Unleash AI Potential with Cisco's Scalable GPU Server Solutions

last week I walked into the office of a CIO at a large Railway company former military guy gray hair big beard generally kind of an angry individual he looked at me and he said infrastructure for AI is kind of like riding a unicycle in a marathon and I said Gary what are you talking about and he said we've realized that it's a very long race but we're going to need to balance and adjust to line a business every single step of the way I looked at him and I said it's okay we got you we've got a solution for you and that's what Nan and I want to talk to you about today so first of all thank you for being here we appreciate the opportunity thank you for listening questions comments we it's been fun to sit in the back of the room and listen today um keep them coming that we we appreciate that um nant is our senior engineering product manager for our AI servers it's a pleasure to have him with us today we're we're very fortunate to do that uh my name is Jason McGee I'm part of our office of the CTO for our Computing business unit and uh think I have just a incredible job at Cisco get to come in and talk to customers a lot about where we're at from an AI perspective and there's all sorts of models that are out there we break it down there's really five big areas that we're focused on what makes Cisco unique is the ability to tie all of those layers together La our last tech Field Days SAA and tar came in and talked about some of our Solutions but the breadth and depth of Cisco is what allows us to take advantage of the high-powered AI servers that you heard about during the keynote and the ones we'll hear about a little later on today those Solutions can be broken down into some big buckets our Cisco validated designs or where we take our best practices our lab testing our certifications performance testing and we roll that into a user consumable document they've been around for years we do them across the company they're several hundred page documents but the really cool thing is we've made them as code so they're consumable through get repositories where customers are able to get those cvds and deploy that validated configuration or blueprint based upon the knowledge base of Cisco as a whole but we looked at that and we said you know what we could probably make that easier for our customers so we took and we put a product wrapper around our Cisco validated designs for AI and you may have heard this last fall we Ed our AI pods and what AI pods are is they're simply a purchasable Cisco validated design including the networking the Computing components the software components all those layers that we showed on the previous Slide the servers that nant and I are going to talk about today are what powers aipods using the information from our CBDs and if you want that available as a service like Dan just talked about we have hyper Fabric and so that's that glue that's the solutions that pulls AI together for us but it all starts with the fabric and H hyper fabric AI that portfolio is incredible piece of that so we're going to spend the rest of the time that we have together half an hour or so with some time for some Q&A we really focusing in on that next layer up is what does the compute form factor look like how do we power the Computing engine that's driving the workload that's generating all the traffic that's populating our networks how do we get the work done when we look across our our customers we've seen the hyperscalers some of the larg the large model trainers really be focused on the very the large language models when we get into Enterprise customers we're finding that your mileage is going to vary As you move across the AA landscape some of them are doing training which you see on the far left here all the way over they're adding their own data into it through retrieve logm generation they're doing light tuning they're doing retraining some of these existing models moving out into the inferencing space we're seeing massive growth in the inferencing space over the next several years uh and we think that's going to dwarf the training space as the industry moves toward more specialized models we're seeing the the large large models generate into smaller specialized models that are industry specific or use case specific that's the what we're building the platforms for to take advantage of you know where where the world is going from that AI Continuum if you look at the compute portfolio that we've been selling for over 15 years a phenomenal amount of capabilities in the AI space but it really fits in to that inferencing light training tuning part of the of the model and as a as a quick Glimpse if you look at what's there if you map the gpus to the the server platforms that we have available today there's a lot there don't try to consume it but the takeaway is that we can handle we think about 60% of the AI workloads that are running in Enterprise today with the gpus that are in our blade servers and our rack mount servers um quick refresher we call our current modular or blade based systems X Series we have a rackmount line of service that we call C Series just so that you and I know as you move across the Cisco landscape we've got all sorts of nomenclature and Landscape but the X series of the are the blades and that's what you see on the left hand side the existing one kind of the typical oneu two rack servers are what you see on the right hand side with a variety of gpus from different vendors that are available in them what we're going to be talking about the rest of the day is going to fall into that rightand side and it's going to fill out some of that Continuum that we looked at earlier we heard loud and clear from our