Ryzen 9 9900X CPU Reviews

conclusion is at 27:32
its better in blender...
Pretty wild to go from, "It isn't Intel" to "It's better in blender".

Can't wait until people stop beating this dead horse and move on. Let's hope Zen 6 (or will it be 5+, does anyone even know anymore?) is a better showing that Zen 5; or we will all be making Zen++++ jokes for the foreseeable future.
 
Pretty wild to go from, "It isn't Intel" to "It's better in blender".

Can't wait until people stop beating this dead horse and move on. Let's hope Zen 6 (or will it be 5+, does anyone even know anymore?) is a better showing that Zen 5; or we will all be making Zen++++ jokes for the foreseeable future.
If AMD & Intel keep this up we'll be telling "remember x86?" jokes.
 
Pretty wild to go from, "It isn't Intel" to "It's better in blender".

Can't wait until people stop beating this dead horse and move on. Let's hope Zen 6 (or will it be 5+, does anyone even know anymore?) is a better showing that Zen 5; or we will all be making Zen++++ jokes for the foreseeable future.
"beating this dead horse"

Bruh it hasn't even been a week...
 
"beating this dead horse"

Bruh it hasn't even been a week...
To be fair, the hardware is the hardware. It's not going to get magically better. After two years we've gotten marginally better performance, despite AMD's claims.

Based on this, the X3D parts will only be marginally better than the current ones - unless AMD jacks up the clock speed. He's right to be disappointed.

At this point, we can only hope that Intel will save us with something spectacular. Just feels like a very depressing time to be a PC enthusiast.
 
I'm quite excited myself as I'm a 70% non gaming, 30% gaming person so a 9900X will be perfect. Plus quite intrigued by the new boards.

I am a bit dissapointed as I was going to go Intel again as the past 8 years have been excellent (I was AMD from 1999 to 2016). However, will 15th gen be untainted?
 
Zen 5 is pure crap. From Anandtech

Compared to the Ryzen 9 7950X, we are seeing a slight increase in latencies within a single CCX. The SMT "advantage", where two logical cores sharing a single physical core have a lower latency, appears to be gone. Instead, latencies are consistently around 20ns from any logical core to any other logical core within a single CCX. That average is slightly up from 18ns on the 7950X, though it's not clear what the chief contributing factor is.

More significantly – and worryingly so – are the inter-CCD latencies. That is, the latency to go from a core on one CCD to a core on the other CCD. AMD's multi-CCD Ryzen designs have always taken a penalty here, as communicating between different CCDs means taking a long trek through AMD's Infinity Fabric to the IOD and back out to the other CCD. But the inter-CCD latencies are much higher here than we were expecting.

For reference, on the Ryzen 9 7950X, going to another CCD is around 76ns. But in Ryzen 9 9950X, we're seeing an average latency of 180ns, over twice the cost of the previous generation of Ryzen. Making this all the more confusing, Granite Ridge (desktop Ryzen 9000) reuses the same IOD and Infinity Fabric configuration as Raphael (Ryzen 7000) – all AMD has done is swap out the Zen 4 CCDs for Zen 5 CCDs. So by all expectations, we should not be seeing significantly higher inter-CCD latency here.


I'm quite excited myself as I'm a 70% non gaming, 30% gaming person so a 9900X will be perfect. Plus quite intrigued by the new boards.

I am a bit dissapointed as I was going to go Intel again as the past 8 years have been excellent (I was AMD from 1999 to 2016). However, will 15th gen be untainted?

As a 70% non-gamer, what exactly are the redeeming qualities you're seeing? Blender and Cinebench?
 
As a 70% non-gamer, what exactly are the redeeming qualities you're seeing? Blender and Cinebench?

Lower launch pricing than Zen 4 and significant efficiency gains, maybe? I'm not super stoked on the Zen 5 release, but the efficiency improvements seem to be the biggest selling point. It's definitely something I consider - less for the energy bill savings and more to avoid the extra heat output in summer (which then needs to be counteracted by running the AC more).
 
