Name: Announcing the Next Generation of Dell/NVIDIA Engineering Workstations for SOLIDWORKS Users
Uploaded: 2025-08-18T18:33:43.268Z
Duration: 57 min 58 s
Description: Announcing the Next Generation of Dell/NVIDIA Engineering Workstations for SOLIDWORKS Users

Transcript for "Announcing the Next Generation of Dell/NVIDIA Engineering Workstations for SOLIDWORKS Users": Good afternoon, everybody. Thank you for, joining us today. Today, we're gonna talk to, our partners with Dell and NVIDIA, taking a look at the, the new precision workstations as well as, the, the new, NVIDIA product line. And very happy today to have, Hamashi with us, from NVIDIA. And a little bit later on, we will have, Ken also join us, from Dell, to go through their points. Good morning afternoon, Hamashu. Yeah. Thank you, Brian. Yeah. Like I said, we're we're, we're always very happy to, get together and, and talk about everything, you guys have going on over there. And, we've all been watching the, all the latest news and Nvidia's been very active, right now. So we're, definitely Lot lot lot going on in the world of AI for sure. Yeah. Absolutely. Absolutely. And we're all waiting to see how it comes into, into play with, SolidWorks. So Yep. Happy to take you through a few slides, some of the new products that we introduced, and, ultimately, how they will benefit, SolidWorks users. Alright. So, my name is, Brian Paulk. I'm a senior applications engineer, at GoEngineer, and I spend a bunch of time, benchmarking, hardware, with Dell and NVIDIA. And I'll be going through some of that information, a little bit later on. As I mentioned, we have Hamashu Yair, from NVIDIA, And we'll bring on, Ken Flanagan, from Dell, to talk about the Dell products. And if I we always just like to, remind you guys of, of Go's in GoEngineer's mission, and that's to empower, the creation of today's products and tomorrow's innovation. There's nothing better than to do that with our partners here, with state of the art hardware, and that just makes getting our products, designed easier. So, everything with SolidWorks runs better, when we have the right hardware, and that's hopefully what we're gonna provide you with today. And with that, I'll hand it over to you, Himanshu. Yep. Cool. Thank you, Brian. My name is Himanshu Ayer. I lead industry marketing and strategy for the manufacturing industry at NVIDIA. And my background is is is mechanical engineering. I have worked for, SolidWorks in the past. Actually, that's how I started my career as an applications engineer, at SolidWorks and and, had several roles there before, you know, taking on other opportunities and such. So, thank you for for joining the call today. I would like to give you a quick background of, some of the hardware and software solutions that NVIDIA, has for the manufacturing industry, and then specifically, some new announcements and GPUs that are very well suited for SOLIDWORKS workflows. We can go to the next slide, Brian. Yep. Thank you. So for over twenty years, twenty five years now, NVIDIA's, professional visual computing solutions have helped drive, innovation for thousands, millions of users across many different industries, manufacturing industry being one of them. And and specifically for manufacturing, NVIDIA RTX, which will will will mainly cover today, has been, accelerating the time to market, for new products, by, you know, powering workflows and and use cases across photorealistic, visualization, design, simulation. And, over the last few years, you know, areas such as, digital twin and, obviously, AI, have been playing a role in this area. So we'll quickly, you know, take you through some of those. So RTX is is bringing all of this technology. RTX is essentially a technology platform, for all of these industries, you know, specifically here, manufacturing industry, but also other industries like, you know, scientific research and and data science, by, you know, making all of these technologies for deep learning, machine learning, AI, visualization, design, simulation, to to, you know, multiple platforms. So if you look at this slide, essentially, if we start at the bottom, these are the hardware platforms. These could be, RTX GPU powered, laptops, RTX GPUs in your workstations, desktops. And, ultimately, as we scale up, these same GPUs go into, data centers. You know, it could be a small data center, on prem data center to these massive, cloud data centers from these, service providers and such. So it is the most advanced visual computing platform available for professionals, and it enables customers or customers to take advantage of these technologies such as, you know, GPU accelerated rendering, design, simulation, data science, across, different types of computes, with with with machine learning and AI. So once we, you know, take a look at that hardware layer, over that is are are the different platforms that