customers from our partners even Cisco it internally you know one of our one of our biggest customers is that we needed something to fit into that training space and so we've introduced a large training server dense GP GPU server using the high-speed fabrics for GP GPU to GPU communication so if I'm running the Nvidia gpus inside the system then we use the Nvidia in NV link fabric for that fast fabric internal GPU to GPU communication if I put in the AMD gpus into it then we use the infinity fabric from AMD and Nan's going to show you what those architectures look like as we get into that and then very exciting you see with the the yellow blob onto it what G2 announced during the keynote today what we call our c845 a server is our dense GPU server but designed for flexible workloads a lot of our customers don't need the full 885 8 gpus high-end gpus they're either running an HPC environment today and they kind of want a crawl walk run into the AI space or you know they're getting into some of the smaller language models they're beginning to get into training and they need something they can stepwise grow into what we're hearing repeatedly is that in the Enterprise space a lot of the software teams are not ready to impl imp AI they're Gathering their data they're cleaning their data and and getting that ready to go and a lot of times they need they don't need to start with eight gpus they can start with two they can grow to four six eight depending on what they need and so that's the box that we're very excited to introduce today and we'll dig into a littleit little bit more details around that no matter what these servers are if I can't operate them in a simple easy manner then it's kind of pointless we're going we're make life harder for the end users and so what we've done with all the compute portfolios is when we look at our traditional compute whether that's the rack servers whether that's the blade servers we've managed those servers locally for years we call that very creatively our unified Computing System Manager well all of those servers in those Legacy environments take take advantage of that large number of gpus that we put up earlier we can also take and plug those existing servers into our as a service management architecture and what we call intersight that gives us capabilities to do some of the AI Ops functionality for our servers so we can look at proactive tack we can dismiss Parts before the the customer even knows that they fail um we can do over the horizon notifications for our customers kind of help them see around the corners uh whether that's a security advisory whether that's a vulnerability whether that's a failing Hardware comp component in the system using the metrics and the monitoring that we have on the on the back end of that system we're collecting a little over 13,000 data points a second on that platform an amazing amount of telemetry that we're using in the uh AI Ops functionality to help out our customers going forward what we're doing and we're talk about today is the AI servers that you see in the middle here will also connect those servers to the same operations platform and the takeaway is that when we move AI into the Enterprise space the governance the operation the tooling the apis that are already being used in Enterprise space can be used for AI I don't need a separate Silo I don't need a separate set of Stack tooling training I can use what I have in house today so the sovereignty the governance the best practice that our it has put in place we can take advantage of with that exist ing management operations infrastructure so with that we're we're going to dig deep into the 885 first and then we'll jump into the the newer box that was introduced during the keynote today so Nan I'll turn it over to you all right uh so I'm going to go into the details of uh UCS c85 a M8 uh it's a dense GPU server uh which was announced uh and late last year uh it was our newest addition to the Cisco UCS product family uh and it's designed for some of the most compute intensive and data intensive uh workloads including uh llm training fine-tuning uh deep learning uh and depending on the size of the model and uh number of concent users you want to support you can actually even use it for ragon INF fren kind of use cases uh as I mentioned earlier it's a DSE GPU server it can support uh eight of nvidia's hgx h100 or h200 gpus uh as well as amd's Mi 300X gpus and on the CPU side it's a a dual socket server uh with support for uh amd's fourth generation and fifth generation uh CPUs here is a bit of uh technical details uh it's uh it's an 8ru uh air cooled server uh and supporting as I mentioned earlier both fourth and fifth generation AMD uh epic uh uh uh CPUs but instead of uh supporting the entire set of skes from these Generations we have picked certain skes uh of CPUs which are more optimized for AI kind of workloads so there the focus is more on the the clock rate so basically if you see this uh table here it shows CPUs with a Max boost rate around 500 5 GHz say 4.3 GHz as well as 3.75 GHz so those are the CPUs optimized for AI workloads it has 24 ddr5 uh dim slots uh which can operate up to uh 6,000 megat transfer per second uh standard storage M boot drives uh 16 nvme for uh local data storage would be check pointing or any other purpose uh gpus already mentioned uh h100 and h200 from Nvidia sxm P factor and amd's m300x which is oam p Vector uh in terms of network cards so it does have quite a few PCI slots in there so these servers are sort of designed to be uh placed deployed in clusters the computer fabric is extremely important for these so we have one neck per GPU for that East West connection to deploy it in a cluster and then we have up to five Nicks which can be used for uh the front end Network or the north south connectivity I'll go to more details on that uh what we are supporting there right now is uh the connectx 7 or b314 the 400 gig NX or or super ni for the East firstest connectivity and 2x 200 gig uh uh NYX and uh dpus uh for north south and then of course there is an ocp card which can be used for host management uh in terms of cooling there are quite a few fans there 12 fans up front and there are four fans inside to cool the ssds and then uh power supply it's a power hungry server uh there are six uh three kilowatt psus operating in 4 plus2 redundancy mode for uh the gpus and there are two 2.7 Kow psus for the CPU

