AMD Ryzen 7 9700X

[sodium]

Thank you for including idle power numbers, so many reviews ignore it.

 

[evernessince]

So, the question then becomes is how do we know that the same isn't true for the 9700x? How do we know if they come out with a 9700 non x that there won't be even more power savings? We don't, not until they come out with one.

All signs points to the 9700X already being efficiency tuned out of the box. It's certainly possible that additional efficiency gains can be had (the same could also be said of the 7700 (non-x)) but the end result is not an interesting product on the efficiency front as there simply is not enough performance gains over last generation that would allow you to make massive cuts to the power consumption.

If we look at the 5.3 ghz OC power consumption that you are referencing it is within one watt of the 7700x in the 47 application average.

If we look at the performance of the 5.3 ghz we can see in the cinebench single thread that the 5.3 ghz OC gets 7.1 points per watt and the 7700x gets 5.5. In Cinebench that is a 29% better performance per watt for single thread. If we do multi thread the advantage is 10.2% in favor of the 9700x. If we do gaming the 5.3 ghz OC of the 9700x is 3.01 fps per watt, the 7700x is 2.66 fps, that means the 9700x gets 13% more fps per watt when overclocked to 5.3 ghz when compared to a stock 7700x. If the architecture wasn't any more efficient then we wouldn't be seeing these results. Maybe the architecture only shows it efficiency when you go up the wattage curve when compared to Zen 4.

You are trying to compare a tweaked config to a stock config, it should go without saying that's not an apples to apples comparison.

If you go back and look at TPU's review of the 7700X to compare apples to apples you can see the same OC setup (the exact same one you are using in your example so it's fair game) achieving a 21.56% increase n multi-threaded efficiency:

Now let's compare that to the 9700X in MT efficiency:

That is a mere 10.2% increase in efficiency over the stock 7700X. When you extrapolate the 7700X OC config onto the 9700X review data, which allows us to compare apples to apples, the 7700X OC 5.1 GHz setup is actually 11.36% more efficient in multi-threaded workloads than the 9700X.

Now let's look at gaming efficiency:

3.01 for the 9700X config:

4.85 for the 7700X OC config:

The 7700X tuned config is actually more power efficient than the 7700.

This is what I meant when I said there isn't enough performance headroom with the 9700X to make tuning the power interesting. There isn't a lot of power to shave off to begin with and Ryzen 7000 chips already get amazing results when tuned for power savings.

 

[Bavor]

What does price got to do with the review?

If you follow the context of the previous messages, its pointless to use slower RAM for both a review and as a consumer when it has a negative effect on the benchmark results and real world use cases. Especially when the slower RAM doesn't cost less than the much faster RAM. The DDR5 CAS 30 A die is among the lowest priced DDR5 6000 RAM available and currently the best performing in most benchmarks, games, and other real world use cases. For example, one of the other review sites showed a 9% difference in FPS between CAS 30 DDR5 6000 and CAS 38 DDR5 6000 in non X3D CPUs.

 

[wheresmycar]

How would AMD be able to do that? It's not lower clocks, that's for sure after looking at OC headroom. It is what it is.

Simple... they nerfed it. Thats all it takes. Lets not be under any illusion performance segmentation or product differentiation isn't a thing. The most likely culprit being enhancing the Zen 5 X3D appeal with that wider performance gap and no doubt costing an arm and a leg to land one. Problem being, if this trend continues for the gaming crowd the entire forward Gen support plan on AM5 seems pointless except for big wallet spenders eyeing up some of that X3D magic. I'm certainly not of the opinion we need more CPU power by todays gaming demands but it is what it is Zen 5 being a performance deprived AM5 add-on with some minor perks which are not worthy of a shout-out. As for other workloads which may fare more advantageously with that reasonably well-attired IPC uplift and efficiency... i'm only in it for the game hence ain't gonna dwell on that.

 

[Minus Infinity]

Man, what a perfect advertisement to wait for Arrow Lake. Maybe in typical AMD fashion the early bios and agesa suck, but this is pretty bad and the power is only good because of the performance ruining low TDP. The hype was strong yet again from AMD and having learned nothing form the RDNA3 fiasco they ploughed on ahead. It's like AMD has just released a 7700XT.

Steve at Hardware unboxed was only getting all core clocks of 4.4GHz, vs 7700X getting 5.2GHz. No wonder it uses less power. Hardly an architectural improvement.

I would love for these results to be due to be due to bad bios/agesa/memory issues, but if they aren't Intel will be finally catching a break. But for now you'd be insane not to buy 7800X3D or if you do both gaming and productivity 7950X3D and for productivity only, 7900X/7950X.