NVIDIA offers for AI development, for digital twin development, and and such. And on top of that are the libraries and, SDKs that are integrated into ISV application. So here, if we take a look at, SolidWorks, right, SolidWorks incorporates several of these NVIDIA libraries and toolkits to accelerate, the performance, of these applications. Desso systems also incorporates many of these libraries and toolkits in their broader three d experience, suite of applications, that, you know, that that our users are, you know, are working on on a day to day basis. So, essentially, it's it's it's it's a full stack solution. It's it's hardware plus software plus these libraries and toolkits that are essentially powering these workflows. Next slide, please. Yep. Thank you. So when I talk about the hardware layer of it, right, these hardwares, these the the the GPUs essentially, have seen phenomenal, gains in performance based on the different architectures that that that we have come up with. So, the latest architecture we announced earlier this year is the Blackwell architecture. So the GPUs essentially will be, you know, called the they will have reference to the Blackwell architecture. The one before that was the ADA architecture, and Ampere architecture. So, every architectural advancement, every new architecture will see the latest generation of these different types of cores. Right? These cores are are what accelerate these workflows, and and these are, CUDA course, Tensor course, RT course, and each course have a specific workflow attached to it. So for example, the tensor cores are really the ones that are powering all the AI workflows. The RT cores or the ray tracing cores are what are powering the visualization and ray tracing type of workflows. And CUDA course are essentially the ones that are powering all of these, parallel computing, type of workflows. So each generation, brings more capabilities, more power, to these different applications, to these different use cases. Next slide, Brian. So this slide kind of gives you, a look at the portfolio that we have right now, the portfolio of desktop GPUs. Obviously, there are there is a counterpart to this on the laptop GPUs, and then there are data center GPUs. But it is a very broad portfolio. If you start on the right, these are some of the entry level solutions like the RTX a 400 or the a 1,000. These are smaller work smaller GPUs ideal for, you know, lighter workloads, at at a lower price point. And then as we go further to the middle and to the left, sorry, to the to the left middle and to the right, the two that you see, the RTX Pro 2,000 and RTX Pro 4,000 SFF or small form factor GPUs. These were the ones that were announced last week at SIGGRAPH. So these are, brand new GPUs in the Blackwell, family of, GPUs, which are which we call as mainstream products. These are really, the most cost effective, GPUs, that are ideal for majority of everyday professional workflows, whether it whether it comes to design or or to visualization. These, fit into multiple, OEM systems, Dell workstations and such. They can fit fit in small form factor, workstations or desktops that are becoming quite common, but they can also fit into the more traditional larger, desktops, based on how, you know, the bracket that is used. And then as you go to RTX Pro 4,000, 4,500, 5,000, and above, these are really, the higher end GPUs that support, you know, extreme, you know, if you're working with very, very large models or doing real time simulation or or rendering and such, These are the GPUs that have that, right amount of VRAM, right amount of, core counts and such that that that are powering these, workflows. So, ideally, it's it's it's a broad, you know, portfolio of products. And, you know, based on your needs, based on your workflows, you can choose choose the right, GPU, for for your use case. Next slide, please. So, you know, when I mentioned, applications like SolidWorks, for example, we have been working with the SolidWorks development team for a long time. Right? Somewhere around 02/2019, 02/2020, the whole SolidWorks graphic graphics pipeline, went through, rework. It was significantly, revamped, and and and, most of the visualization workflow in in SolidWorks now goes through a a a GPU. So to be able, you know, to have that fluid experience while using SolidWorks when you're working with, large models or assemblies or, you know, creating photorealistic visualizations and such. An NVIDIA GPU is going to give you that performance, that you need. Right? You know, when you're doing just simple operations, the span, zoom, rotate, or when you're doing section views, and such, you don't have to, you know, cut down the or or or simplify the model or suppress parts and assemblies and such. You can load the model in in in full, fidelity and still be able to work with it without, you know, facing any any any issues or