AMD Ryzen 5000 series processors

Four CPUs are launched recently. The Ryzen 5950X, Ryzen 9 5900X, and therefore the Ryzen 7 5800X. And wait, where's our 5600X? Read below and get you answer.

Our test setup may be a bit more extensive because we're packing in all four new CPUs, four of their predecessors and three team blue CPUs (INTEL) to mark this potentially momentous occasion. Getting right to business. in comparison against our Ryzen 3000 Series chips, it's clear that in games, there's a serious performance advantage with the 5000 Series. In fact, that advantage tends to hover around 20%, extremely on the brink of AMD's claimed 19% instructions per clock performance uplift. And this is often especially true in notoriously CPU bound games like Microsoft trainer 2020. Same goes for GTA V, CSGO.  Even Civilization VI has turn times that are roughly five seconds faster. Doesn't matter what game it's , the Ryzen 5000 Series is significantly faster compared to the last gen. But none of this is often a surprise. I mean, obviously AMD wouldn't release a replacement product if it had been worse than their old one. So then, how does Ryzen 5000 stack up against INTEL? Even the Ryzen 5 5600X, a CPU that costs 300 U.S. dollars beat every single one among Intel's CPUs, more often than it lost. And where it did lose, it had been within a couple of percentage points. These are all games that are traditionally CPU bound  would you check out that performance in CSGO? that's over 200 more frames per second. Only the 5600X, the lowest within the new lineup, falls below 600 points. And for reference, a Corei9-10900K scores a measly 535. AMD continues to flex everywhere not only their previous generation but team blue's entire lineup, with the Ryzen 7 5800X meeting or beating Intel's Core i9, even in multi-threaded workloads like POV-Ray, our Mozilla Firefox compile test and Blender. And even that lead quickly evaporated once we introduced the value competitor Ryzen 9 5900X to the combination , to mention nothing of the 16-core 5950X. Intel just cannot catch an opportunity here. And if you were wondering whether their upcoming Rocket Lake CPUs are getting to help. Well, after doing some rough work, if we assume Rocket Lake's performance are going to be between 10 and 20% above 10th Gen, that basically brings us back to established order . So assuming that Intel can keep core clocks as high as they're now, the best case scenario is that Intel will get back their gaming performance throne but because they're losing two cores on the highest end skew, meaning that anywhere that thread count may be a factor, Intel's claimed IPC uplift is unlikely to form up the difference. So at a minimum then, AMD has the gaming crown for a few of months and can still absolutely pulverize Intel in productivity, even after the 11th Gen launch. I mean, this is often something that we've not seen in overflow a decade now. Clearly then, surely this comes at some kind of major cost in power consumption or thermal output, NO.  Under an equivalent Blender workload, we are bang on between all of our 3000 and 5000 Series CPUs, with none of the pairs appreciably hotter than the opposite . Not only that, the 5000Series CPUs maintained roughly an equivalent or even higher core clocks throughout the test. meaning the improvements in performance mostly came straight from improvements in core architecture. Power consumption bizarrely seems to be actually down a touch for Ryzen 5000, with each CPU drawing at the worst, the same, or at the best , a touch bit less juice from the socket than 3000 Series. which means we can corroborate AMD's claims that they're literally not drawing such a lot as one watt more. This really is a fantastic achievement. As for a way they came , AMD claims that each one of this was done through pure engineering. and there is tons to speak about here, starting with the demise of the compute complex as we all know it. Now they still call them CCXs, but they are not partitioned into groups of 4 cores anymore. and each eight-core die is now ready to access its full level three cache without having to affect the additional latency involved in splitting it up. This is often an enormous part of what gives AMD such a huge boost in performance, because as we've recognized by now, Zen loves fast memory and Zen 3 is not any exception. So having a unified pool of cache memory means Ryzen 5000 can more effectively feed all the cores all the time. and that is not all. AMD tweaked their branch predictor and their operation queue. and therefore the TLDR of that's that the CPU can do more per clock and there is less of a penalty for it having to copy and redo work if it mispredicts something. and beyond just Ryzen 5000 itself. AMD's newly launched Radeon 6000 Series GPUs, are ready to work with these CPUs with a feature that they're calling Smart Access Memory, which can allow a Ryzen CPU to access all 16 gigs of a Radeon 6000 Series GPU's memory. In some cases providing a serious speed up. As for compatibility, AMD says that Ryzen 5000 CPUs are going to be compatible with motherboards using 400 Series chipsets also because the recommended500 Series chipsets. you will need to wait though for a BIOS update so as to use one of these new Zen 3 CPUs in an older board. And unfortunately for early adopters and thrift seekers, 300 Series motherboards are officially unsupported. Early 1st Gen Ryzen boards had some issues and that they can't guarantee anything at now and admittedly , AMD has still done pretty okay with intergenerational compatibility for the foremost part, especially compared to Intel's hard and fast two generations of CPU. Bottom line then, if Ryzen3000 was Ryzen matured, Ryzen 5000 may be a continuation of that. I mean, reviewers experienced no weird stability issues even right at launch. And it also marks AMD's returned to the upper echelon of desktop CPU performance. Intel against this , outright missed the vacation season this year. And if AMD is in a position to further refine subsequent big Ryzen CPU, for TSMC's shiny new five nanometer process, Intel's holiday next year, isn't gonna look far better . Like, to be clear, it's not all about nanometers. Nvidia famously squeezed a full generation of additional performance without changing process nodes once they released Maxwell in 2014. And, AMD, just did it on TSMC's seven nanometer. But Rocket Lake is gonna be pushing Intel's geriatric 14 nanometer node to its freaking limit unless something really surprising happens. 

So, it is a big win, here's. Once you buy new stuff and it gets replaced by something way better a brief time later, because Intel and AMD's leapfrog game looks to be heating up, not slowing down. I just hope that AMD's supply can continue with the demand.