Also got to love the straight up lies from AMD earlier saying 6400 was memory sweet spot and now it's back to 6000 again.

I can only therefore expect RDNA4 rumours to also be full of crap too.

I wouldn't say it's disappointing - progress is progress even when it's minimal.

It's literally not progression. If you get 5% more performance for 20% worse price, it's regression. If Intel does this with Arrow Lake, it will get smashed. Hell, I wouldn't be surprised to see E-cores only competitive with stock 9700X, they can boost to ~4.7GHz and IPC is said to be ~ 2% up on Raptor cove.

 

[A Computer Guy]

For some reason these new AMD CPU's really don't like to do AES or Powerpoint compared to their predecessors. Some office exec's will see this and be like sorry AMD we still need to buy Intel for now because Powerpoint. Maybe this is where AM5 EPYC fills the void since it's still 7000 series chips? I'm jesting a bit of course. Normally users probably won't be upgrading from 7700x to 9700x but from lower SKU's. I love this format of chart and it would be awesome to be able to pick any two CPU's from the test run to compare like this. I would want to see 5700x vs 9700x.

 

[Minus Infinity]

I use 65 Watt CPU's so i'll never see such temps.

95C on a 65 Watt CPU is just weird.

When does it run at 95C. 95C is the temperature it could safely run all day every day. Even in Cinebench MT it hit what, 59C. That's cooler than my undervolted 5800X which runs at 63C when I set PPT at 118W.

What's with the low quality CAS 36 DDR5 6000 in the reviews? I've seen other benchmarks where CAS 30 vs CAS 36 had a significant improvement in gaming benchmarks on Zen 4. I assume the same would be applicable to Zen 5.

As long as it's apples to apples. 7700X benefits too from lower CAS.

Was expecting a bigger performance gap and higher FLCK but will need to wait for the newer boards to see how that pans out.

also

• No NPU for AI acceleration

Not really needed on a desktop with a DGPU.

NPU on current laptops do about 45-50 Tops

And old 2080 ti does something like 300 tops.

The 2080 Ti is 129TOPS. But 4080 is 820TOPS. My 6800XT though is a rather pathetic 67TOPS and no wonder my stable diffusion using Topaz is so slow.

RTX TOPS

 

[Punkenjoy]

For me, i think the current way of doing chiplets with Infinity Fabrics have been pushed to the maximum on desktop. They probably need a better way to connect the i/o die to the CCDs and need to be able to run much faster memory. By now, those CPU are probably memory starved and need to be able to run faster memory and faster I/O communication.

No progression on the memory side is a big downside but isn't really a surprise since the I/O die is the same as Zen4.

Without a new I/O die, those CPU really seems like a refresh. Also it look like they are doing pretty well in other area. AMD is making the big bucks on the server markets and i wouldn't be surprised if they sacrificed gaming performance for server performance.

 

[Minus Infinity]

So much pessimism. According to phoronix, the situation should be much more similar. They didn't use a 14700K but I think the 14900K gives a clear reading.


View attachment 357892View attachment 357893

All well and good but in the real world looking at productivity and scientific tests it is much less impressive, even showing regression in some tests and barely ever hitting 10% improvement. Now if this were labelled and priced as a 9700 with that same 65W TDP it would be much more palatable.

Super disappointing for sure and while X3D will fix the gaming performance hopefully it will still come with the almost negligible mt improvements over 7000.

WTF happened is my take on this what was AMD doing for 2 years.....

9950X reviews will be interesting I guess.....

Maybe, just maybe 9800X3D will get better bins, and all core clocks will surely need to be better than the terrible 4.4GHz of the 9700X. That's like Zen 3 5800X.

 

[Dyno]

So did anybody find out where the sweet spot is for these new 9000 series AMD cpu's? Is it still 6000 or is it 6400 for the memory portion? There is a review that i noticed they even put in an 8000MHz memory and it only made a difference in one game.

 

[Minus Infinity]

So did anybody find out where the sweet spot is for these new 9000 series AMD cpu's? Is it still 6000 or is it 6400 for the memory portion? There is a review that i noticed they even put in an 8000MHz memory and it only made a difference in one game.

AMD told hardware unboxed 6000 is indeed the sweet spot as IF only runs at 2000MHz, same as in Zen 4.

 

[TrantaLocked]

This is for people who want a low TDP with no serious performance sacrifices. The 100+ watt parts are going to be monsters.