performance related issues and such. So, we do extensive benchmarking, like I mentioned. Right? Every new architecture, every new family of GPUs, that that that we release, we work with the SolidWorks team to do these benchmarks. So we see the performance, gen to gen performance, across these, different family of, GPUs. The latest being Blackwell architecture where you can see, you know, up to 10 x faster performance. And and Brian has done a lot more benchmarking as well, so he will go through some some more of this in in in details. Next slide, Brian. Yeah. So like I mentioned, right, we do a lot of benchmarking. We we work with the SolidWorks development team in terms of certifications and such. So we have created a solution overview of how SolidWorks performs on, these GPUs. So it it has benchmarks of both desktop GPUs as well as laptop GPUs. There is a link, there in the in the subheading of the slide. I think when you get the slides, you should be able to, get to the the solution overview, or you can just search, based on the title of the solution overview, and and you should be able to access it and download it from the NVIDIA website. Next slide, Brian. Again, just to, go back to that point that I mentioned, we work very closely with the SolidWorks development team to make sure that, SolidWorks users get the best performance, on NVIDIA GPUs. Some of you might have worked with Renan, from the SolidWorks team in the past. He has, you know, he he he's able to, you know, access and test SolidWorks visualization workflows and such, on on NVIDIA GPUs and and and has been, you know, a strong NVIDIA supporter. Next slide, Brian. So, just to go back to one of the GPUs that I mentioned specifically, right, the RTX pro 2,000, GPU. This was something that was announced just last week at SIGGRAPH. And, I think it is one of the best, GPUs for running SOLIDWORKS. It is at that right price and performance level where it can handle majority of SOLIDWORKS workflows even when it comes to, you know, large assemblies and rendering and visualization. You know, only if you are working with very, very large models and datasets and doing very high end photorealistic rendering, is when you will have to go up to 4,000 or 5,000. But for most, average SolidWorks, use cases and such, RTX pro 2,000 and RTX pro 4,000 are are very well suited. I'll just leave you with one, customer example if we go to the next slide, Brian. So, Gluckskin, some of you who had attended, SolidWorks World or three d experience world, they were one of the keynote speakers with Sujit on the main stage, I don't think last year, but the year before that. They are a Canadian company who make these, smart strollers. They are AI powered strollers. Pretty amazing, actually. It's essentially like a like a robot, that that you can control the motion, and such. So so they are developing and and they are a SolidWorks customer, obviously. So they are developing these, these strollers, parenting solutions, on these RTX 2,000 GPUs. And and not just SolidWorks. Right? Like I mentioned, this is, a very cool almost like a robot. So they do a lot of the robot training on the GPU as well, and more and more, some of their AI powered workflows are now running on these RTX Pro, GPU. So so a great example of, you know, a customer that that that that you might have seen at three d experience world who are also using these hardware and software solutions from NVIDIA in their product development workflows. Next slide. Yeah. Thank you. I will leave you with just two, new products or announcements that we have done, recently. On the on the left is, product called DGX Park. And this is like a a small form factor desktop, product, but it is essentially for doing AI development work. So for example, if you or your company is, you know, if you're taking some of these open source models from from from OpenAI or or Meta and such, and you want to do some fine tuning, of that for your specific use cases or workflows, but your regular laptop or desktop may not necessarily have that compute power, this is a solution that can provide you with that compute power. This is essentially an AI desktop, to augment your, you know, your other solutions. And and, it is, it is a combined CPU GPU board, ARM based CPU with a with a black wall GPU, up to a 128 gigabytes of memory, and it can easily support LMM models up to 200,000,000,000 parameters for for development, fine tuning, and such. And then on the right, what you see is the big brother of the DGX part, essentially, is the DGX station. Again, single, system, CPU, GPU system, with up to 700 and, 84 gigabytes. So it obviously can handle much, much larger models for AI development and and such. But just something that, you know, I I wanted to share with the with the with the folks here, to be aware of for as you are entering this era of AI, You know, it could be agentic AI. It could be