sled components I I have a question regarding this yeah Power hungry means also the gpus will produce more heat I assume yeah I see this on the gaming computer from my kids yeah they are always running hot so what do you see overall on the let's say cooling requirements compared to a normal server without all this GPU density is it double or do you have kind of a relation how much more heat they will produce U so so when you say in comparison to a regular server I mean that's sort of like subjective but these are designed to be kept in regular data centers it's air cooled I mean amount of heat dissipation uh and the number of fence it basically you can put in a regular data center right now it doesn't require need special cooling so it is not that uh let's say magnitudes higher it will still work where let's say systems without GP you have have that would be correct okay yeah yeah you don't need any special requirements to cool these I mean there are uh newer gpus coming uh on Nvidia and emd's road map that would probably require some level could cooling so but that that's coming that's next on the road map okay and and so here is an exploded view of the server as and as you can it's see it's a modular design on the top left uh you can see the GPU sled which is 4 Ru right below that is the CPU sled and then below that uh a layer of uh uh psus up front uh there are 12 fans and you but at the bottom you can see there those are the uh Drive slots and I of course I'll get into more details on the specifics of these components uh here's a rear view of that and in the middle there you can see those uh eight PCI slots for the East West connectivity a closer look at the rear view of the server so as I said earlier GPU TR on top uh CPU in the middle and then the psus now taking a closer the look at the GPU tray uh it as I mentioned earlier it has support for 700 wat h100 h200 sxm gpus from Nvidia and 750 wat m300x gpus from AMD and if you look at uh the the sort of like a logical block diagram of the GPU board here this is for NVIDIA you can see that there are eight gpus which are connected to four NV switches on that board through multiple NV links what it allows is actually is non-blocking communication highspeed communication between any pair of gpus here on the board and you can achieve up to 900 GB per second by Direction bandwidth and to provide connectivity to the outside world uh when again when you're putting it on a cluster uh these gpus are connected to an basically sort of like their own uh sort of like CX7 or supern via a set of PCI switches as you can see here so that's Nvidia GP board design uh quick question those super Nick are those Bluefield 3 these are BF3 b314 okay so this is Spectrum x no this is uh this is just the GPU board actually okay uh there is no Spectrum or or any external switch on this all of so up till this part this actually sits on the GP board and this is part of the CPU sled uh the PCI switch and whatever you see above that that's part of the CPU slide there is no external component this is all of it is on the server itself okay are you getting to the infin ban question eventually cuz well you knew that's where I was going yeah I mean that you know the blue filter reason that connects s support either or right all of our reference architectures are ethernet based that's where we see the industry going but recognize we do have some customers particularly the HPC that have an infiniband fabric so we will support infiniband connectivity on this it's not our primary use case or where we see the future going but as a transition period it's something that is supported with this box today well I mean I don't want to speak out of turn but I think Nvidia realizes that too which is why they put so much development into Spectrum X recently is because they realized that infiniband basically has a ceiling and Fe you're complete at this point and they're wanting to get people to migrate off of it as much as possible because especially if you're going to start uh deploying this in a cloud scenario clouds aren't not going to deploy infiniband if they don't have it because that's a cost that they don't want to incur exactly and we looked at a hard stance and we said let's just make it easy and acknowledge that there's some inventan out or there's a lot of inventan out there so you know we'll support if somebody wants to connect it that way and transition over with Nvidia and the rest of the industry fantastic okay thank you and on the next slide I here I have the AMD uh GPU board design so the the top half as as you can see is actually pretty similar to the what you saw on the previous slide but the bottom that's the uh AMD GPU connectivity so there are eight uh o form factor uh gpus from AMD Mi 300X in the scenario and they are all connected via this full mesh kind of network uh and this basically ens stores 128 GB per second uh bisection bandwidth and this of course is based on the infinity fabric from Infinity uh from MD uh closer look at the CPU tray uh uh support for as I mentioned earlier and