 

[Makaveli]

When does it run at 95C. 95C is the temperature it could safely run all day every day. Even in Cinebench MT it hit what, 59C. That's cooler than my undervolted 5800X which runs at 63C when I set PPT at 118W.

As long as it's apples to apples. 7700X benefits too from lower CAS.


The 2080 Ti is 129TOPS. But 4080 is 820TOPS. My 6800XT though is a rather pathetic 67TOPS and no wonder my stable diffusion using Topaz is so slow.

RTX TOPS

i stand corrected it has been awhile since I looked at that link but I knew it was alot more than what you get on a laptop NPU.

 

[Launcestonian]

I haven't read through every single post so this may have been mentioned before; but with bios immaturity & chipset drivers as well influencing the deal, this plays a part here with this not very impressive 9700X performance review.

 

[A Computer Guy]

Hey they gotta get people off those A620 motherboards somehow… especially since these pull even less power.

Perhaps this power consumption improvement even leaves an opening for an overdrive socket adaptor for AM4? One can dream anyway.

 

[mosawo]

What is making so many people so unnaturally excited?
The TDP is low at 65W, and compared to the previous generation, the performance has increased and the MSRP has decreased. My impression of the 9600X and 9700X is thumbs up.

Gaming performance enthusiasts can just wait for the 9800X3D, and multi-threaded performance enthusiasts can just wait for the 9950X. The 9600X and 9700X are not their domain. That is what we experienced with the 7000 series and the 5000 series.

Furthermore, just as it took time for the AM4-CPU to mature, the AM5-CPU is also in the middle of maturing. When we found out that the ZEN5 was 4nm instead of 3nm, we should have expected and prepared for there not to be a big jump up. And it is very rare and not normal to get a big jump up by replacing it with the same named product every generation. It is a rare case to win a jackpot at the casino, and it is unreasonable to think that it should always be that way.

By the way, I have no plans to replace my 5600non-X. Even if the 9000 series is fantastic, it's not the time for me to invest in a CPU yet.
Again, my impression of the 9600X and 9700X is thumbs up.

 

[W1zzard]

Just in case some were claiming higher power usage on am5 than previous platforms, I just checked my 7800X3D with -25 on pbo offset at idle at its sitting at 28 watts. View attachment 357914

You need to measure the full system at the wall plug, not look at CPU power software reading, which can be pretty wrong, too.

We are interested in platform power, which includes the chipset, too, and all the other little things that could consume power.

 

[Craptacular]

All signs points to the 9700X already being efficiency tuned out of the box. It's certainly possible that additional efficiency gains can be had (the same could also be said of the 7700 (non-x)) but the end result is not an interesting product on the efficiency front as there simply is not enough performance gains over last generation that would allow you to make massive cuts to the power consumption.

Possible, but we won't know that until 9700 non-x is released.

You are trying to compare a tweaked config to a stock config, it should go without saying that's not an apples to apples comparison.

I wouldn't say I'm the one doing the comparison as it was you who made the point that the wattage was the same when doing a comparison of the OC'd 9700x vs stock 7700x, so if anyone was making the comparison it would be you.

If you go back and look at TPU's review of the 7700X to compare apples to apples you can see the same OC setup (the exact same one you are using in your example so it's fair game) achieving a 21.56% increase n multi-threaded efficiency:

Again, I'm not the one using it, you were the one who brought up that the 9700x OC had the same wattage as the stock 7700x. But it wouldn't be an apples-to-apples comparison to use the 7700x OC results as it isn't the same software, different games and different rendering software such as cinebench versions are different.

Honestly, the best comparison is to just do a 9700x vs 7700 non-x comparison as both are 65-watt tdp CPUs and in the 9700x review both CPUs were benchmarked using the same software at their stock configs, whereas the 7700x review with the 5.1 OC doesn't use the same software. So, yes on efficiency the 7700 non-x is more efficient at stock settings than the 9700x, the question that now needs to be answered is if AMD will come out with a 9700 non-x and what power consumption will it be and what will the power consumption of the 9800x3d be. You have already made up your mind on that, I have not.

 

[ShogoXT]

Il be eagerly awaiting more content on PBO and RAM tweaking. Based on some reviews there seems to be large ranges with the power targets.

As for RAM, that's never been stable on the Ryzen platforms unfortunately. Either way this is a loss for drop in default settings users.

Slightly off topic but even for my x470 and 3950x I never could get it stable and not cook itself until I learned how the tweak the PBO. Default settings were straight garbage. Ram b dies were unusable. Everything had to be tweaked. At least I finally learned...

