AMD Ryzen 7 9800X3D

[JustBenching]

No. The whole point of testing in lower resolutions and variety of settings is to remove GPU from working hard as much as possible and see the full stretch of the CPU. They absolutely did the right thing and other reviewers should learn from them.
If you actually watch the video, you will aslo notice that they do not use RT, and rightly so.

I don't care about the resolution, they can drop to 240p for all I care. But lowering settings is silly, lots of settings including RT are CPU demanding.

 

[Makaveli]

I almost wonder if 7800x3d is still technically the better chip, look at those power savings, for not much difference in performance: source: techspot

View attachment 370535

Isn't there a 20% performance difference in this multicore test between the 7800X3D vs the 9800X3D?

 

[evernessince]

But... you're not a 7800X3D owner...

Look at my system specs. "7800X" was a typo.

 

[Space Lynx]

Isn't there a 20% performance difference in this multicore test between the 7800X3D vs the 9800X3D?

My initial look was just that page and the 1440p gaming page, where it was only a 3% average gain over 7800x3d. But after looking at more reviews, several games actually get 15%+ gains 7-8% in 1440p. Still decent. I am still waiting for 11800x3d before I upgrade, and my 7800x3d will be retired to the storage bin in the event I need a warranty replacement or emergency cpu.

 

[Tek-Check]

I don't care about the resolution, they can drop to 240p for all I care. But lowering settings is silly, lots of settings including RT are CPU demanding.

Did you read my post?
- no RT was used
- lowering settings makes CPU work harder and GPU lighter. That's the whole point of stressing the CPU.

 

[btk2k2]

Did you read my post?
- no RT was used
- lowering settings makes CPU work harder and GPU lighter. That's the whole point of stressing the CPU.

In fairness RT is demanding on both CPU and GPU so running at 720p DLSS P with RT on can still alleviate a GPU bottleneck while also ensuring that the CPU is being stressed by having the CPU demanding settings enabled.

 

[Godrilla]

It's alive.

 

[Tek-Check]

In fairness RT is demanding on both CPU and GPU so running at 720p DLSS P with RT on can still alleviate a GPU bottleneck while also ensuring that the CPU is being stressed by having the CPU demanding settings enabled.

They will certainly run more tests in coming weeks and months. Initial reviews should never be as complicated, with hundreds of graphs.
Reviewers did a good job with 15-20 minutes reviews, to give an overall picture of the CPU. More details will follow. Hardware Unboxed will release tomorrow 7800X3D vs 9800X3D in 45 games, plus other videos next week.

 

[JustBenching]

Did you read my post?
- no RT was used
- lowering settings makes CPU work harder and GPU lighter. That's the whole point of stressing the CPU.

Did you read mine?

Oh nevermind...

 

[Tek-Check]

Did you read mine?

I did. You have other reviewers who used RT. Find them.

 

[3valatzy]

120 watts is extremely high. AMD must stop with these insane voltages and wattages!

The chip should have been 65 watts, maybe 95 watts maximum after heavy overclock.
To see where a normal chip ends, look at that:

If it is not for the higher power draw, these chips would offer little to nothing over the already aging AM4 platform.

Speaking of x86-64 stagnation. Not worth it to buy anything.

 

[Kaleid]

120 watts is extremely high. AMD must stop with these insane voltages and wattages!

View attachment 370886

The chip should have been 65 watts, maybe 95 watts maximum after heavy overclock.
To see where a normal chip ends, look at that:

View attachment 370887

If it is not for the higher power draw, these chips would offer little to nothing over the already aging AM4 platform.

Speaking of x86-64 stagnation. Not worth it to buy anything.

If that is a problem one could simply try ECO mode through Ryzen Master

 

[Arkz]

Look at my system specs. "7800X" was a typo.

So you bought a 7800X3D and a 4090, but a 20W power difference has you concerned?

 

[chrcoluk]

Those fuckers .. "sorry but we're temporarily out of stock of your ordered items. Of course we'll send the missing item as soon as it's here. Expected date: unknown"

http://imgur.com/7kcvNZc

There was clearly an indication of "stock available right now" when I ordered, the payment was completed through Amazon Pay, which should be instant .. I am friends with a lot of people at Caseking, but next time I am buying something I'll buy somewhere else

Hate it when companies do that. Back ordering things they say are in stock.

To be honest, I was expecting 9800X3D to be a bit less power-houngry, but the fact is it still a very efficient CPU and is worthy to upgrade from Zen 3. Zen launch 5 is a disappointment for me as well, but then when I look at Arrow Lake launch, I should not be disappointed at all. So my side is currently having better time, but that does not explicitly mean that it's good. Intel should really catch up because AMD needs force which will drive technological innovations further. As we learned from history, lack of competition paves ground for stagnation.

