AMD’s new Zen 5 Ryzen 9000 processors are just around the corner, and today we can reveal a little more information about them. As a quick recap, last month at Computex, AMD announced the Ryzen 9000 series. This includes four main desktop processors in a fairly standard lineup: Ryzen 9 9950X, Ryzen 9 9900X, Ryzen 7 9700X, and Ryzen 5 9600X.
The main benefit we get here is an average IPC increase of 16%, similar to AMD’s previous Zen architecture upgrades. Other aspects of the design remain unchanged from the Ryzen 7000, including the number of CPU cores (which is still between 6 and 16 across the board) and peak frequencies that push up to 5.7GHz. There are also no 3D V-Cache models yet, meaning the initial batch of Zen 5 processors tops out at 32MB of L3 cache on the CCD.
We now have a release date for these processors: July 31, 2024, right at the end of AMD’s first July release period. What we don’t have though are the prices, which is a bit odd to be honest. AMD will start selling these processors in just over two weeks, and there’s no word on how much they’ll cost or how much buyers will have to save to get Zen 5. In contrast, when Zen 4 was announced in 2022, nearly a month before launch, we they had complete information on prices and specifications.
But we also don’t really know what it means or how much to read into it. Usually, when a company doesn’t want to divulge specific information, it’s because the response isn’t very positive.
Most of the Zen 4 lineup is now discounted below its original MSRP, sometimes by hundreds of dollars, like the 7900X, which you can get right now for just $360. So earlier price comparisons with full MSRP Zen 5 models may not be so favorable. But it could also be the case that the prices are quite competitive and AMD doesn’t want to cannibalize Zen 4 sales in the next few weeks. We’ll probably get the prices in time for reviews, but we’re still in the dark right now.
New performance requirements for Ryzen 9000
What we’re getting today are more snapshots of the first party’s performance. Back at Computex, we got very limited performance information, mostly focusing on the Ryzen 9 9950X compared to the Core i9-14900K. Now AMD is expanding these comparisons with other models.
Most of these images are quite typical in the information they provide. The six games shown for each comparison are the same for all models. AMD advertises that the 9950X is an average of 13% faster than the 14900K for gaming across these titles, although as always, first-party benchmarks should be taken with a grain of salt due to the level of cherry-picking that can only happen on 6 titles. With the Ryzen 9 9900X, AMD suggests 12% more performance than the 14900K, which should mean very similar gaming performance to the 16-core model.
With the 9700X, AMD now compares it to the 14700K, again showing a 13% performance improvement for the Zen 5 part. Based on this data, the 9600X is expected to be 14% faster than the Core i5-14600K.
What is perhaps more interesting are the productivity numbers, which appear to be significantly more selective. For example, for the 9900X versus the 14900K, AMD shows Blender as one example of higher performance, but for the six- and eight-core models, this benchmark is swapped out for 7-Zip, which is usually a pretty favorable benchmark for AMD CPUs. There’s also a strong focus on single-threaded results like Procyon Office, Geekbench and Photoshop, as opposed to multi-core tasks, outside of Handbrake, which is expected to benefit significantly from AVX-512 Zen 5 support.
Reading between the lines, this suggests that the Ryzen 9000 CPU may not be as competitive for multi-threaded workloads compared to single-threaded workloads given the closest Intel parts. E-cores are a significant part of Intel’s architecture, which massively helps with multi-threaded productivity applications: the 14700K has 12 E-cores in addition to 8 P-cores, and the 14600K gives you a 6P+8E configuration. As a result, Intel is likely to maintain its leadership in multi-core productivity; we previously found the 14600K to be more than 50% faster than the Ryzen 5 7600X in some workloads like multi-core Cinebench, which is a huge gap to bridge in a generation that doesn’t see any increase in core count.
AMD also compared the Ryzen 7 9700X to the Ryzen 7 5800X3D, claiming that the new Zen 5 part is an average of 12% faster than the first generation 3D V-Cache models while using less power. This is an interesting comparison given that it’s effectively Zen 5 versus Zen 3, not a comparison to the newer Zen 4-based 7800X3D. However, AMD also mentioned that a similar comparison to the 7800X3D would show the new 9700X losing “a few percentage points ” faster.
All of these claims are not fully consistent with our current gaming CPU data, which found the Ryzen 7 5800X3D to be slightly slower than typical non-X3D Zen 4 CPUs. It is 4% faster for the 7600X versus the 5800X3D and 8% faster for the 7700X versus the 5800X3D. There’s almost no crossover between our test suite and AMD in the games tested – and we’re using an RTX 4090, while AMD typically uses a Radeon RX 7900 XTX – but that would only mean a small increase in gaming performance for the 9700X over the 7700X. .
At the same time, for the 9700X to be 13% faster than the 14700K across a wide range of games, it would actually be faster than the 7800X3D in our data and over 30% faster than the 5800X3D. It’s hard to know exactly where things will land, as AMD claims the 9700X is slightly faster than the 7800X3D. I guess you shouldn’t really trust first party benchmarks, but if it were to be true, the gap over the 5800X3D would have to be bigger, which would at least contradict one of the claims AMD made, somewhat contradicting others and our testing.