amd's uh fourth and fifth gen uh CPUs and we have picked uh uh 9554 from the fourth gen and 9575 F uh CPUs from the fifth gen uh uh from the fifth gen of uh AMD CPUs uh and then for certain configurations where the customer does not require uh as much compute power uh or as many cores uh we also support for 9535 uh and then on the dim side we support three sizes currently 64 96 uh 128 GB DMS depending on uh what the requirements are and what gpus they are using uh they can pick uh appropriate uh GM now looking at more of like um closer look at at at the especially the networking components here at one you can see those eight PCI slots for uh the computer Fabric or the backend connectivity and as I mentioned earlier we currently support dx7s and uh b340 super mix at number two you can see uh it's a B3 to20 dpu from Nvidia we also have support for uh CX7 if you don't require uh dpu functionality any offloading and then uh but if you want dpu but then 400 gbps connectivity then there is support for b32 40s also at number three uh you can see the two psus these are 2.7 kilowatt and 1 plus one redundancy mode these power uh primarily the CPU slide components as well as uh the drives number four is the dccm which is data center secure control module and has basically it houses a BMC and then uh there's an RJ45 port for outof band management and the mini display port and a few USB ports uh for KVM connectivity just one question we had before the presentation for hypers sheet so on two you could run then ebpf yeah if we have a DPO networking card that would be supported yeah yes okay wonderful that's the super neck is the dpus the branding thing so we have both super neck and DP dpus I mean I think Nvidia uses ter Nvidia prefers the term super neck because they're because in because AMD calls them dpu and then Intel calls them ipus for some reason yeah and then uh finally at number five we have uh couple of RJ45 ports that's an Intel card ocp card which can be used for host management here so quite of few ports in terms of how the network connectivity we basically see four uh types of networks on these servers there is of course management Network which is out of band management as I mentioned on the previous slide there's front end Network which can be used to connect to the the broader uh data center it can be used for getting access to certain data for inferencing storage network of course to connect to the storage uh and then the most critical one I think in this scenario would be the the inter GPU backhand Network also known as compete fabric and as you can see like eight Nicks uh uh which will support this and to create a small cluster all you basically need is one uh 64 uh Port uh switch in the scenario it's Nexus 93 64d x2a and uh you can create a really small cluster of just eight servers using it uh but if you want to scale it let's say a larger cluster of 32 nodes then of course you need a fine Leaf kind of architecture uh the important thing here is that uh all of this network has to be uh non-blocking um nonsaturated so basically if you are connecting a 400 GPS uh neck that uh leae switch connectivity to the spine switch from the uh Leaf also has to be 400 GPS or higher do you guys support gigabit next and not not not yet not in this one yet uh as we move move to the Future generations of this we will have support for it power and cooling uh again it requires a lot of power to power all those gpus and uh CPUs and everything so uh it has six psus primarily for uh the gpus at the bottom there you can see those are three ,000 wat 80 plus CP uh psus running in 4 plus two rency mode and let's say if uh three of them actually uh become inactive then of course the server still operates but the GPU performance will be kept at 60% across all of them that's 12.5 Kow per server that is the maximum possible use power cons that we expect so that's basically primar used for sort of like a provisioning purpose but in normal scenarios I don't think you will see more like more than 8 to 9 kilowatt do you have like a recommended density that you're planning for in a rack so that you don't melt through the floor uh it depends on basically what power supply customer has but there is a customer that's trying to put four of these in one rack so I think they have uh 60 Kow power supply in that track and so they will put uh four of these so around 48 to 50 kilowatt will be for for these servers and then there is top of the r switch and other things in there currently air cooled correct it's all air cooled plans to make it water cooled uh that will be a future platform the only reason I ask is NVIDIA has Wells are water cooled because they're hot yeah yeah is the storage um accessible from all the servers um so server one can access the storage from server 2 as well so it's not the local storage uh if that's the question uh that's all local but yeah uh you can have a distributed storage software def storage uh connected to these but that's a different architecture of course there are okay so do you do that yourself or do you use other solutions for we are looking at certain Partners uh to support it I can think of one think you could all right so Hardware configurations so these are uh what we are doing here is we have created these sort of like optimized fixed configurations for these server so uh let's say if you're doing uh training uh and then you want to deploy it in a cloud-like environment then you will have certain specific requirements you might want a b31 40h as well as a B30 to20 but if you're doing inferencing you might not even require those eastwest necks right so for for different deployment scenarios we have created these fixed configurations which are optimized for again as I said like those