• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD Ryzen 7 2700X 3.7 GHz

Uuumm...no. matter of fact, exactly opposite is true. Decrease the visual fidelity and the cpu has more work to do, insresease the settings and resolution and the cpu has less work.

Unless it is some physics running on CPU. If I remember correctly some graphics options have cpu bottleneck too(was it geometry, can't remember).
 
Unless it is some physics running on CPU. If I remember correctly some graphics options have cpu bottleneck too(was it geometry, can't remember).
Some postfx effects and geometry too. They like fast single core. This (I think) is actually the reason for a very curious thing -in games like watch dogs 2 (a lot of gemoetry in the city), Ryzen 7 can scale on all 16 threads equally, but it still loses to an i5 which hits 80-90% utilization on 6 threads.
https://www.purepc.pl/procesory/test_procesorow_amd_ryzen_7_2700x_vs_intel_core_i7_8700k?page=0,42
Like I said, more cores is not equivallent to faster core and vice versa, some games prefer one to the other.
Physics is done on the gpu these days, at least most of the times.
 
Last edited:
Stock 7600K-like performance in games from an overclocked 8c/16t is disappointing to say the least . 2700X has serious workstation performance tho. I just can't explain people putting ryzen in rigs that are mostly for gaming.

Things in bold for speed.

1 - I upgraded before the 8700k was released, at the time, a nice 8 core was a 'psychological' boost over my old 6 core.
2 - The 8700k, which wasn't out, was pushed out at a lower Intel price point (had I been able to see into the future and see Intel's post Ryzen pricing restructure, I may not have gone with Ryzen).
3 - The Intel tendency to ditch motherboard for new chips was a factor to me. I can upgrade to the next Ryzen mArch change and keep the motherboard.
4 - My Ryzen was a sideways upgrade, willing to experiment with a new tech arrival, you know, like tech enthusiasts like to do.
5 - I own a 1080ti. Going by your logic of not being able to explain putting a Ryzen in a gaming rig, I can't explain why anybody would buy a Vega 56 or 64.
6 - I gamed at 1440p till last week when I then went 4k. Intel has nothing exceptional to offer me over Ryzen.

Point 4 is the most delicate point. Buying the fastest thing doesn't make it interesting. I enjoyed my Ryzen build. And again, when I bought it, the 8700K wasn't in existence.
 
AMD - we make your intel CPU a better buy for the money !
Well, yeah. It's why, pre Ryzen launch I always said that even if you buy Intel exclusively, it's still wise to root for AMD.
 
Hmm .. maybe we could list them as Base/Boost. ie "Core i7-8700K 3.7 / 4.7 GHz". But wouldn't that add a ton of noise to the charts?
What do other readers think?
yeah it may look so but that would be more accurate and less misleading. especially for new comers. also adding CPU Name - Defualt Clock/Boost Clock label top of the chart may make it perfect :)
 
Things in bold for speed.

1 - I upgraded before the 8700k was released, at the time, a nice 8 core was a 'psychological' boost over my old 6 core.
2 - The 8700k, which wasn't out, was pushed out at a lower Intel price point (had I been able to see into the future and see Intel's post Ryzen pricing restructure, I may not have gone with Ryzen).
3 - The Intel tendency to ditch motherboard for new chips was a factor to me. I can upgrade to the next Ryzen mArch change and keep the motherboard.
4 - My Ryzen was a sideways upgrade, willing to experiment with a new tech arrival, you know, like tech enthusiasts like to do.
5 - I own a 1080ti. Going by your logic of not being able to explain putting a Ryzen in a gaming rig, I can't explain why anybody would buy a Vega 56 or 64.
6 - I gamed at 1440p till last week when I then went 4k. Intel has nothing exceptional to offer me over Ryzen.

Point 4 is the most delicate point. Buying the fastest thing doesn't make it interesting. I enjoyed my Ryzen build. And again, when I bought it, the 8700K wasn't in existence.
Lol I'm not judging you. You don't have to explain dick to me :) You made your decision, and even though I'm more inclined towards intel cpus in gaming rigs, I can't argue with any of the points you just made, they make sense.
 
Last edited:
Great review. Thank you @W1zzard

With this kind of performance I'd rather pick a 8700K TBH. Just overall a stronger CPU in gaming, which is what the price segment is for.
 
Great review. Thank you @W1zzard

With this kind of performance I'd rather pick a 8700K TBH. Just overall a stronger CPU in gaming, which is what the price segment is for.

definitely is. Gaming performance, especially for me as a 144Hz user is what prevents me most from going AMD.
Still, these are OK. Can be recommended to most people.
 
Well so that's it, now we wait for lesser prices and for next gen cpu-s.
 
At thier weakest the Ryzen 1600 is 100% faster than the 7600K
How does 7600K do at Ryzen 1600's weakest?

