a bunch of processors on one giant wafer.

[@mathiasjacquelin2146]

You could also rent some time on a larger scale machine (regardless of the vendor) on some cloud computing platform

[@MrWilderNapalm]

Tesla just announced 100 ExaFLOPS by October of 2024 from its DoJo system.

[@happyandhealthy888]

ANdreas+NAstja=Anna