specific scenarios so we have about 16 fixed configurations here uh using H2 200 and then there is for H 100 a few are for H 100 of course because you know we have H 200 now and then uh for mi30 about five or six of those and uh this is also used of course in hyper fabric cluster uh for AI uh you mention again so I think the key differentiator for the servers are these three things it it delivers unmatched performance it has this highspeed GPU interconnect on the on the server itself the board uh that has NV link NV switches and scalability uh those eight Nicks which allow it to be become part of a really large cluster and that helps us still over uh Ser basically all these use cases uh AI uh starting from geni training model training all the way to deep learning reinforcement learning and even some large model inferencing uh HPC quite a few of our customers actually deploying these in HPC uh uh environments and then you can also uh do some real time data processing which is through uh you know by accelerating through uh gpus uh so again with Jason has mentioned earlier not every customer requires such a powerful server uh for some of them they just starting on their AI Journey they require fewer gpus a smaller density so for that we announced uh C8 845a I it today and just will share more details on that wonderful thanks toan thanks appreciate the question anything else on the 885 The Big Box before we move on okay it's big it's powerful uh it'll it will do a tremendous amount I'm just trying to I had around 60 kilowatt in a single wreck like that could servers hazardous environment with a high voltage coming through don't don't look up tech specs too closely cuz Nvidia can get 60 kilowatts and a half and even that I mean and one of those Network diagrams was 32 of those servers put together um I think you know one of the big takeaways you know I started with Solutions and I and I really want to emphasize that that the validated design on how you put it together the aiod where I can order is one or I think the demo for hyper fabric is incredible because Dan started out with you know user guidance what cable to plug into what port if I'm paying that amount of money for those highend gpus I want to maximize performance um I mentioned Gary the CIO that I talked to last week we had a a another uh CIO that mentioned to to us he said he said my job is not to build the AI applications it's not to transform my company my job is to keep the gpus as busy as possible to fill them with data he's like that's where our investment's at and I need to ensure that I'm doing that and that's where those guided configurations the hyper fabric AI can come in and really benefit around that okay but the conversation we're having is a lot of customers are not ready for that massive massive box and you know we really think the 845 is going to fit a wonderful Niche architecture that's out there I say Niche that's a I shouldn't use that term but being able to start with two gpus grow an increments of two 2 4 68 of that and have a variety of gpus I can use the H1 100 the h200 kind of the top end from Nvidia today or I can use uh the l40 gpus for um that that are just fine for some of the workloads that are needed out there when you get an Outreach for something like this does that create like a Professional Services engagement so you can talk the customer through that kind of discussion or is this something that you're really only Contracting through other partners that have experience doing this deployment both so yeah we're doing a lot of that when I mentioned the aiod that comes with the service wrapper around it that that's a piece of that it's kind of interesting to know especially because of everything that's going on in in in Europe and um probably uh others will follow as well soon with ESG and everything that's going on yes making sure that you're going over the regulations that going on with a exactly yeah it's and even as we get into the switches the optic density you know it's it's going to go beyond just the GPU the server our entire data centers are you know are being affected um and we hear it every day but when you actually see it on paper with the specs it's it's mindboggling um you know what what we can do or you know what's out there and the amount of work that can be done from from this um with the I'm going to have a hard time not using the code name for this box but the 84 since it's just brand new today but the the 845 box it is a a a 2 CPU AMD epic based system I said it it can grow grow in the gpus what I do want to emphasize on this is that we're using the 885 we talked about before that's an hgx reference architecture from Nvidia this box is the mgx reference architecture from Nvidia um we looked at that um and there's some things with the offthe Shelf mgx specs that most of the mgx servers in the world um are doing that we thought we could do better there's some simple Cable Management cable routing items that we improved that drastically reduced the amount of air flow improve serviceability of of the Box uh one of the things that we did was that most of the time on mgx architecture all the power supplies are on one side of the machine that prevents the box from passing a drop test because it you know goes one side we evenly distribute the power supplies horizontally across the bottom of the box so we have to do something like a drop test for reliability again those things that are just standard for our Enterprise compute customers that they're you know they're used to using and have to have you know in that um one of the other things is that uh typically it takes about 38 to 40 screws to remove the gpus from an mgx design this box one screw so if I need to replace a GPU we made serviceability