your own models that you're working with. These are the solutions that can really help you with those workflows. And that's that's it. That's where I am going to stop my presentation. Feel free to send any questions in the chat. I will monitor it for a little bit. If not, feel free to reach out, to me via email. That's it. Thank you, Brian. Hey. Thank you very much, Masu. Yeah. I was definitely very excited to see last week's announcements, get that that 2,000 level, Blackwell out there. I think you're absolutely right. That's gonna be a a a perfect fit for for many many of our users. So I'm excited to, get my hands on it and and, see how how it stacks up. Yeah. You know. Very soon. Very soon. The availability will start, in fall, so you'll be able to get your hands on it very soon. Alright. I see just a couple, questions here. What do you know about the, let's see. What about FEM and CFD? Yeah. Are we there yet? What do you know, with the graph cards on the CFD? Yeah. So, over the last two or three years, many, many applications, specifically CFD applications, all run natively on GPUs. Right? Traditionally, these applications have been all CPU based, but now pretty much all applications have a GPU native solver. So, essentially, the entire solver will run on the GPU. And, obviously, the performance is is is much, much better. We have seen speed ups of 10 x, 20 x, even 50 x in in in in in many use cases. When it comes to SolidWorks specific applications, right, SolidWorks simulation or SolidWorks flow simulation, to my knowledge, they are still CPU based. They are still running Well, I believe that's correct. I believe that's correct. Now when we get into the platform applications, go into the Symblio tools. Yep. Those are, can definitely take advantage of the hardware. Absolutely. So so when it comes to the Desource Simulia family of products, both Simulia Abacus and, their CFD solutions like Powerflow, they can all run natively on the cheap GPU, which is going to give you those performance benefits as compared to the more traditional workflows. Yeah. Right. The traditional simulation tools inside SolidWorks still depends on the solver, and, you know, how much threading they can do. But they're they're still heavily CPU dependent at at this time. So so hopefully, hopefully, we'll continue to see that advance from from the r and d team. But Perfect. Great. We had a question here on, was interested in the, laptop, GPUs. Mhmm. What what is available? I know there are, what I'll show it in some of the Dell slides here. But, what are the I know there's some Blackwell cards coming out there as well. Yep. Yeah. So so there are, Blackwell GPUs available for laptop as well. And, more and more new products, more and more laptop GPUs are are, you know, will be available to the users through our OEMs like Dell. What I had on that slide was the desktop, family of GPUs, but there is an equivalent of the laptop, GPUs, as well. Right? So I imagine I imagine here that we're gonna see those, 2,000 level, cards coming coming, very soon in on the g laptop side as well. Yeah. The laptop GPUs fail the desktop GPUs, a little bit. So the laptop GPUs will come towards the end of the year or early next year, the equivalent versions of the laptop GPUs. Alright. Very good. Alright. The next question here is a question about what we recommend for for, assembly is about 5,000 components. I'm gonna table that for the moment. I'll get to that a little bit later because I have some data that it talks about that. There's a lot of factors that come in, but I think that's gonna I'll be able to address some of that in my, benchmarking data here. So Awesome. That's, great. If any other questions, feel free to send them to me. I will, turn off my camera and will be on mute. Yep. Thank you very much, Amashu. You're welcome. Alright. Unfortunately, Ken had some technical problems and was not able to join us, so I will continue on with our, Dell product line. Right now, Dell is, revamping and and moving us into the Dell Pro Max, product line. So there's a rebranding that is going on, from the Dell Precision products, that we're used to, into the new Dell Pro and Dell Pro Max, series. So, the Dell Pro computers are gonna be the more of your general, office, level computers. And for engineering, we wanna be in the Dell Pro Max, series. So in March, they announced this, and the equivalent towers. April, they started, introducing the same thing on the laptop line with the Dell Pro Max fourteen and sixteen inch. In July, they rolled out the 16 and the 18 inch, laptops, and that's kinda where we're at. Okay. So that's what we're we're looking at. We're looking at, cost effective systems, powered by, Intel and Nvidia, to give us a full range of solutions, that we can, use for SolidWorks on a daily basis. So that with everything from an entry level 14 inch, laptop, these are are real good