With all that said, AMD compared the Ryzen 9000 to the Ryzen 7000 directly in Blender with multiple threads, claiming various performance gains while boasting a lower TDP. But here’s an important reminder: TDPs aren’t a true measurement of power consumption, just a rating of the cooler’s required performance. Just because the Ryzen 5 dropped from 105W TDP to 65W TDP while improving performance by 17% doesn’t mean that the actual power consumption of these parts dropped by 40 watts.
New features of the Zen 5 processor
So what else can you learn about Zen 5 processors? AMD expects decent gains from using Precision Boost Overdrive, especially for the Ryzen 7 9700X. There have been rumors claiming that AMD is going to bump the TDP of the 9700X from 65W to something higher like 120W at the last minute, but that is not the case based on the information AMD is showing just two weeks after launch.
Instead, AMD claims that due to the lower default TDP, the PBO will provide more extra headroom than usual, especially for the 9700X, where it claims a 15% performance boost.
The Ryzen 9000 processors will run cooler than the Ryzen 7000 thanks to a 15% improvement in thermal resistance, which AMD claims will result in a 7 degree Celsius reduction in temperature at the same TDP.
In general, the Zen 4 CPUs were easy to cool with only modest power consumption, especially compared to 14th Gen Intel, but ran above 90°C under full load. This is a nice benefit for people who are temperature sensitive and worried about cooking their CPUs, and should also allow the CPU frequencies to run higher at a given temperature.
There are other overclocking improvements, especially for memory. Base JEDEC memory support has increased from DDR5-5200 in Zen 4 to DDR5-5600, and there’s OC support up to DDR5-8000, which AMD tells us is actually achievable in real-world workloads, but the sweet spot will be lower. The most interesting inclusion, however, is something called the Memory Optimized Performance Profile.
Basically, this means that memory frequencies and timings can be adjusted on the fly depending on the workload. This allows for the benefits of high-frequency and low-latency memory without having to constantly tweak things in the BIOS.
The way it was explained is that if you buy a high end memory kit like DDR5-8000 in some workloads it would push the frequency up to 8000 speeds on looser timings like CL38 because usually when running at higher speeds you can’t have super tight timing.
But for some games, for example, the system will automatically reduce the frequency and tighten the timing if necessary. This way, your DDR5-8000 kit can actually be beneficial in a wide variety of use cases, instead of being somewhat unusable for gaming due to looser timing and worse latency. We’ll see how efficient it is and how well it tunes memory on the fly.
For CPU overclocking, there is also an evolution of the curve optimizer function called the curve shaper, which allows more control over the voltage and frequency curves at different operating temperatures.
Zen Architecture 5
AMD has made some deeper revelations about the Zen 5 architecture, the highlights being a full 512-bit data path for AVX-512 instructions, increasing the L1 data cache from 32 KB 8-way to 48 KB 12-way, doubling the maximum bandwidth of the L1 cache, and better branch prediction.
There was an increase from a 6-wide send in Zen 4 to an 8-wide send, an ALU increase from 4 to 6, and there is now a dual decode channel on the frontend.
This is how AMD sees each of these architectural improvements affecting final IPC and overall performance. A big driver of the gains was improvements in execution and decoding, with bandwidth also a significant factor.
New AMD B850 and B840 chipsets
AMD has announced two new chipsets to join the 800 series. At Computex, we got the X870 and X870E as the flagship chipsets and saw a lot of new boards from various manufacturers. Now AMD is also unveiling the widely known B850 and B840 chipsets, the breakdown of which can be seen in this handy chart.
Bottom line, the X870 and X870E command USB 4 and PCIe 5.0 for graphics and a primary M.2 slot. The B850 drops the USB 3.2 20Gb/s PCI 5.0 requirement for the primary NVMe drive only, with 5.0 for graphics optional. The B840 limits things further, only supporting PCIe 3.0 instead of 4.0 as the default, dropping USB 3.2 support to 10 Gbps and removing CPU overclocking support.
So where do things fall compared to the existing 600 series?
The AMD X870 and X870E are a step up from the X670 series in that all boards will now support USB 4 and the non-E boards will have more PCIe 5.0 connections. The B850 is a mix of the B650 and B650E, but mostly an update of that chipset.
The B840 ends up below the A620 – while both chipsets drop overclocking support and drop USB speeds to 10Gbps, the B840 further drops the PCIe bus to 3.0, while the A620 remains PCIe 4.0. This could make other aspects of the B840 end up above the A620, such as the number of USB ports supported. All 600 series chipsets support memory overclocking.
AMD also says the new boards have improved memory routing to allow for more reliable memory performance at higher speeds, such as the DDR5-8000 mentioned earlier, although we’ll have to see which boards get these improvements. However, as mentioned at Computex, these new 800-series motherboards will not be released alongside the Zen 5, instead coming a few months later.
Finally, we also got a minor update on the roadmap, which features a picture of the Zen 6. Absolutely zero other details, so honestly, it’s really just hype, no indication of when we can expect the Zen 6 or what kind of processor it uses, just nothing. what would be funny
All AMD has said is that Zen 6 is on track, whatever that means. The gap between Zen 3 and Zen 4 was roughly 22 months, and between Zen 4 and Zen 5 there is the same 22 months. Based on that, we should expect Zen 6 in May 2026 at full estimate.
And that goes for today’s Zen 5 news. You should see our full reviews of these processors in a few weeks, along with plenty of follow-up content in the coming months.