With the recent launch and imminent release of the new AMD Ryzen 5000 CPUs, we have seen an uplift in some AMD fans unsure as to the higher pricing vs the Ryzen 3000 series and whether it is competitive.
The initial launch shows that AMD is unwilling to step on their own toes, knowing that the older Zen 2 chips still hold a lot of value, with the latest CPUs pricing aimed at new consumers, rather than those looking to upgrade their 3000 series processor. AMD will undoubtedly launch more value-focused SKUs in the near future, possibly to compete with Intel’s next launch but this is speculation at this point.
So really the question isn’t whether you should upgrade your Ryzen 3000 series processor, as that makes little sense, and instead, we should assess whether or not new consumers should be buying the latest Ryzen 5000 vs 3000.
Before we delve into any specific differences behind these processors let us take a glance at the Ryzen 5000 and 3000 series CPU SKUs as advertised.
When we line up the Ryzen 5000 and 3000 SKUs in order of price we start to see the full Zen 2 and Zen 3 stack covering almost every price point and performance need. We are unsure if AMD plans to continue making Ryzen 3000 processors but, in any case, the price of the Ryzen 5000 processors will probably remain a premium until the Zen 2 stock has gone.
The table below shows the stack of Zen 2 and Zen 3 CPUs covering most key price points.
“Zen” is the term used by AMD to designate the different generations of their CPU architecture since 2017: with “Zen” being the first, superseded by “Zen+” in 2018, “Zen 2” in 2019, and “Zen 3” in 2020, so, confusingly, Zen 3 is actually the 4th generation of architectural upgrades over this period.
Both Zen and Zen+ used a 14nm process node, whereas Zen 2 and now the new Zen 3 is based on a 7nm process. Generally, the smaller the process, the more transistors can fit into the same amount of silicon on the CPU, meaning greater processing power. AMD’s stated aim has been to leapfrog reductions in process size for every other generation, with each generation in the middle being a refinement of the preceding generation’s technology.
Many of the performance increases in the new Zen 3 can be attributed to the changes in the cache layout. Improvements in manufacturing mean that AMD has been able to give eight CPU cores access to one large 32MB L3 memory cache, rather than splitting it into two lots of 16MB with four CPU cores each. This is an especially powerful improvement for gaming as most games rely on the memory subsystems within the CPU quite heavily – much of the data required to maintain a ‘world’ within a game, with all its physical laws and persistent objects – is held here as a ‘world thread’. With these improvements, any one of the cores can directly access any part of data stored in the single large memory pool, rather than the data having to traverse across the ‘Infinity fabric’ in between the separate cache segments and the cores having to communicate with each other to access these separate cache segments. This means a greater speed of process resulting in less latency for games.
The higher-end enthusiast processors for the 5000 series will have two of these 32MB layouts on the chipset, with eight cores each (essentially a doubling of the image on the right) but this is an improvement over the similar level chips for the 3000 series that effectively had 4x16MB cache sections, with four cores each.
If the claimed performance increases between Zen 2 and 3 are accurate, this is exciting for computing tech enthusiasts, as it points to an alternative avenue of performance enhancement that AMD has tapped into which they can then apply to the next Zen 4 generation which AMD has confirmed will be using a 5nm process.
Looking at the clock speed difference between Zen 2 and 3, we can see that the given base clock speed is actually higher in the 3000 series than the 5000 series in some instances. This may seem counterintuitive at first considering the improvements in performance that AMD is claiming, however just as important to performance is the number of Instructions Per Clock (IPC), and here, if AMD’s numbers are to be believed, the 5000 series excels.
Essentially the improvements in architecture detailed above mean that the amount of data being processed by the new series, per clock, is greater than the 3000 series, so the amount of Instructions the new CPUs can crunch over any given amount of time is greater.
The naming conventions of the new Ryzen 5000 series, of course, follow on from the Ryzen 3000 and are considered a direct replacement for new consumers.
For those that are unaware, Ryzen 9 is the designation for AMD’s ‘enthusiast’ level processors – those aimed at high-end gamers and workstation users. Below these come the Ryzen 7 ‘performance’ models, also targeted towards gamers and workstation users but with slightly less money to burn or simply lower requirements. The next lowest are the Ryzen 5 ‘mainstream’ processors, then the Ryzen 3 ‘entry-level’ in descending order of both processing power and price.
For the purposes of this article, we will be looking at the Ryzen 5, Ryzen 7, and Ryzen 9 offerings for both the 5000 and 3000 series, which are the ones of biggest interest to gamers.
It is worth quickly noting that until we get our hands on the new Ryzen 5000 processors we are unable to truly test the performance at the moment. There is already extensive data on the Ryzen 3000 processors but we did get a little insight at the launch.
While there was no indication of what the Ryzen 5 5600X and Ryzen 5800X can do, the Ryzen 9 5950X stole the show and was touted by AMD as the “best of both worlds” CPU.
This mammoth 16 core/ 32 thread processor provided the highest cinebench score of any desktop CPU with 640. The Ryzen 5950X was also shown to smash the 3950X with heavy workloads, with a 27% increase in performance for CAD, 9% compiling, and a 5% uplift in video editing performance too.
AMD also gave us a glimpse into the gaming potential of the Ryzen 9 5950X, showing a 5% increase over Intels i9-10900K in Shadow Of The Tomb Raider and Total War: Three Kingdoms.
As soon as the Ryzen 5000 series processors are released we are going to be benchmarking each SKU across a multitude of games and benchmarking software to bring you the comparative data.
AMD has sought to maximize efficiencies for systems that combine both the newly-released Ryzen 5000 processors and newly announced Radeon RX 6000 series GPUs with something called Smart Access Memory.
Taking advantage of the RX 6000 series’ Infinity Cache, the software allows the CPU to directly access the graphics card’s full memory buffer. Smart Access Memory essentially means more efficient use of the combined memory between the CPU and GPU, reducing buffering and latency.
Until we get our hands on the hardware for benchmarking we aren’t sure what kind of gains you can expect but AMD showcased the use of Smart Access Memory with “Rage Mode”, resulting in increased performance across the board.
So normally we come to the end of an article like this and ask Ryzen 5000 vs 3000, which should you choose? Well, the answer isn’t definitive and you should certainly hold off until we see third-party benchmarking, which includes our own, as it will give you a Ryzen 5000 comparison too.
One conclusion we can feel quite safe saying is that if you are building a new PC, whether you have 1st-gen Ryzen or old-gen Intel CPU, the new Ryzen 5000 series is certainly worth your investment.
The Ryzen 5000 series is tipped to become the best consumer-level desktop processors we have ever seen, both from a gaming and workload perspective. What is going to be interesting is the performance uplift in combining an AMD CPU and GPU, something that could boost the overall value of AMD products even further.