here for, the mobility factor. Real compact. They support the, Core Ultra, seven, processors. We have now moved out of the, the Core I family, that that, wrapped up with the thirteenth and the fourteenth gen. And we've now moved into the latest generation of the core ultra processors. That can support a, up to a r t x 5,000 graphics card, two terabytes of storage, and up to 64 gigs of RAM. Many of our users, move up into our 16 inch. It's really kinda kinda fit right in between that, 15 and the old 17 inch. We've kinda, upped our sizes and gone now to a 16 and an 18 inch. So a lot of our users, kinda settling in on the 16 inch. The Pro Max, gives us the, Core Ultra nine, with 16 cores, up to the, RTX, 2,000 level graphics card, and, up to four terabytes of storage on there. When we jump into the Pro Max, 16 premium series, that's gonna allow us to get up into a little bit of the bigger graphics cards up to the 3,500, and, up to eight terabytes of storage. And then if we, really need those upper end or we get into things like, three d scanning and things like that, we can go into the plus level models, running the core ultra nine processors, but those can get us up into the RTX 5,000 level, graphics cards. So on the, tower side, kind of, revamping the entire product line once again on the on both sides of the products, with the Dell, Pro Micro, the slim, and the t two. So, anybody familiar with the, thirty six sixty series? That's gonna be equivalent to our t, t two tower. If you did, the smaller form factors, those are gonna be now the, slim or the micro. And then for our big applications, or high end SIM applications, Dell does have a full series, that can go up into the, higher multi core processors, or even go into a rack mounted system. Here are our our Dell Pro, Max Micro. The smallest form factor that we get, can run the Core Ultra five through the Core Ultra nine, and could still with that small form factor graphics card get us up to a, an RTX 4,000 ADA. We still get our 64 gigs of RAM on there. So if we're looking to, kinda reduce our footprint, so it gives us a real nice option that we can, can move to. The Dell Pro Max Slim, slightly larger machine, still can give us that, Core Ultra, nine, processor. And the, as well as the 4,000, level graphics card. So very two, two machines, very similar. This one, the slim model can give you, up to a 128 gigs of RAM of the DDR five. And, so just a little bit more capacity, on that that motherboard. The standard, Dell Pro Max, t two tower, This has been a, very common, family of of products, for Dell, going from the, Core Ultra five, to the Core Ultra nine, or even the nine, ultra nine k, the higher version chips. We can support all the way up to the, RTX 6,000 level graphics cards, and a 128, gigs of, DDR five RAM. So the most versatility, most flexibility, in this system. The 58, the precision fifty eight sixty tower, still has has not, switched over to the new generation of of products. So that's still available. These run on these Xeon processors. Those give us the higher core counts. For workloads, we typically get into those conversations when we're dealing with, a lot of the simulation type applications. Seventy eight seventy five, has the latest rise and thread ripper, up to 96 cores. And then you have the, Dell Precision seventy nine sixty tower, which is the equivalent with the, Xeon processors. And up to up to four terabytes of memory, if needed. So, most most of our customers, most SOLIDWORKS customers, are are on the left end of this where we're working with the, the Dell Pro Max, micro slim or the t two. So if you have a high end application, specialty application, be sure to reach out to us to see if one of these, larger models, is a better fit, for your application. Alright. So I'm gonna move into some of the, the benchmarking that I've been working on with GoEngineer. I've been working on this for for about the last seven years. So I've been, pretty extensively, running through our own tests. And and we've developed our own process, and we narrow in on the four main areas, and we take a look at the graphic cards, the hard drives, the RAM, and the processor. Those are the biggest things that we can control, in our system. So it's in as, Hamas, you mentioned, in 2019, SOLIDWORKS introduced the enhanced graphics pipeline, and you see it in the system options as the enhanced graphics. That was that was when beta started. And what we did was we opened up, the graphics pipeline on the back end of the software and we push more data over to the GPU. And in the early days, in those early versions, there was a performance hit and we studied that extensively, with the enhanced, enhanced graphics on versus off. In 2024, we have now added, the silhouette edges. So we can do accelerated silhouette edges now on those graphic cards. And so now that adds another level because we test everything, in