Hi, this is Wayne again with a topic “THIS is how AMD Wins. – EPYC Genoa Announcement Reaction”.
While AMD endured a relatively lackluster ryzen 7000 launch, thanks to Fierce competition from everyone’s favorite budget, offering Intel in the data center AMD is pulling no punches. With today’s announcement of Epic, Genoa, they’re bringing double-digit IPC improvements, a maximum of 96 cores per socket, a whopping 12 channels of ddr5 memory, I don’t even have that many fingers and so much more we’re going to tell you what’s new and why Intel’s upcoming Sapphire Rapids, Probably won’t be able to compete with this spicy salami Mamma Mia, oh quasi, Picante, comic, questo, sponsor Segway Phil redox, build Redux makes it easy to configure your new build with support, guides and competitive pricing compared to building a PC yourself head to build redux.com, Linus and Start your new build today for those that haven’t been paying attention. Let’S get you up to speed. Amd has been releasing their epic line of server processors based on the Zen architecture since 2017, with the first generation dubbed Naples, not until the more recent epic Rome in 2019 in Milan in 2021, however, did AMD actually manage to compete and then beat Intel taking the Performance Crown this is in part due to core count: Intel’s monolithic style processors capped out at a measly 40 chorus per socket compared to amd’s 64 core chiplin designs back in 2021, Intel hoped to compete against epic Milan with their own chiplet-based design, codenamed Sapphire Rapids, but Ended up delaying due to technical difficulties and poor yields to a now confirmed January 2023, launch and they’re still on 10 nanometer or Intel 7. Now they call it Intel.
Seven, we don’t don’t say nanometer come on. Things are not looking great for Intel’s Enterprise sector. Right now, because honestly genoha looks awesome, the most obvious change with the latest generation is the New Zen 4 architecture running on tsmc’s five nanometer process. Zenfor is more of a refinement of Zen 3 rather than a complete redesign. But while much of the micro architecture layout is the same, improvements have been sprinkled all over a big chunk of the clean 14 IPC uplift comes from optimizations to the architecture front end with improved Branch prediction that can predict two branches per cycle rather than just one And a hefty bump to L2 cache, they also introduced AVX 512 support, which is going to be a huge help for machine learning and AI workloads in the data center, speaking of which the other major upgrade this time around is core counts across the 18 total skus. Launching for Genoa, the max cores per socket has increased by a whopping 50 from 64 to 96 cores, with the minimum core count, seeing a bump from 8 to 16 cores on the lowest end skus. Despite the powering efficiency improvements of the new architecture, maximum socket power consumption is jumping from 280 Watts on the last generation to a whopping, 400 watts per socket a level that amd’s high performance. Compute customers have actually been asking for, though the company tells us they don’t expect the power envelope to climb much Beyond this point and don’t expect to get much further into this video without a mention of our beanies on ltdstore.com uh.
They put LOL in there, but you know what it’s actually getting colder. Unfortunately, with great power draw comes great, socket demands and Genoa is no exception. Gone are the days of rectangular. Oppression say hello to our new Square overlords: Jokes Aside: genoa’s new sp5 socket packs a whopping 6096 pins, nice um more like a like a nice backwards.
It’S a similar trade design to the previous sp3 socket, but it’s now heatsink actuated rather than socket actuated. That’S fancy. Speak for the cooler providing the necessary pressure to ensure proper contact between the CPU and the socket’s pins. It’S not dissimilar to the approach Intel has been taking for a few years now, except AMD thoughtfully included the lockdown plate for the CPU to prevent you from accidentally dropping a chip into the socket which you wouldn’t want to do. Since 6096. Nice pins are tiny and expensive, however, they are not as expensive as the copious amounts of memory. This thing supports across 12 channels we’re talking up to six terabytes of ddr5 running at 4 800 Mega transfers per second, allowing for a theoretical Peak throughput, but a 480 gigabytes per second across the whole thing. Six terabytes all without taking too big of a hit in total latency, oh, and if that isn’t enough fast memory for you AMD, is introducing support for cxl wait.
What the hell is. Cxl cxl is short for compute express link and it could very well be a game changer in the service space and potentially for Consumer Tech in the future. It’S a cash coherent interconnect standard built on top of PCI Express 5 that allows for huge increases to. I o memory and storage flexibility in the same way that AMD uses PCI Express Lanes as Infinity Fabric and their dual socket servers.
Cxl uses PCI Express Lanes as low latency high bandwidth interconnects, allowing the system to pool or expand memory and improve the performance of various types of Hardware accelerators like gpus or network cards. Maybe you need an extra four terabytes of RAM for more virtualization. I know I do slap in a cxl memory card and you are good to go and cxl lets. You allocate that extra Ram to other devices if you want to Genoa, supports cxl 1.1 plus type 3, which is the catchiest name.
I’Ve read all week and is primarily focused on memory. Expansion rather than accelerators. Cxl is only just coming to the server Market, but is primed to be a very important technology for the future of our computers, so keep an eye on it. It might come to Consumer Hardware sooner than you think, if you’re worried that you won’t have enough pcie lanes for all this stuff, then don’t be.