Is HS soldered like first Gen Ryzen chips or they using TIM between the die this time around.
It is soldered, mentioned a couple of times in the review
 
Some postfx effects and geometry too. They like fast single core. Physics is done on the gpu theres days, at least most of the times.

Two of the most common physics middlewares PhysX and Havok are both running on cpu most of the time, but granted one can't even necessary disable those physics. Well maybe some games have cloth and debris options but yeah it varies.
 
Wow, at that price point it almost makes sense to upgrade my 1700X.
 
DDR4-2933? Why? Isn't DDR4-3000 more available? (Cascade Lake-X is rumored to feature DDR4-2933 as well…)

I wish they sold versions without these boxed fans. This CPU is hot and needs a proper cooler. Stock fans are fine on low-end products, but at least >$300 products should be offered without them.

-----

It's sad to see that AMD have taken such extreme measures to close the gap with Intel. I'm afraid products like this which boosts outs of thermal specifications will be the new norm, and unfortunately, Intel will probably respond by doing the same thing. We have already seen boosting taken too far on Vega and Pascal, where real world builds easily loose 5-10% when putting the card in a case with normal airflow. But this is now taken one step further by boosting beyond thermal specifications, rendering all open case or open bench tests useless since the results are no longer reliable. This is now effectively an overclocked product, and should be regarded as such.

These chips hit Intel's 7th generation Core "Kaby Lake" series so hard, that the company cut its yearly generational product cycle by half and rushed in the 8th generation Core "Coffee Lake" series, with 50-100% core-count increases across the board, to restore competitiveness.
The planned bump in core count has been known to the public since around the launch of Skylake and is not a result of Ryzen in any way. Even the greatly delayed Ice Lake was done in tapeout early last summer. These are designs made before Intel knew what Zen would provide. Please refrain from such attempts to misrepresent the history when you know better.

Have AMD forgotten their own product from 2011? Or are they no longer considering their FX line as "8-core"?
 
Well, yeah. It's why, pre Ryzen launch I always said that even if you buy Intel exclusively, it's still wise to root for AMD.
if only rooting alone would help AMD make better cpus...
Actually the only way to not let Intel and nVidia inrease prices and to gèt their future products for less money is to buy AMD products now so they can spend more on devoloping their products.
it really works like that. i am not an amd fan nor an intel hater. if AMD was in Intels place i would have said the opposite
 
Excellent review, I have been very curious about this chip for awhile. I am guessing its temp limit is about the same as before. Did you manage to ever make it throttle?
 
How does 7600K do at Ryzen 1600's weakest?

I don't quite understand the question? when your gaming performance depends on the CPU, not the GPU, the Ryzen 1600 is roughly around 100% faster than the 7600K.

Your reviews don't reflect that, they appear to be best case for low core count CPU's, in other words your CPU reviews are not based on CPU performance, they mislead people into thinking 4 cores are better than 6 or 8 when in fact in real life the higher core count CPU's are vastly better gaming CPU's.

Where are your testing methodologies? i want to replicate them so i can help you see exactly where you are going wrong.
 
Last edited:
Commendable effort. by AMD and comes pretty close to many of the pre-launch hoopla. Obvious choice for the game who's using workstation apps, at least a third of the time. And for the gamer, while I understand the cost argument logically, I don't think $70 is going to impact gamer choices in this $300+ CPU price niche.

On thing I'd like to see is the effect of overclocking in a couple of gaming titles. For example, when comparing AMD and nVidia GPUs, TPU reviews let us see the impact of overclocking. I think the best example of that was the 480 review which showed:

The MSI 480 overclocks 8.6% and the MSI 1060 overclocks 15.1%.. So when the 1060's (10% performance advantage @ stock) is overclocked, the relative difference would be:

https://www.techpowerup.com/reviews/MSI/RX_480_Gaming_X/26.html
https://www.techpowerup.com/reviews/MSI/GTX_1060_Gaming_X/27.html

110% x (115.1 / 108.6) = 116.6% of the 480s speed or 16.6 % faster

In some cases, one card @ stock is faster but when both are OC'd the situation reverses. For those that OC, stock performance is something they will never see. It would be an overwhelming task to do 18 games but a chart with 2 - 4 with just the new CPU and the most apt price niche competitor would be very welcome.

The solder thing however is clearly serves as no more than a marketing ploy .... why do folks delid ? so they can avoid the temp wall and get higher OCs ...they aren't getting any OC.

Can ya turn off the RGB Stuff ?
 
Last edited:
if only rooting alone would help AMD make better cpus...
Actually the only way to not let Intel and nVidia inrease prices and to gèt their future products for less money is to buy AMD products now so they can spend more on devoloping their products.
it really works like that. i am not an amd fan nor an intel hater. if AMD was in Intels place i would have said the opposite
I get that you think you know what people should buy, but that's people's choice, really.