all of these scenarios. So that takes me up to about 42 different settings that I run through, in our testing to try to isolate and understand, what is happening, in SOLIDWORKS. I work with a host of models. We range everywhere from, about 1,500 components up to almost 90,000 components, and we try to cover the spectrum, with everything in between. So the first model I work with is a, customer called Chip Star. This assembly, came in. It was, it's about 1,500 components, 480, unique, 1,600 bodies. And it's it comes in with about 5,000,000 graphic triangles. The, next data that we're gonna take a look at is the Olmax water jet, CNC. This machine is about, this comes in about 4,700 almost 4,800 components. About 768 unique and about 5,300 bodies. This has, about seven seven point two million graphic triangles on this one. We then go to the CFH, the Canadian France Hawaiian Telescope. This gets us up to about 7,500 components, 1,500 unique, but it's got 22,000 bodies in here. So a lot of imported, multibody geometry, and this takes our graph trial count up to 28,300,000 trials. My next model is the Racine Railroad, rail car. This one, is the the, model that started, our whole benchmarking process, all the way back in about 02/2010. And we've done a lot of white papers and and best practices off this doc this data. So, but to make it a a large model, we stuck two two train cars in there. Gets us up to about 20,000 components. There's about 1,900, unique and about 30,000 bodies in here. And a lot of these models, these are customer datasets, that we've, obtained over the years, and these, by all means, are not perfect models. So this one takes us up to about 22,800,000 graph triangles. Alright. So the graph triangle, the top level graph triangle is a, a feature that we can now get in SOLIDWORKS 2025. So in our, evaluation report, we now can capture the, top level triangle. So so that's kind of a newer piece of information that we've been able to get, in the recent, version. When I take these models and I look at the rail car, so that's about, 10,000 components, 15,000 bodies, using the default settings right out of the box. So that has large assembly mode on, lightweight mode on. We're using shaded without edges, and the enhanced graphics on. With the enhanced graphics on, we can see here, how the cards keep improving and the larger cards, are able to keep, producing the higher frame rates. Okay? When the models are smaller, the the the frame rates, are pretty pretty similar across the board. We don't see a lot of variation. We don't see the variation until we get into larger datasets. When we start really pushing the datasets up, that's where we start to see a a big variation. And every time, we add in, the new generation of cards, they fall right in top, they fall right in line, keep improving and pushing these, these limits. When we look at visualize, so this is a graph of, rendering, on the different cards. And so the lower the score, the faster the render. And what this really shows us was the transition from the older generation cards, the Maxwell and the Pascal series, when we moved into the RTX series, that ray tracing, that Himanshu talked about. Being able to push that ray tracing onto those video cards, use that compute power made a significant impact on digitalize. And the larger we go with the, the graph card, the more CUDA cores we have, the more processing power we have in that graph card, the faster we can render. So we're almost moved off of the Pascal and the Maxwell cards. There's still a few of those kinda lingering around. The the p series cards are still supported by SolidWorks, but they will be the next ones that kinda move off the list. Everything now is is these RTX cards. So all the cards that that most of us are running today, can definitely handle, some visualized rendering. So we've been taking a look trying to understand, the performance and we did a study and we're taking a look at, a gen over gen, the a 2,000, versus the 2,000 ADA. So the same version of SolidWorks, just one generation newer with graphics card. So this is, taking a look at our first model, 1,500 components. And what we're seeing is about a 20 to 25%, improvement in the, shaded and shaded without, edges modes, And about a 10 to 10 to twenty, ten to 30% increase, in the hidden lines removed hidden lines visible, and wireframe modes. Now some of that comes in from the addition of the silhouette edge, function. So we're getting, the silhouette edge option is really targeted to help those hidden lines removed, hidden lines visible, and wireframe modes. Looking at a slightly larger model, still looking at the a 2,000 versus the 2,000 ADA. When we get up to about 5,000 components, we're seeing about a, five to 10% increase, in a shaded or shaded with edges mode. But we saw a very, significant jump, in the in lines, removed in lines visible, modes and