AMD Ryzen 9 5900X

[RandallFlagg]

No one cares about the apps and productivity benchmarks? Everyone just went straight to the gaming results? AMD is destroying Intel in every benchmark.

At the very top end yes. The 5800X and 5900X are winners if you're in the market for a $450+ CPU. I don't think anyone is disputing that. At least I haven't seen anyone disputing it. Most contention revolves around the much more mainstream 5600X.

 

[X71200]

No one cares about the apps and productivity benchmarks? Everyone just went straight to the gaming results? AMD is destroying Intel in every benchmark.

There are cases in games even where the heavy core amounts come into usage. Even in that poor Counter-Strike configuration, when you're loading bot routes to a map for the first time, a HCC CPU makes you and your players wait less. Honestly you get almost nothing over a specific framerate. Like if this CPU gives you 150 FPS in your *name any generic looter shooter*, you will not really get gains by getting the Intel that does 155 FPS. As for apps, definitely. The architecture is superior especially now for most.

Meh, I wouldn't do LN2. Needs to be something a slightly skilled layman can put in, maybe limit it to AIO cooling solutions.

There are frequency "walls" in the Ryzen architecture. Beyond a specific point, it just doesn't scale as well as Intel. If you go for extreme methods of non-everyday cooling, Intel might reach a higher frequency - that however is irrelevant since your normal PC won't do it. You also have to keep in mind that these CPUs increased the bar against clock walls. You can reach higher frequencies now.

At the very top end yes. The 5800X and 5900X are winners if you're in the market for a $450+ CPU. I don't think anyone is disputing that. At least I haven't seen anyone disputing it. Most contention revolves around the much more mainstream 5600X.

You're free to post that to the review of that CPU here.

 

[W1zzard]

While 3200C16 is most common and affordable now, I believe it would be fair to go with 3600C16 for Ryzens since that would be the best balance between cost and performance that is recommended for AMD systems. 3200C14 definitely has low latency, albeit it does limit the top end performance.

AMD Zen 2 Memory Performance Scaling with Ryzen 9 3900X

We take a close look at memory scaling on AMD's new Zen 2 Ryzen 3900X, testing both application and gaming performance at seven different memory speed and timing combinations ranging from 2400 MHz all the way up to 4000 MHz.

www.techpowerup.com

The difference between 3200CL14 and 3600CL16 will be very small I think

 

[X71200]

They're pointing to the fact of Infinity Fabric clocks, not just RAM, you probably already know this but heads up. The price difference between 3200 and 3600 RAM is also pretty much insignificant money.

 

[RandallFlagg]

This is what happened with Zen 2 vs 9900K, overclocking RAM, CPU. Intel will get a big boost as well and if you're going to OC the RAM, it should be on both platforms. I've seen a lot of sites with false comparisons in this Zen 3 release. Look at the 9900K with DDR4-4133 CL 17 OC'd.

Anyway, this with a 10700K 10900K and the 5600X/5800X/5900X would be a good comparison.

 

[Cheeseball]

AMD Zen 2 Memory Performance Scaling with Ryzen 9 3900X

We take a close look at memory scaling on AMD's new Zen 2 Ryzen 3900X, testing both application and gaming performance at seven different memory speed and timing combinations ranging from 2400 MHz all the way up to 4000 MHz.

www.techpowerup.com

The difference between 3200CL14 and 3600CL16 will be very small I think

Looks like it does depend on the game still. SOTR likes less latency, yet Wolfenstein II/Youngblood favors more bandwidth.

Will be looking forward to your RAM latency/bandwidth tests!

They're pointing to the fact of Infinity Fabric clocks, not just RAM, you probably already know this but heads up. The price difference between 3200 and 3600 RAM is also pretty much insignificant money.

Sometimes the difference can be $50 to $75 depending on the timing. The pricing between 3200CL16 and 3600CL18 may be minimal, but overclockable RAM (Samsung b-die or Micron e-die) costs more.

 

[X71200]

Sometimes the difference can be $50 to $75 depending on the timing. The pricing between 3200CL16 and 3200CL18 may be minimal, but overclockable RAM (Samsung b-die or Micron e-die) costs more.

I mean $25 is not big money compared to what you're shelling out for the CPU, and you can get cheap B-Die and E-Die. They're not as expensive as you might think. I can get relatively good results out of my now cheaper 3600 B-Die. Like 3800 or 3600 at CL16 non-issue.