I did notice it seems to now have higher power usage in some scenarios, but it still isnt as bad as other chips on the market and there was a performance uplift, so yeah I do like the look of the 9800X3D.

Agree on needing one competitor pushing the other.

 

[evernessince]

So you bought a 7800X3D and a 4090, but a 20W power difference has you concerned?

Both of which I've power tuned.

I run AI and encoding tasks 24/7 so yes, power consumption is absolutely important to me.

You keep quoting that cherry-picked 20w figure but that's only in some games. For heavily threaded applications or in my case, power consumption will be more than double with a 9800X3D vs my 7800X3D.

 

[AnotherReader]

Both of which I've power tuned.

I run AI and encoding tasks 24/7 so yes, power consumption is absolutely important to me.

You keep quoting that cherry-picked 20w figure but that's only in some games. For heavily threaded applications or in my case, power consumption will be more than double with a 9800X3D vs my 7800X3D.

In some applications, it's 15% to 40% faster than the 7800X3D. Power limiting it to the latter's default would probably still allow it to be significantly faster.

 

[Godrilla]

Update one thing 1 noticed even at 4k at 120hz vsync on on Horizon zero dawn remastered the 99% gpu rendering went up from 94 to to 102 fps. vs 7800X3D. Could be a coincidence because the game has been patched to version 1.1.4.2.
So far everything seems smoother subjectively speaking lol.
PBO offset to -40, curve optimizer all cores and 200 mhz added although my overall Cinebench 24 MT score is slightly lower than what this review has at stock. (My score 1289, 1272 pts second run) ST score matched the stock it's score of 134.
Maybe it's my strix itx x670e motherboard vs x870e.
Core temps seems to be giving wrong clock readings as well.

 

[evernessince]

In some applications, it's 15% to 40% faster than the 7800X3D. Power limiting it to the latter's default would probably still allow it to be significantly faster.

I agree but the argument you're jumping into isn't that you couldn't tune the 9800X3D to be more efficient. Arkz was making the argument that these CPUs are primarily for gaming and therefore efficiency outside of gaming doesn't matter in addition to implying a 20w difference isn't notable.

A cinebench stress test?
In gaming, what people are actually buying these for, on average it's a 20W difference.
If 20W more is too much for you on a more powerful chip that costs this much money then you should probably change your gaming habits. Like to using a Steam Deck instead.

My counter-argument was that these CPUs are indeed used outside of gaming (my own personal use case was used as proof of that). I'd also say that 20w figure is misleading for a few reasons, including it's assumption that said user only plays light to moderately demanding games, assumes they never use any other heavy application, and assumes that shrugging at 20w doesn't materialize as a policy towards part selection in general which results in the end user selecting more power hungry components across the entire build thus resulting in a larger increase in power consumption as compared to those who rank efficiency as a more important criterion.

Only a small selection of enthusiasts will have the know-how to be able to tune these CPUs which is why out of the box stats are so important. You could argue that nearly any CPUs is efficient when tuned, even the 14900K (and certain TPU members do make that very argument and it is valid, for a very small group of people).

 

[umeng2002]

It's alive.

How's the uplift in Kingdom Come Deliverance?

 

[Godrilla]

How's the uplift in Kingdom Come Deliverance?

Unfortunately I don't have the before performance figures. From my recollection the most demanding seens were at night and performance would drops to 60 fps but here I am getting 100 fps although it could also be Noctua d15/7800X3D vs 9800x3d/360 aio cooler. 4k maximum settings. I guess if you have the 7800X3D/4090 @ 3ghz you can compare your performance.

 

[Makaveli]

My initial look was just that page and the 1440p gaming page, where it was only a 3% average gain over 7800x3d. But after looking at more reviews, several games actually get 15%+ gains 7-8% in 1440p. Still decent. I am still waiting for 11800x3d before I upgrade, and my 7800x3d will be retired to the storage bin in the event I need a warranty replacement or emergency cpu.

That I understand but you used the cinibench image and not the gaming one in your post. So abit confusing.

 

[Hankieroseman]

If all I ever bought was AMD the last 20 years, does that make me a fanboy or just a satisfied customer?

I signed up for one @microcenter Dallas so I hope one is there in the morning. I should go right now but I ain't feelin' it.

 

[Godrilla]

I signed up for one @microcenter Dallas so I hope one is there in the morning. I should go right now but I ain't feelin' it.

They hold reservations for 3 days as per Google search.

 

[Hankieroseman]

Look at my system specs. "7800X" was a typo.

That Noctua Air Cooler works fine don't it?

 

[evernessince]

That Noctua Air Cooler works fine don't it?

It's great, nothing to complain about there. Easy to install, doesn't make much noise, and keeps the CPU cool. Running 77c under 24/7 load.