Lower launch pricing than Zen 4 and significant efficiency gains, maybe? I'm not super stoked on the Zen 5 release, but the efficiency improvements seem to be the biggest selling point. It's definitely something I consider - less for the energy bill savings and more to avoid the extra heat output in summer (which then needs to be counteracted by running the AC more).
Oh boy, I hate to break it to you, but this CPUs are not any more efficient that anything that was already out.


View: https://www.youtube.com/watch?v=6wLXQnZjcjU
 
Lower launch pricing than Zen 4 and significant efficiency gains, maybe? I'm not super stoked on the Zen 5 release, but the efficiency improvements seem to be the biggest selling point. It's definitely something I consider - less for the energy bill savings and more to avoid the extra heat output in summer (which then needs to be counteracted by running the AC more).
This is the most ill informed post I’ve read since Zen 5s release. Not only are they not more efficient, but even if they were, it certainly wouldn’t equate to a significant enough delta to make any appreciable difference in your AC run time.
 
This is the most ill informed post I’ve read since Zen 5s release. Not only are they not more efficient, but even if they were, it certainly wouldn’t equate to a significant enough delta to make any appreciable difference in your AC run time.

Lol. Snarky reply is unnecessary. If this is just a thread to bash people I'll bow right out. Thanks.
 
Not sure how factual a ranking of how much post are ill-informed can be ?
Pointing out something as misinformation when it is indeed misinformation is not an insult. Zen 5 is not more efficient. The post clearly states that the content of the post are not factual. Nothing in the post is directed at the poster, at all. Saying that it is the most ill informed is strictly an opinion, we can agree there; but to sit there and take that post as an insult is just going overboard. Facts do not care about feelings.
 
Pointing out something as misinformation when it is indeed misinformation is not an insult. Zen 5 is not more efficient. The post clearly states that the content of the post are not factual. Nothing in the post is directed at the poster, at all. Saying that it is the most ill informed is strictly an opinion, we can agree there; but to sit there and take that post as an insult is just going overboard. Facts do not care about feelings.
Zen 5 consumes less power for the same or more performance than Zen 4 while running lower clockspeeds on multi-core loads. It's not only more power efficient but more efficient in having higher IPC.

How is that not more efficient?
 
Zen 5 consumes less power for the same or more performance than Zen 4 while running lower clockspeeds on multi-core loads. It's not only more power efficient but more efficient in having higher IPC.

How is that not more efficient?
The claim was “significant” efficiency gains that’ll lead to using your AC less. I’ve been watching and reading reviews for the past week and haven’t seen it. Have you? Beyond that, it often times performs worse. Did you see those numbers or is this a case of blind leading the blind?
 
The claim was “significant” efficiency gains that’ll lead to using your AC less. I’ve been watching and reading reviews for the past week and haven’t seen it. Have you? Beyond that, it often times performs worse. Did you catch those numbers or is this a case of blind leading the blind?
And you stated there are no efficiency gains at all.

Not only are they not more efficient,
 
And you stated there are no efficiency gains at all.
I said they’re not more efficient because they’re not. They trade blows.

Now can you or can you not show me these significant efficiency gains that’ll allow you to adjust your thermostat? While you’re digging for that info, dig for me saying “at all” since that was your claim. Kind of amazing you can quote and misquote me at the exact same time.
 
Last edited:
Zen 5 consumes less power for the same or more performance than Zen 4 while running lower clockspeeds on multi-core loads. It's not only more power efficient but more efficient in having higher IPC.

How is that not more efficient?
Yeah, we've heard this marketing bullet point ad nauseam. But hey, don't take it from me, take it from GN. Someone who people on this forum laud when he is singing praises about Ryzen chips.


View: https://www.youtube.com/watch?v=6wLXQnZjcjU&t=2s
 
AMD achieved Intels levels of disappointment on this launch, blah, , , ,. X3D maybe better but not holding my breath. Hope Intel actually has something this time around with Arrowlake.
 