Edit: Oh I just remembered. If AMD is copying Intel naming stack now, they have 65w and 125w CPUs locked. If AMD releases an 9800x, it would be the same...

 

[W1zzard]

For some reason these new AMD CPU's really don't like to do AES or Powerpoint compared to their predecessors. Some office exec's will see this and be like sorry AMD we still need to buy Intel for now because Powerpoint. Maybe this is where AM5 EPYC fills the void since it's still 7000 series chips? I'm jesting a bit of course. Normally users probably won't be upgrading from 7700x to 9700x but from lower SKU's. I love this format of chart and it would be awesome to be able to pick any two CPU's from the test run to compare like this. I would want to see 5700x vs 9700x.

View attachment 357927

Made the chart for you:

http://imgur.com/HLLGSXh

No plans for dynamic charts at this time, even though I get why you're liking them and how random combinations could be useful.

So did anybody find out where the sweet spot is for these new 9000 series AMD cpu's? Is it still 6000 or is it 6400 for the memory portion? There is a review that i noticed they even put in an 8000MHz memory and it only made a difference in one game.

AMD says "Memory: Some processors may be able to achieve EXPO 6400 1:1 with manual settings. By default AGESA will set any memory profile above 6000 MT/s to 1:2 mode, but an end user may override this to 1:1. Stability of this configuration will vary based on the specific processor. A latency optimized 1:1 EXPO memory profile will provide the best performance in a wide range of applications. There is no need to make any other adjustments. If you desire to make further tweaks AMD recommends trying to tighten the timings as much as possible as AUTO:1:1 DDR5-6000 MHz remains as the “sweet spot” for price and performance"

6000 works really well, 6400 requires a bit of luck and some tweaking of voltages, i.e. making it non-stock. DDR5-8000 MHz is possible, but due to the 1:2 mode it won't be that much faster. I have a G.SKILL kit coming, so will have data on this soon (not until after the 2nd round of reviews)

 

[Dr. Dro]

I will say hmm with a bit of erhmm.
The power draw is impressive and frequency is so low. No wonder it performs not what I have expected. Maybe AMD is leaving some headroom for the 9800x and/or x3d?

Frequency is not low for the node and even if you let it loose it doesn't seem to scale alongside power consumption. It is what it is. But it's still new, hopefully AGESA level improvements turn things around

 

[A Computer Guy]

Made the chart for you:

http://imgur.com/HLLGSXh

Many thanks!

 

[Pumper]

This should've been a regular 65 W 9700 for $300.

Considering the huge 28% increase in the number of transistors, this is incredibly disappointing. AMD quickly got too comfortable in the lead. Rooting for Arrow Lake now, hopefully it can shake things up.

I would bet on AMD just moving from non-X/X naming to X/XT, so instead of 9700, we will see a 105W 9700XT later on.

 

[Minus Infinity]

I forgot to check the benchmarks vs the 7700 non X. This is essentially GPU renaming where everything in RDNA3 other than 7900XTX was named a tier higher than it should have been. This 9700X is a direct replacement for the 7700, not 7700X. In applications it's 7.8% faster on average, in 720p gaming 5% faster on average, has basically identical power efficiency and has same TDP. I agree with Pumper, they need to release a 105W 9700XT now, as this thing is power starved and looks like bandwidth starved. Needs faster memory as well to shine.

 

[JWNoctis]

AMD says "Memory: Some processors may be able to achieve EXPO 6400 1:1 with manual settings. By default AGESA will set any memory profile above 6000 MT/s to 1:2 mode, but an end user may override this to 1:1. Stability of this configuration will vary based on the specific processor. A latency optimized 1:1 EXPO memory profile will provide the best performance in a wide range of applications. There is no need to make any other adjustments. If you desire to make further tweaks AMD recommends trying to tighten the timings as much as possible as AUTO:1:1 DDR5-6000 MHz remains as the “sweet spot” for price and performance"

6000 works really well, 6400 requires a bit of luck and some tweaking of voltages, i.e. making it non-stock. DDR5-8000 MHz is possible, but due to the 1:2 mode it won't be that much faster. I have a G.SKILL kit coming, so will have data on this soon (not until after the 2nd round of reviews)

So it's still the same IOD and IF draw-of-the-luck as on Zen 4, but as long as it's better luck than 2000MHz on Zen 3 I'd take it. Seems like they kept that to heart this time.

I would bet on AMD just moving from non-X/X naming to X/XT, so instead of 9700, we will see a 105W 9700XT later on.

Makes sense, especially if they would leave the non-X to OEM.