the wireframe was kind of a wash. But very nice gen over gen, improvements. So every time we get a new card, we are seeing, that continual improvement that, Hamashi is referring to. So the next question is to take a look at a 2,000 ADA versus a 4,000 ADA. And how much improvement are we gonna see, going up a card? And, again, I'm using this hours twenty twenty four, hands graphics on, silhouette edge is on. And we're getting about a 25% improvement, or increase in frames per second. Okay? So we're about a 25% increase in the shade and shade with edges, and we got, a 70% increase in the in lines removed in lines visible. If I take a look at, larger model, so this is the double rail car, 20,000 components, 30,000 bodies. I got a 40 about a 40 to 50% improvement. So we're starting to cross over, and one of the things I'm looking for and I'm trending here is trying to figure out if we can really start to determine when do we cross over, limits. When when does a when should we move from a 2,000 to a 4,000 or 2,000 to 3,500 or anything like that, and what kind of ranges. But definitely, once we start getting up into these fifteen, twenty, 30,000 component assemblies, there's there's a convincing argument, to start looking at these higher higher level cards like the 4,000 series. And I just, finished taking up some, some of the latest data I've collected, looking at the 4,500 ADA versus the 4,000. And I'm seeing an an additional, 20, percent improvement, in the frames per second over a 4,000 level card. And we're seeing, about, almost 20% across the board. You know, 17, 17 to 20%, in the 4,000 4,500 over a 4,000 level card. So definitely, if the, if the models, the right models are there, have the right demand for it, we can definitely see, some nice improvements there. I always talked a lot about the enhanced graphics, and I've worked a lot with the, the r and d team and the manager over there. And the r and d team has done a big push and big cleanup of the enhanced graphics. So if you tried this, in 2019, 2020, weren't happy with the performance, definitely go back and turn it on. They've cleaned that and they've converged the performance. The only place I'm seeing it is maybe in large assembly mode when we're turning things off, but, that's kind of the one spot where where we're still seeing a little bit of discrepancies. But otherwise, it doesn't matter whether I have it off, have it on, have it on with the silhouette edges. We're not trading off a big performance hit. What you are getting is much better pan, zoom, rotate and it really is more effective when we get into the larger models. So the larger the model gets, the bigger and more important this gets. If you're working on small assemblies, you're probably not gonna notice much difference whether it's on, or off, because the data just doesn't demand, and push the system that much. Performance, looking at, the laptops. This looking at a seventy seven sixty, which has an I seven eleven eight fifty h versus the precision fifty six eighty with the I seven thirteen seven hundred, h versus the seventy six eighty with the third, I seven thirteen eight fifty h. Okay. So, eleventh gen processor versus two of the newer thirteenth gen processors. K. We can see that, we get there is about a 25% improvement in performance going from the eleventh gen to the twelfth gen architecture. And that's illustrated there by that that step in there. So, getting off the older architecture, we got a surprising jump in performance, getting into the twelfth gen architecture. So 12, thirteenth, fourteenth gen, gave us a very nice boost in performance. This is looking at a 5,000 level component. We can still see where the, seventy six eighty, slightly larger performance, gives us a it gives us a more significant drop in that performance. So the lower, the score, the, faster the benchmark, overall. But, it's how they stack up side by side. I'm going up to 7,000 components to 20,000 bodies. This is again, the larger the model gets, the more significant that, drop appears. So the three columns on the right side of each stack is the precision seventy six eighty. And that that step does, indicate about a 25 to 30% improvement in performance. We can see the same thing looking at the, the towers looking at a precision thirty six fifty with the I seven ten seven hundred k, the tenth gen processor, versus the Dell Precision thirty six sixty with the, twelfth gen, I seven twelve seven hundred. Very clearly here, that step, we can see that jump for going from the tenth tenth or eleventh gen, that tenth to the twelfth gen architecture. Again, moving up, to a slightly larger model, 5,000 components. Still holding true. And, basically, now it's, I can tell you that if you're still on any of the older architecture, ninth, tenth, eleventh gen, Intel processors, there is a significant improvement in performance, getting up into the new