Yeah, we've heard this marketing bullet point ad nauseam. But hey, don't take it from me, take it from GN. Someone who people on this forum laud when he is singing praises about Ryzen chips.


View: https://www.youtube.com/watch?v=6wLXQnZjcjU&t=2s

You obviously didn't read TheFPSReview's reviews and maybe they were the only ones test and watch actual clockspeeds. Considering most of them are former [H]ard|OCP reviewers I trust them more than anything else you link.

The AMD Ryzen 9 9950X has a base clock of 4.3GHz and a boost clock of up to 5.7GHz. In our all-core/full-load testing in Cinebench R23, running for 10 minutes we experienced all the cores at the same time running at 4.7GHz-4.8GHz on all cores. We also noticed an interesting trend, most of the time half of the cores were at 4.7GHz while the other half were at 4.8GHz. If we compare this frequency to the previous generation Ryzen 9 7950X, we find that the new AMD Ryzen 9 9950X runs slower on all-core/full-load. The Ryzen 9 7950X was running at 5GHz-5.1GHz on all cores in this same test, using the same AIO. Perhaps that extra 30W of power and temperature allowance helps the 7950X to run at a higher clock speed at all times. The fact remains, on all cores the AMD Ryzen 9 9950X runs slower than the previous generation.

Less power used. Lower clockspeeds. Same or more performance. Explain how that is not more efficient.
 
HWUB, TPU, GN... all show the same thing, the efficiency gains are minuscule and often none. And there are games and apps where Zen 5 performs worse, leading to factually worse efficiency at times.

This becomes particularly evident when you run Zen 4 and 5 parts with the same TDP to make it a fair comparison (which not all reviewers did, and certainly not AMD).*

On average most reviews find it a little more efficient... but I think it's quite natural that most people are not excited about single digit improvements across the board, not to mention the occasional regression which makes the overall picture just bad.

Very likely has to do with the latency issue that has been reported, Rocket Lake on Intel side suffered from that too, and struggled to outperform 10th gen consistently because of it.

*and for gamers, let's not forget the 7800 X3D, a Zen 4 part, that still outperforms every other CPU, both in terms of framerate and watts. For them it makes Zen 5 look like a joke. Perhaps not a surprise, but since AMD was hyping up Zen 5 to be some gaming monster....

edit: and now AMD is backpedaling on their claims and agreeing with the reviewers:

View: https://www.youtube.com/watch?v=IaPo5io-Gfw
 
Last edited:
You obviously didn't read TheFPSReview's reviews and maybe they were the only ones test and watch actual clockspeeds. Considering most of them are former [H]ard|OCP reviewers I trust them more than anything else you link.



Less power used. Lower clockspeeds. Same or more performance. Explain how that is not more efficient.
One dude makes the most ill informed post ever, you top it with the funniest. So cinebench is the hill youve chosen to die on? When was the last time someone recommended new build recommendations for 24/7 cinebench runs? No wonder no one is seeing what you’re seeing, they’re looking at actual applications and games they use. Never witnessed such a desperate and sad cope.

So your best case scenario is cinebench and 30 watts… two led can lights essentially. Exactly how long do you plan to run cinebench that you’ll get to adjust your thermostat?
 
One dude makes the most ill informed post ever, you top it with the funniest. So cinebench is the hill youve chosen to die on? When was the last time someone recommended new build recommendations for 24/7 cinebench runs? No wonder no one is seeing what you’re seeing, they’re looking at actual applications and games they use. Never witnessed such a desperate and sad cope.

So your best case scenario is cinebench and 30 watts… two led can lights essentially. Exactly how long do you plan to run cinebench that you’ll get to adjust your thermostat?

1) Maybe in your mansion you don't notice that more heat means the A/C runs more often, but if your PC is in the room above the garage like mine, it's something you pay attention to.
2) The claim wasn't that you have to adjust your thermostat. The claim was that if there is more heat generated in the room, the AC will have to run more frequently. This compounds when you have a variety of electronics running like I do. If you pay some attention to power draw and thus heat, efficiency sums across devices. Not rocket science.
3) My post said I didn't think Zen 5 was a very exciting release.
4) Your neck vein must be close to bursting. Calm down.
 