A single CPU provides you with 128 PCI, Express Gen, 5 lanes and up to 160 in two CPU systems, providing ample flexibility in terms of expansion, and if you want to see just how much you can fit into these things, super micro, let us unveil their new H13, Genoa Suite check out the link in the description to see some Hands-On Hardware, content and last, but certainly not least, AMD introduced a few improvements to their security, such as support for over a thousand encrypted virtualizations. That is pretty neat, especially for vps’s. What kind of perform from all of these slick new feature enhancements, though, for starters more than double the performance in integer arithmetic, over Intel’s 40, core Xeon, 8380 Flagship, which is expected considering AMD, has more than double the cores and ends up being just an eight percent Increase per core, but the customer is buying 96 core CPUs like this one. They would probably actually just want the more cores rather than faster cores, so does amd’s dominance then carry over to the rest of the product stack according to AMD. Yes, very yes, but before we get too far, please note that AMD has been pretty liberal with the scaling in their performance graphs, so try to focus on the numbers themselves, rather than the pictures AMD claims that their 24 core epic 9274f will compete with that same 40, core Intel CPU, at least in integer operations, calling integer operations. General purpose Computing is a bit of a stretch, but even if you cynically assume that all of these CPUs will perform 25 percent worse than AMD promises, you’ll still see comparably cord AMD skus competing with Intel’s flagships.
Maybe you shouldn’t be quite so cynical because the way AMD tested their CPUs for this could be leaving performance on the table. Let me explain: EMD has embraced the Silicon Lottery with their new power management feature called determinism wow, just like gon na take a philosophy there. Well, it comes in two flavors performance and power.
Oh, it is philosophy, it’s kind of like amd’s Precision boost found on their ryzen chips and allows you to bump up your performance within spec. If you have the thermal or power Headroom to do so in the performance determinism mode, the CPU will Target the maximum performance at a given TDP with power varying to maintain stable performance power. Determinism is flipped with the CPU targeting the highest frequencies, while maintaining stable power drop, while a consumer might want more performance at all times.
Well, if you run servers that vary and load having a constant power draw, can actually improve efficiency and reduce stress on the components increasing the lifetime of your server. Now, back to the graphs, for these reported benchmarks, AMD didn’t use their determinism modes, which means that customers could end up seeing extra performance on their chips. Should they decide to take advantage of this feature and even without that extra Edge AMD, still claims massive per core increases over Intel’s lineup promising as much as twice the performance per core in high performance Computing, which makes them an obvious choice for simulation workloads AMD has Far faster cores, which helps them win an HPC and they have way more cores which gives them the upper hand, virtualization and hyperscaling. But here is the most damning comparison.
This performance per server chart AMD can fit the same number of cores in 60 percent. Fewer servers using just half the power, despite the extra socket draw at this point, Intel isn’t even competing they’re sitting on the bench, but all this performance of course means nothing if the price isn’t right is AMD, leveraging their performance leadership and charging through the nose. Well, their Flagship 9654 that I’m holding in my hand right here is coming in at effectively 12 000 US Dollars, which means the price per core, remains virtually identical to their Milan Flagship, but they are charging around a 10 premium on most of the products that compared To Zen 3, which is a bit of a bump, considering customers are likely going to need all new servers in addition to all new Chips, thanks to the socket change and probably also all new Ram yeah, it’s gon na get pricey, but with the ever increasing amount Of computing being done in the cloud sure that major players like AWS and Google will happily pay the premium or you know just fab their own CPUs like you do and what’s coming in the future. Well, we know that we’ll be seeing 3D V cash versions of these chips with the upcoming Genoa X lineup, and that the Bergamo lineup will max out at 128 cores on a single die while we’ve yet to see what kind of upgrades Intel will bring with sapphire Rapids, it’s clear that AMD isn’t resting on its Laurels, employing every trick in the book to maintain the massive growth they’ve seen in the past couple of years in the server space and with the ever-growing presence and threat of arm-based servers. It’S going to be tough for Intel to keep up now. Just make me some non-pro, Zen, 4 threadrippers and I’ll be really happy almost as happy as I am to tell you about our sponsor MSI.
The holidays are coming quickly, but MSI has you covered for gifts? This year, msi’s holiday deals are on now, where you can save on their 12th gen laptops, desktops, peripherals and components. Do the kids need a new laptop for school, or maybe your partner needs an upgrade to their outdated gaming rig, there’s something for everyone on sale now and for those of you in the US. There’S the hashtag, timsi holiday giveaway, tell MSI what you’re thankful for and enter to win gift cards apparel and more. So what are you waiting for check out, msi’s deals and the hashtag Ty MSI holiday giveaway at the link below to start saving today, thanks for watching guys, obviously we haven’t had a chance to put this guy through its Paces yet, but maybe if you want to See some epic action, we have our holy on the Epic Milan, which was wow 2021. How time flies .