 

[W1zzard]

Sometimes the difference can be $50 to $75 depending on the timing. The pricing between 3200CL16 and 3600CL18 may be minimal, but overclockable RAM (Samsung b-die or Micron e-die) costs more.

That, and I rather spend that money on a faster GPU, or bigger SSD, or more 32 GB instead of 16 GB

 

[X71200]

That's your own choice though. You can get higher quality SSD by paying $25 more, or alternatively you can get still good quality SSD + slightly faster RAM to run Infinity Fabric faster. I wouldn't disagree that fast RAM doesn't do miracles, but saying what you would do in this case seems to mostly represent your choice and not everyone else's.

 

[W1zzard]

seems to mostly represent your choice and not everyone else's.

Absolutely. Every reviewer has his own preferences and ways of doing things, and that's good, because it gives you more data to base your buying decision on

 

[Cheeseball]

I mean $25 is not big money compared to what you're shelling out for the CPU, and you can get cheap B-Die and E-Die. They're not as expensive as you might think. I can get relatively good results out of my now cheaper 3600 B-Die. Like 3800 or 3600 at CL16 non-issue.

I don't have pricing for the 16 GB kits anymore, but here are the ones I was looking at:

G.SKILL Trident Z Neo - DDR4 3200 - 14-14-14-34 (Samsung B-die) - $224.99

G.SKILL Trident Z Neo - DDR4 3600 - 16-19-19-39 (Hynix CJR/CFR) - $173.98

While not a big $50 gap, it is still a significant cost. This is taking into fact just sticking with stock/XMP/DOCP settings (for the average consumer). Obviously if one is experienced, it would be best to go with the 3200CL14 and overclock it.

Also there is the fact that those who would be upgrading to any of the Ryzen gen. 4 CPUs (like from Coffee Lake or Zen 1/1.5) would most likely be using their previous RAM, which could be similar to the ones I mentioned above.

 

[zlobby]

We need the memory subsystem numbers, e.g. cache latencies!

 

[tsillis7]

wtf is going on with you techpowerup,all the grafs shows that intel w

ins in all game benchmark??????this is a joke!hardware unboxed,and gamers nexus shows that 5900x is better than 10900k!

 

[N3M3515]

No one cares about the apps and productivity benchmarks? Everyone just went straight to the gaming results? AMD is destroying Intel in every benchmark.

They only care about de 1% lead in games from intel (that's only on this site by the way...)

 

[RandallFlagg]

wtf is going on with you techpowerup,all the grafs shows that intel wView attachment 174541ins in all game benchmark??????this is a joke!hardware unboxed,and gamers nexus shows that 5900x is better than 10900k!

Even at Tom's, it's by no means a clean sweep. Not even in productivity. If that's what you think happened, you might want to look at the charts again. It's good, better than Intel even, but there's no clean sweep.

Also, Toms' tested with a 3090, while TPU / Guru3d and some others tested with a 2080Ti which is their standard platform. Related to that comment, Tom's has a habit of re-using their benchmarks. Meaning their tests may be with updated drivers and such.

 

[tsillis7]

Even at Tom's, it's by no means a clean sweep. Not even in productivity. If that's what you think happened, you might want to look at the charts again. It's good, better than Intel even, but there's no clean sweep.

Also, Toms' tested with a 3090, while TPU / Guru3d and some others tested with a 2080Ti which is their standard platform. Related to that comment, Tom's has a habit of re-using their benchmarks. Meaning their tests may be with updated drivers and such.


View attachment 174548


View attachment 174546


View attachment 174542

View attachment 174543

my problem is that this site shows that 5000 cpus loosing in every game and making amd look bad!only in this site i saw that!
(We gave it away already in the very first lines of this review; the original ZEN design was bronze, ZEN2 silver, and ZEN3 truly is gold. The release of Ryzen 5000 probably will make a whole bunch of people at Intel nauseous as from every and any angle AMD now is faster. So that last bit you could nag about with a Ryzen proc is gone] !!!!!!!!!<this words are from guru3d site nothing else!!!

 

[RandallFlagg]

my problem is that this site shows that 5000 cpus loosing in every game and making amd look bad!only in this site i saw that!
(We gave it away already in the very first lines of this review; the original ZEN design was bronze, ZEN2 silver, and ZEN3 truly is gold. The release of Ryzen 5000 probably will make a whole bunch of people at Intel nauseous as from every and any angle AMD now is faster. So that last bit you could nag about with a Ryzen proc is gone] !!!!!!!!!<this words are from guru3d site nothing else!!!