easy on it uh for the customer so you know really trying to bake in some of those things that our Cisco customers are are used to using in those systems you're only supporting the AMD CPUs you're not going to support the Intel CPUs this box is AMD CPUs only um we will support the Mi series of gpus in it from uh you know and then you know going forward I think we'll see a a broader variety of the vendor CPUs in the AI specific servers but just point in time where we're at today um it's what makes the most sense for you know for this box um the 885 that we looked at that is more of an appliance model where you buy a fixed config this box you have a have a lot more flexibility uh and how you can you know buy it out as far as the the AMD CPUs the storage uh gpus and then this the super Nix and the uh North South Nix that we have from Nvidia as well the connect x7s so we are announced the box today we anticipate April time frame is when begin shipping this box out to our servers or I mean to our customers but taking orders today and we do have it on the show floor uh if anybody wants to take a walk down and go take a look at it um key thing is it is obviously optimized for AI we have the capability to do some incredible workloads on it but it does a lot of other things as well so if I need to scale up into AI or I need to have a dual function server this fox has the capability to do that and I can't stress enough the importance of the consistent management so it plugs into the as a service based management platform that we have which you can run on Prim or can be consumed from the cloud of Cisco whichever is easiest for the for the customer but that inight platform allows that consistent management of it so it fits in into the overall what what what do you mean with a dual purpose server so if if I am maybe I have a traditional large database workload HPC environment um I also have the ability to run AI workloads on that box I could transform over the life cycle of that box you know or I could you know actually have multiple applications running out if I needed to yeah I would I would see that happening for um BDI environments hospitals doing uh a lot of these things so um yeah that's at our at our recent customer Advisory Board last fall we heard that loud and clear we we kind of came all in on hey here's this 8gpu server and they were kind of like hold off most of us are not quite ready for it yet um we need something that and we have a perception because AI is new and sexy everybody's talking about it that it's a thing by itself but there's never an AI system or an AI application that doesn't have a lot of other traditional servers wrapped around it so if you know especially if I'm putting my if I have a off-the-shelf model that I'm plugging my own data into with retrieval augmented generation I've got a vector day sitting out there that's on a server somewhere I've got traditional storage probably coupled along with the um you know vast or ddn or wcas or others of the world that that are wrapping that AI specific storage so it's it's an ecosystem and so that's where that Dual Purpose comes in especially the university uh hospitals doing all the hospital stuff during the day doing the research the yeah so yes pretty or just nor oldfashioned machine learning yeah absolutely right we don't see it in practice a lot but one of the unique things about the unified Computing system and I'm sure you're all you know very familiar with it but we have the concept of a service profile so when we deploy the the blade servers or traditional rack servers out we go in and we carve out what that server looks like in software how do I want it to be configured what bios do I want on that box uh and in our blade based system we've had some customers have actually convert that Pro file day and night where I can run an application during the day and research during the night something like that and so that's something that you know we'll be able to do in the future on on these boxes as well um kind of a corner case but but it's out there but I think you know the HPC space is a big place some of our traditional database uh I mean some our traditional workloads very large databases those type environments uh it's going to be a a really good fit we're see our customers grow into it and these will be part of the AI post validated design so they're not part of the aipods today but that is coming when so we're not going to start shipping these until April and so you know we anticipate the validate designs the aods and all of that will be available around that same time frame so you know we're working to get all that together okay question no okay I saw the mic I was um as we just kind of wrap up and look at these layers so we talked about the Cisco fabric hyper fabric we didn't go a lot of to store stage that's there but we know that's a very big piece we feel that we've got a full portfolio at that compute layer whether it is running inferencing on the CPU itself we have a radiology demo running downstairs in the Intel booth on doing uh AI inferencing looking at pneumonia images with just the Intel CPU no GPU all the way up to the large H GPU systems that we have and you know with that uh c845 you know fitting in here with the uh you know two all the way up to a GPU systems that are there um and then being able to wrap the solutions the validated designs the AI pods the hyper fabric around that so the Nvidia NV the pytorch framework whatever the customer needs to run these top layers comes along with that all in one package okay so it is a massive amount of complexity um but it's an exciting place it's a really really neat place to be and we're excited to have the portfolio built out

2025-02-24 12:58

Other news