Key findings
- Given the average performance increase of 5%, neither gamers nor developers have reason to be happy.
- AMD's big promises regarding performance and energy efficiency evaporated as soon as the new CPUs saw the light of day.
- AMD may have targeted the server market with Zen 5 for its EPYC CPUs, thereby compromising desktop performance.
Earlier this month, several publications began running reviews of AMD's latest Ryzen 5 9600X and Ryzen 7 9700X. Reactions to AMD's Zen 5 processors have been mixed, and the incredible performance gains (especially in gaming) that AMD claimed have largely been proven false. Both gamers and developers had expected the Ryzen 9000 CPUs to take the AM5 platform to new levels, but what we've gotten is stagnation.
Focusing on the most anticipated chips in the lineup, the 9600X and 9700X, it's easy to conclude that Zen 5 is far from the no-brainer many of us were hoping for. So let's analyze the three big areas where AMD has disappointed consumers with its latest CPUs.
Related
Windows has affected the performance of the latest AMD processors
It turns out that the poor performance of the new Ryzen 9000 series could be partly due to Windows.
3 Gamers generally got higher-priced Zen 4 CPUs
Stick with your 7600 and 7700
Two years after the launch of the Ryzen 7000 “Zen 4” CPUs, this is not what you'd expect from a “new and improved” architecture. Sure, AMD may have managed to pack a lot more transistors onto an even smaller chip (by switching from TSMC N5 to N4P for the CCDs), but that engineering achievement didn't translate into any notable improvements in gaming performance. It turns out we only got 2-3% more FPS on average compared to the Zen 4 chips.
It makes more sense to compare the 65W 9700X and 9600X directly with the 65W 7700 and 7600 models.
Also, this comparison is between the newer chips and the Ryzen 7 7700 & Ryzen 5 7600, rather than the 7700X and 7600X. This is because it makes more sense to compare the 65W 9700X and 9600X to the 65W 7700 and 7600 respectively, apples to apples. Even taking into account that Windows impacts the performance of Zen 5 and Zen 4 CPUs, the gaming gains are still small compared to the previous generation.
The Ryzen 7 and Ryzen 5 chips consume significantly more power than their predecessors while offering identical performance
Enabling PBO on the Ryzen 7 and Ryzen 5 chips does basically nothing for gaming. Hence all the talk about Zen 5 CPUs being “more power efficient” is also untrue, as the Ryzen 7 and Ryzen 5 chips consume significantly more power than their predecessors for identical performance, sometimes over 30% more. And when you consider that we didn't get any real “price reductions” on the 9700X and 9600X compared to the 7700 and 7600 respectively, things look even worse.
After 2 years, AMD is basically asking you to pay significantly more for the exact same CPU that just has a lower TDP. The lower power cap doesn't translate to better FPS per watt either. AMD could have rated their newer chips at 105W by default instead of 65W to squeeze out 1% more FPS, but it made more sense to stick with 65W. The sad thing is that this has done next to nothing for gamers in terms of either performance or temperature.
Related
Why my upgrade from Zen 3 to Zen 5 has nothing to do with the new Ryzen 9000 CPUs
The new Ryzen 9000 chips look decent, but I'm saving an upgrade for something better on
2 Creators also have no reason to worry
More multi-core performance for more watts?
While the Zen 5 chips don't give gamers anything to look forward to, developers at least have something to look forward to. However, they'll have to contend with a much hotter CPU. At the default 65W TDP, the Ryzen 7 9700X is only about 5% faster than even the 7700X in multi-core performance. Only when you enable PBO do you get a nearly 20% improvement over the Zen 4 chip.
However, this performance increase comes at the cost of significantly higher CPU temperatures, which hurts AMD's value proposition. AMD seems to have focused a little too much on the rated TDP of its new chips rather than delivering strong out-of-the-box performance. Perhaps the idea was to portray its CPUs in a better light compared to Intel's usually hotter offerings, but this has once again resulted in unspectacular gains after 2 years.
AMD seems to have focused a little too much on the rated TDP with its new chips instead of delivering strong out-of-the-box performance.
As a result, we have a situation where neither gamers nor professionals have a valid reason to consider the 9700X or 9600X. Even the flagship Ryzen 9 9950X, which many developers might rely on, is only about 3% faster on average than the Ryzen 9 7950X, so it can't be recommended instead of the Zen 4 chip. So who is Zen 5 really for?
The 9950X may be faster than Intel's Core i9-14900K in many professional workloads, but so was the previous generation 7950X. The glaring problem with the launch of Ryzen 9000 is the small generational improvements we're seeing over Zen 4. Consumers have been used to seeing big improvements with each generation of Ryzen CPUs, so these results have obviously disappointed many users, especially since AMD made big promises that turned out to be simply untrue.
Related
The new Ryzen 9000 CPUs show that AMD is playing it safe
Granite Ridge is almost here, but should you consider buying these new desktop CPUs?
1 Zen 5 is probably only for the server market
Looks like desktop users suffered
Source: AMD
I may be clutching at straws here, but given that the Ryzen 9000 desktop chips are nothing special, I think AMD might be interested in optimizing Zen 5 for its server CPUs. In the past, we've seen almost every subsequent Zen architecture handily beat the previous one, but this time around, AMD doesn't seem to care if Zen 5 is basically a like-for-like replacement for Zen 4, despite being inferior in several areas.
This may have resulted in a lackluster architecture for desktop users, with neither gamers nor developers seeing any point in considering these new chips.
AMD may be focused on maximizing its profits from the EPYC line of server processors. And considering that's where the money is, that's not surprising. AMD has increasingly followed in Nvidia's footsteps and focused on the data center market for GPUs. It now appears that the same story has played out in the CPU space, at least for this generation.
However, this may have resulted in a lackluster architecture for desktop users, with neither gamers nor developers seeing any point in considering these new chips. For every type of user, there seems to be a cheaper or more powerful option than the Ryzen 9000 CPUs.
The only people who will ultimately buy the Ryzen 9000 CPUs are probably reviewers – just to produce more data with which to criticize AMD.
Gamers like me will likely wait for the 9000X3D CPUs or be happy with their Ryzen 7000 chips. Pros will simply go for the Ryzen 7000 flagship or the Core i9-14900K/13900K. And those building entry-level systems won't consider these new chips anyway, given the price. The only people who will ultimately buy the Ryzen 9000 CPUs will likely be reviewers – just to produce more data to trade to AMD.
Related
Nvidia and AMD are in the AI ​​era and gamers can do nothing but watch
The AI ​​wave has become a tsunami for gamers as Nvidia and AMD focus all their energy on data center GPUs
Will the Ryzen 9000X3D CPUs also disappoint?
I'm one of the many users waiting for the Ryzen 9000X3D chips in hopes of upgrading to the fastest gaming CPU in the world. Personally, coming from the Ryzen 7 5700X, it would be a big upgrade for me, but I fear the same Zen 5 issues could plague the upcoming Ryzen 7 9800X3D and other X3D CPUs.
Unless the 9800X3D brings a significant generational shift over the Ryzen 7 7800X3D, AMD could lose the gaming crown to Intel. Even now, the 14900K is only about 5% slower than the 9800X3D, so it's not inconceivable that the upcoming Arrow Lake chips could outperform not only the 7800X3D, but the 9800X3D as well. We could see a shake-up in the desktop CPU space soon, and as always, it will be exciting to report on whatever happens.