architecture. So, there's a good ROI on that investment and it should should definitely be something that we're considering, if not today in the near future. So again, the the larger these larger these models get, the the more distinct, that that performance separation, is. So the other question I'd I commonly get is the comparison of a tower versus a laptop. Today's laptops are mighty powerful, And what this shows is the two, two groupings on the left are the tenth and the twelfth gen tower, and that, step is that, 2025% improvement. The three on the right are the two thirteenth gen and the eleventh gen, and we can see that they kind of fall in between, those processors. So we can see that we get nearly the same, performance. We can always push the tower a little bit harder. We can cool it a little bit better, so you can get a little more performance when it comes to a tower. But switching to laptops, we're not giving up significant performance like we did in the past. So there's not that trade off. So I know a lot of us, want that flexibility, want that mobility. So the, the laptops perform nearly nearly as well as as many towers. So we're we're definitely happy to see that. So here's another look, the the latest, Dell thirty six eighty. This is the fourteenth gen processor. Looking at the improvements of SOLIDWORKS 24 versus 25 on the same system. This is looking at the, 1,500 component assembly. The bottom chart is looking at the 5,000 component assembly. So the green lines are twenty twenty four. The blue lines are twenty twenty five. So even gen over gen in SolidWorks, we're seeing a nice significant improvement in our our performance. And then if we compare, the three generations of the Dell tower, the thirty six fifty versus the thirty six sixty versus the thirty six eighty. Our blue line represents the thirty six fifty with a tenth gen processor. The orange line represents the, twelfth gen processor with the I seven twelve seven hundred, and the gray line at the bottom, represents the fourteenth gen processor, the I nine fourteen nine hundred k. So again, the lower the score, the faster, the the benchmark completion of the benchmark. Alright. So we're seeing that significant improvement, in gen over gen towers. K. So takeaway, CPU is still king. Alright. If we're looking for performance, it's clock speed. SOLIDWORKS is still single threaded, so most operations are gonna perform on one core of your CPU. So we want that fastest core speed possible. And we look at that boost speed because that boost speed is the fastest that processor can run. Alright. It is time to consider thirteenth, fourteenth or even the new core ultras. So if you're looking at today's lineup of, systems, it's gonna be very hard even now to find the thirteenth or fourteenth gen. We are gonna be into the core ultra, the seven or the nine. Not all CPUs are equal. So do your homework. Take a look. I like to, do a Google search on the, on the CPU model, and look at the stats. They have different power consumption, different boost speeds. Just because they're a thirteenth or a fourteenth or same gen, doesn't mean they have the same output. K. If you're working on large models, enhanced graphics is a must. Turn it on. Use it. You're gonna enjoy it. For those of you that might have turned off large assembly mode, didn't like it in the past, 2024 changed how it behaves. Go back and try it again. Everything in 2024 kinda auto resolves. One of the big things user didn't like was the manual toggling of the lightweight mode. So we've gotta change how that behaves. It's all kinda seamless. As you drill down and and look at products in the tree, it's resolving that data in the background. Much more, user friendly, and not so manual. Alright. So with that, I'll tell you, I'll be happy to look through and see if we have any more questions. I appreciate everybody's time today. If you have any questions, or wanna have a specific conversation of how, how these new products, fit into your needs, please reach out to me. I'm more than happy to, go through this in detail with you. The one question I still see is, the recommendation for a 5,000 phone assembly. So 5,000 components, a 2,000 level graphics card, can should typically be able to handle that. Still handle that fairly well. There's there's always a lot of questions, on the modeling side. There's there's modeling practices, image quality issues, components, and things like that that we can, may also be able to add to the the overall performance of the assembly, whether using, Speedpack and Simplified Configs. So, performance is always kinda two sided. There's hardware and there's modeling. And when we pair up both of those together, that's when you can get, significant gains. Once again, I do wanna thank you, all for joining us today. And please feel free to reach out to GoEngineer if you have any questions, on, any of these Dell.