What you're suggesting doesn't work. It requires everybody turning their purchases into a political statement. It also requires rewarding the player that falls behind, which is not always desirable.
They are businesses, it's their job to compete with each other. Buyers are seldom concerned with the politics behind that.

I would buy from the underdog if a comparable product is $10-20 more expensive or at the same price I can get 5-10% less performance (all other things being equal). But I never picked Bulldozer over Core. Nor do I expect people that do not follow CPU development as religiously as I do to have the knowledge to do the same.
 
During each benchmark? Average? Which cores? Average the clocks? Or the highest? Or the lowest? (serious question)

I'm thinking about this right now for the request further up on reporting clocks at various thread-counts
Hate to sound greedy, but how about all the above.

Edit: side note i guess...if the results are this good...Only AMD can ruin Zen2(which I doubt) will happen.
 
Last edited:
Stock 7600K-like performance in games from an overclocked 8c/16t is disappointing to say the least . 2700X has serious workstation performance tho. I just can't explain people putting ryzen in rigs that are mostly for gaming.
I wonder that myself as well. It’s like a gamer choosing Vega over a 1080ti for its compute power that he will never utilize and in the end, he’s stuck with the option that produces less FPS.
 
Interesting results from Anandtech... Ryzen 2700X walks all over the 8700K https://www.anandtech.com/show/12625/amd-second-generation-ryzen-7-2700x-2700-ryzen-5-2600x-2600/17

And this may be why....

"We ran our tests on a fresh version of RS3 + April Security Updates + Meltdown/Spectre patches using our standard testing implementation. "

Did TPU do that or was it a clean system without the performance hampering Intel fixes?

97168.png
 
Last edited:
DDR4-2933? Why? Isn't DDR4-3000 more available? (Cascade Lake-X is rumored to feature DDR4-2933 as well…)

I wish they sold versions without these boxed fans. This CPU is hot and needs a proper cooler. Stock fans are fine on low-end products, but at least >$300 products should be offered without them.

-----

It's sad to see that AMD have taken such extreme measures to close the gap with Intel. I'm afraid products like this which boosts outs of thermal specifications will be the new norm, and unfortunately, Intel will probably respond by doing the same thing. We have already seen boosting taken too far on Vega and Pascal, where real world builds easily loose 5-10% when putting the card in a case with normal airflow. But this is now taken one step further by boosting beyond thermal specifications, rendering all open case or open bench tests useless since the results are no longer reliable. This is now effectively an overclocked product, and should be regarded as such.


The planned bump in core count has been known to the public since around the launch of Skylake and is not a result of Ryzen in any way. Even the greatly delayed Ice Lake was done in tapeout early last summer. These are designs made before Intel knew what Zen would provide. Please refrain from such attempts to misrepresent the history when you know better.


Have AMD forgotten their own product from 2011? Or are they no longer considering their FX line as "8-core"?

The amd official system memory for Ryzen 7 2700x is 2933MHz, over than that is considered overclocking(xmp profiles). Last gen. Ryzen 7s it was 2666MHz, thus review were with 2667MHz memory.
 
Interesting results from Anandtech... Ryzen 2700X walks all over the 8700K https://www.anandtech.com/show/12625/amd-second-generation-ryzen-7-2700x-2700-ryzen-5-2600x-2600/17

And this may be why....

"We ran our tests on a fresh version of RS3 + April Security Updates + Meltdown/Spectre patches using our standard testing implementation. "

Did TPU do that or was it a clean system without the performance hampering Intel fixes?

View attachment 99976
I wouldn’t put much weigh in Anandtech. Meltdown doesn’t/barely hurts gaming performance. Their review is just weird compared to TPU and everywhere else (even biased ones like Hardware Unboxed had 2700x trailing by some).
 
I don't quite understand the question? when your gaming performance depends on the CPU, not the GPU, the Ryzen 1600 is roughly around 100% faster than the 7600K.

Your reviews don't reflect that, they appear to be best case for low core count CPU's, in other words your CPU reviews are not based on CPU performance, they mislead people into thinking 4 cores are better than 6 or 8 when in fact in real life the higher core count CPU's are vastly better gaming CPU's.

Where are your testing methodologies? i want to replicate them so i can help you see exactly where you are going wrong.

LOL you act like this is Wizzards first go at a review.
 
I wouldn’t put much weigh in Anandtech. Meltdown doesn’t/barely hurts gaming performance. Their review is just weird compared to TPU and everywhere else (even biased ones like Hardware Unboxed had 2700x trailing by some).

Anand are one of the oldest, most knowledgeable and professional reviewers there are, they have been around doing it better than most for decades.

I think they are on to something, they usually are.
 
Back
Top