Guru3d only tested with 4 games, a pathetically small sample set, for one. For another, they tested with DDR4-3600. They also re-used prior benchmarks for Intel and Zen 2, so there are probably differences in driver versions since they only state 'Latest drivers' and their 10900K review for example was done back in May. TPU actually used the same video drivers as their May review of the 10900K, according to the test system setup.

Last but not least, on the game benchmarks at Guru3d, in one game at least the 10900K is missing and in another the 10700K is missing. Then you've got these 1fps differences @ 180fps+. At 1440p, the 10900K (except where it is missing) basically ties it up with a 5900X and a 5950X which are much more expensive chips. Look at the graphs.

 

[Dante Uchiha]

my problem is that this site shows that 5000 cpus loosing in every game and making amd look bad!only in this site i saw that!
(We gave it away already in the very first lines of this review; the original ZEN design was bronze, ZEN2 silver, and ZEN3 truly is gold. The release of Ryzen 5000 probably will make a whole bunch of people at Intel nauseous as from every and any angle AMD now is faster. So that last bit you could nag about with a Ryzen proc is gone] !!!!!!!!!<this words are from guru3d site nothing else!!!

That's literally the only review I've read that shows intel Comet lake on par with Zen3 or a bit ahead in games.

There's something really wrong with this review, and it's not just a matter of memory, I've seen reviewers using the same frequency(3200Mhz) and getting better results. Maybe the problem is the OS version ? Only the Zen 3 platform is using the updated version of windows ?

Even the OC results are worst than in others sites... most reviewers are achieving 4.7~4.8Ghz @ 1.4v.

 

[SIGSEGV]

AMD Ryzen 5000 Vermeer (Zen3) Review Roundup | VideoCardz.com

AMD Ryzen 5000, Featured, Review Roundup Today AMD launches its Ryzen 5000 series processors. AMD makes Ryzen 5000 series official Zen3 is here. AMD today lifts the embargo on Ryzen 5000 series

videocardz.com

sorry, seems only TPU that has an anomaly.

 

[tsillis7]

Guru3d only tested with 4 games, a pathetically small sample set, for one. For another, they tested with DDR4-3600. They also re-used prior benchmarks for Intel and Zen 2, so there are probably differences in driver versions since they only state 'Latest drivers' and their 10900K review for example was done back in May. TPU actually used the same video drivers as their May review of the 10900K, according to the test system setup.

Last but not least, on the game benchmarks at Guru3d, in one game at least the 10900K is missing and in another the 10700K is missing. Then you've got these 1fps differences @ 180fps+. At 1440p, the 10900K (except where it is missing) basically ties it up with a 5900X and a 5950X which are much more expensive chips. Look at the graphs.
i can post 1000 more grafs in this forum from other sites and big youtubers showing amd cpus even 5600x that is faster in few games than 10900k but in this site i dont see not even one graf amd in front of intel that is pathetic!and why he dont test it with the latest gpus?3080 or 3090 ?

 

[RandallFlagg]

View attachment 174555View attachment 174556View attachment 174557

Be careful that you aren't seeing only what you want to see.

 

[AnarchoPrimitiv]

Paul's hardware on YouTube used 3600Mhz Memory and an RTX 3080 and shows the 5900x beating the 10900k in basically everything

 

[mrthanhnguyen]

I have seen hardware unboxed, Jokerproduction has similar gaming benchmark result as TPU. Tech Jesus has kinda similar too. Im eager to see if someone has 5900x and 3090 so I can compare with my pc.

 

[Dante Uchiha]

Be careful that you aren't seeing only what you want to see.

View attachment 174558

View attachment 174559

View attachment 174560

View attachment 174561

So we can get random results at different resolutions to try to justify the grotesque flaw in this review ? Okay then...

 

[tsillis7]

Be careful that you aren't seeing only what you want to see.

View attachment 174558

View attachment 174559

View attachment 174560

View attachment 174561

you still dont get it man i dont care ho is faster for 1or2% better cpu!my problem are the grafs and test methodology is miseleading!not even 1 game that amd wins omg techpowerup!!!!!!!!!!!!!and is for sure that in some games amd is faster ,why he dont pick 1 only ?????for fan!!!!