When was the last time someone recommended new build recommendations for 24/7 cinebench runs?
Oh, oh. I know this one. Zen 1. People here were throwing cinebench numbers around left and right to promote Zen 1 as the best thing since sliced bread. In reality, it wasn't until Zen 3 when AMD reached parity and bested Intel.
 
Zen 5 is only more efficient in Zen 4 in specific scenarios, while in others it is actually LESS efficient. It's not like by buying Zen 5 you are guaranteed better efficiency all across the board. And even then, the efficiency gains in the majority of scenarios are tiny enough to probably be considered a moot point by many. You really gotta cherry pick to find situations where it is vastly more efficient. If that's what it takes to tout the efficiency argument, then maybe reconsider if that's even worth touting at all.
 
1) Maybe in your mansion you don't notice that more heat means the A/C runs more often, but if your PC is in the room above the garage like mine, it's something you pay attention to.
2) The claim wasn't that you have to adjust your thermostat. The claim was that if there is more heat generated in the room, the AC will have to run more frequently. This compounds when you have a variety of electronics running like I do. If you pay some attention to power draw and thus heat, efficiency sums across devices. Not rocket science.
3) My post said I didn't think Zen 5 was a very exciting release.
4) Your neck vein must be close to bursting. Calm down.
You’re absolutely right. Allow me to clarify my point for use cases like yours instead of my gross over-generalization.

If you’re looping cinebench in your air conditioned mud shack, it may run for a few seconds less with Zen 5.

Better?
 
Zen 5 is only more efficient in Zen 4 in specific scenarios, while in others it is actually LESS efficient. It's not like by buying Zen 5 you are guaranteed better efficiency all across the board. And even then, the efficiency gains in the majority of scenarios are tiny enough to probably be considered a moot point by many. You really gotta cherry pick to find situations where it is vastly more efficient. If that's what it takes to tout the efficiency argument, then maybe reconsider if that's even worth touting at all.

TPU did a power consumption test across 47 applications and the 9900X used 10 watts less than the 7900X on average. For the 9950X it was a 14 watt difference. Not huge, but it's there.

They also did a gaming test across 13 games and there were improvements there too, but even smaller.

I agree it's a pretty lackluster improvement. I'll be curious to see if the results improve once AMD gets the software figured out. The five day lag time in letting reviewers know about the special setup process hints that AMD is itself just learning about its new CPUs lol.
 
TPU did a power consumption test across 47 applications and the 9900X used 10 watts less than the 7900X on average. For the 9950X it was a 14 watt difference. Not huge, but it's there.

They also did a gaming test across 13 games and there were improvements there too, but even smaller.

I agree it's a pretty lackluster improvement. I'll be curious to see if the results improve once AMD gets the software figured out. The five day lag time in letting reviewers know about the special setup process hints that AMD is itself just learning about its new CPUs lol.

Which only highlights one of my points that the gains are so tiny that it's a completely moot point. If you wanna save a measely 10-15 watts of power that's literally how much 1 or 2 LED light bulbs will pull. Also funny how a 10-15 watt improvement in efficiency is all praiseworthy for AMD yet over in the Nvidia GPU section people are more than happy to run their GPUs at 600 watts all day and throw efficiency out the window.
 
When did a new CPU was not more efficient in some power band than the previous one (would they not be just slower if that was the case ?), the term is usually reserved for when it is significant. 14900k was also more efficient than the 13900k who was also more efficient than the 12900k, unlike the 11900k->12900k giant efficacy jump it was too small to be a talking point.
 
Last edited:
One dude makes the most ill informed post ever, you top it with the funniest. So cinebench is the hill youve chosen to die on? When was the last time someone recommended new build recommendations for 24/7 cinebench runs? No wonder no one is seeing what you’re seeing, they’re looking at actual applications and games they use. Never witnessed such a desperate and sad cope.

So your best case scenario is cinebench and 30 watts… two led can lights essentially. Exactly how long do you plan to run cinebench that you’ll get to adjust your thermostat?
I'm not trying to die on any hill. All I did was post proof that Zen 5 is more efficient in terms of power and IPC compared to Zen 4 in some workloads. Something you said didn't exist.
 
I'm not trying to die on any hill. All I did was post proof that Zen 5 is more efficient in terms of power and IPC compared to Zen 4 in some workloads. Something you said didn't exist.
If it "is" more efficient how is it there's several scenarios where it isn't? Sounds like you have a literacy and/or comprehension problem. Perhaps Bill Clinton can help you define "is"
 
more efficient how is it there's several scenarios where it isn't?
CPU efficicacy change from task to task when they change enough between generation, for example the zen5 was worst at it when gaming on some setup:
https://www.techpowerup.com/review/amd-ryzen-7-9700x/23.html
Fps per watt
7700.: 3.02
7700x: 2.66
9700x: 2.59


Cinebench single thread was also below the 7700.

I imagine it will climb back when all said it done fix are in. It is not some fix value, it change for the power used and the task being done (and on which OS-setting)
 
Last edited:
I'm sorry but if efficiency is what's primarily being touted as the selling point for a new generation of hardware, then it sucks, period. Just think for a second what would happen if the RTX 5090 comes out, is only 5% faster than a 4090, but uses 50 watts less than a 4090 to achieve that performance. Is everyone going to be singing it's praises about how much more efficient it is compared to it's predecessor and calling it awesome? Efficiency can go shove it when the raw performance gains are next to nothing.
 
I'm sorry but if efficiency is what's primarily being touted as the selling point for a new generation of hardware, then it sucks, period. Just think for a second what would happen if the RTX 5090 comes out, is only 5% faster than a 4090, but uses 50 watts less than a 4090 to achieve that performance. Is everyone going to be singing it's praises about how much more efficient it is compared to it's predecessor and calling it awesome? Efficiency can go shove it when the raw performance gains are next to nothing.
One possible would be when efficiency was a major issue, AM4 was already extremely good in the marketplace that regard, the 7700 was better than the 5950x which was incredible, that why it is hard to get excited.

If there is some core Ultra something of Intel that is 5% faster than the 14700k while using 40-50 less watt (game and application usage similar to a 9700x with PBO on), that could be exciting
 
One possible would be when efficiency was a major issue, AM4 was already extremely good in the marketplace that regard, the 7700 was better than the 5950x which was incredible, that why it is hard to get excited.

If there is some core Ultra something of Intel that is 5% faster than the 14700k while using 40-50 less watt (game and application usage similar to a 9700x with PBO on), that could be exciting

If Arrow Lake is only 5% faster than Raptor Lake then that's disappointing as well despite Raptor Lake being a space heater. They made big changes to Arrow Lake and are even removing hyperthreading altogether so if all those changes only add up to a 5% performance increase then that will be extremely disappointing to see.
 
One possible would be when efficiency was a major issue, AM4 was already extremely good in the marketplace that regard, the 7700 was better than the 5950x which was incredible, that why it is hard to get excited.

If there is some core Ultra something of Intel that is 5% faster than the 14700k while using 40-50 less watt (game and application usage similar to a 9700x with PBO on), that could be exciting
5950x has twice the core count of a 7700 though. It's not hard to be more efficient with that kind of handicap. Especially when gaming since games don't typically scale past 8 core anyway.
 
If Arrow Lake is only 5% faster than Raptor Lake then that's disappointing as well despite Raptor Lake being a space heater. They made big changes to Arrow Lake and are even removing hyperthreading altogether so if all those changes only add up to a 5% performance increase then that will be extremely disappointing to see.
I agree. But if it's a 5% gain across the board, and doesn't have the deg issues, that's still more impressive than Zen 5 which is 3-5% faster... sometimes. If they can avoid these nagging scheduling issues in Windows, that's another plus.
 
Back
Top