Ryzen 7 1700 vs. Core i7-7820X: 8-Core Royal Rumble

You need to avoid the GPU bottleneck to see the CPU bottleneck on GPU.

You obviously don't even understand what bottleneck means.

You sure like to mince words and read between the line and create your own story. Go on denying the benchmark results that everyone can see. The trend the obvious, ryzen bottlenecks at 1080p very heavily and very obviously, because there is no GPU bottleneck. At 1440p, the GPU limitation is more obvious but, not dominant enough to hide the CPU bottleneck. Then at 4K, the GPU bottleneck is all you get to measure and allows people like you to be blissfully ignorant about CPU bottleneck and pretend it doesn't exist.

Ryzen might be very small bottleneck on 1080p but Rysen is so fast that nobody really cares if FPS is 200 or 250. So generally Ryzen is no bottleneck for games. It may seem so if cherry picking right benchmarks, otherwise no.

Roll forward 2 years, when the GTX1080ti becomes a midrange GPU, and you get substantially faster GPU, guess what, the GPU bottleneck is not going be there from the GPU at 4K, but that GPU bottleneck will be imposed by the CPU, specifically Ryzen will do worse for it. This is fact, you can SLI GTX1080ti now and take a peak at that future.

Not at all. In two years we have more and more Vulkan/DirectX12 games and guess what, those games utilize more cores much better than today's antique DirectX 11 games. So there will not be worse CPU bottleneck than what is now, in fact much smaller one. You are trying to make way too much opinions from simple benchmarks. That simply does not work.

SLI/Crossfire systems are out of luck as both Nvidia and AMD do not really care about them any more and much less in future.
 
You obviously don't even understand what bottleneck means.

blah blah blah


Not at all. In two years we have more and more Vulkan/DirectX12 games and guess what, those games utilize more cores much better than today's antique DirectX 11 games. So there will not be worse CPU bottleneck than what is now, in fact much smaller one. You are trying to make way too much opinions from simple benchmarks. That simply does not work.

SLI/Crossfire systems are out of luck as both Nvidia and AMD do not really care about them any more and much less in future.

You do NOT understand what bottleneck means. GPU bottlenecked by the CPU is SAD, and Ryzen is more liable to do it. You think software development will fix the multi-thread problems in two years.... dream on buddy. How long have they been saying this? Since the introduction of the first Core 2 Duo, or Athlon64 X2, back in 2005. Why is it still a thing in 12 years later in 2017?

All I hear are excuses. 1080p so fast it won't matter... So Ryzen is still significantly slower, and that CPU imposed GPU bottleneck will only become a bigger problem with faster GPUs. No one cares about multi-card setups... that is also an excuse.

What happened to AMD making no excuses? Either 1 provide unrivaled performance across the board no excuses, or 2 provide unrivaled value by pricing better than anything Intel can do no excuses? AMD has only started to come close with the release of the R3.
 
Lol, hardreset you're like the biggest AMD fanboy I've ever seen.

I don't think it's dumb to buy an i5 in 2017 at all. Especially if you're a gamer.

I do think it is dumb to buy an R5/R7 and pretend it's "futureproof" because it has more cores. More cores don't make a chip future proof in my opinion. It's not like any of the older 8 core chips are relevant today. To me the results are obvious, gamers should be buying Intel, they top all the gaming benchmarks. Some of you might think that this will flip round in the future but I don't think so. I can't see games needing such a large amount of processing power to require a multithreaded CPU anytime soon.

There is value in Ryzen for gamers though, if you don't have the cash for an i5/i7 then an R5 or an R3 is probably better bang for your buck. But personally I'd rather have the best for a few dollars more.
 
You do NOT understand what bottleneck means. GPU bottlenecked by the CPU is SAD, and Ryzen is more liable to do it. You think software development will fix the multi-thread problems in two years.... dream on buddy. How long have they been saying this? Since the introduction of the first Core 2 Duo, or Athlon64 X2, back in 2005. Why is it still a thing in 12 years later in 2017?

You provided "examples" how Ryzen causes GPU bottleneck but I proved that GPU is far from bottlenecked on those situations, so your claims are false.

DirectX 9 was released 2003, because DirectX 9 has main rendering thread, we can say DirectX 9 performance is mainly determined by single core performance. Same applies to DirectX 11 (2009). So essentially all DirectX software before DirectX 12 are single thread limited, so there you had reason why.

All I hear are excuses. 1080p so fast it won't matter... So Ryzen is still significantly slower, and that CPU imposed GPU bottleneck will only become a bigger problem with faster GPUs. No one cares about multi-card setups... that is also an excuse.

What happened to AMD making no excuses? Either 1 provide unrivaled performance across the board no excuses, or 2 provide unrivaled value by pricing better than anything Intel can do no excuses? AMD has only started to come close with the release of the R3.

Ryzen is not significantly slower. Difference is so small that it doesn't matter at all. That GPU bottleneck is nonexistent and there is no reason why it will become bigger problem, because it won't. AMD or Nvidia no longer care about multi card setups, so that's very good reason.

AMD offers unrivaled performance per watt and CPU temperature, funny that now when Intel CPU's run hot as hell, power consumption and temps no longer matter. AMD value is simply awesome, even Intel noticed that with and are now panicking with new releases.

And while talking about benchmarks, real life gamers agree that 8-core Ryzen runs games like Battlefield 1 smoother than i7-7700K, so who really cares about benchmarks? Nobody.

Lol, hardreset you're like the biggest AMD fanboy I've ever seen.

I don't think it's dumb to buy an i5 in 2017 at all. Especially if you're a gamer.

I do think it is dumb to buy an R5/R7 and pretend it's "futureproof" because it has more cores. More cores don't make a chip future proof in my opinion. It's not like any of the older 8 core chips are relevant today. To me the results are obvious, gamers should be buying Intel, they top all the gaming benchmarks. Some of you might think that this will flip round in the future but I don't think so. I can't see games needing such a large amount of processing power to require a multithreaded CPU anytime soon.

There is value in Ryzen for gamers though, if you don't have the cash for an i5/i7 then an R5 or an R3 is probably better bang for your buck. But personally I'd rather have the best for a few dollars more.

That i5 vs i7 was taken from article linked above. As stated above, 8-core Ryzen run games smoother than quad core i7. Once again more proof that benchmarks are nearly useless when talking about real world performance. Because very few gaming any benchmark. And so any "proof" how Intel is better because benchmarks is invalid.
 
...
That i5 vs i7 was taken from article linked above. As stated above, 8-core Ryzen run games smoother than quad core i7. Once again more proof that benchmarks are nearly useless when talking about real world performance. Because very few gaming any benchmark. And so any "proof" how Intel is better because benchmarks is invalid.

What proof do have to show that it is smoother? All that is just the placebo effect. The 0.1% low, the 1% low fps all show Ryzen to have lower lows. See more benchmark you like to call invalid:

http://www.gamersnexus.net/hwreview...review-premiere-blender-fps-benchmarks/page-7

This means any smoothness is just the placebo effect your self-delusion induces.
 
What proof do have to show that it is smoother? All that is just the placebo effect. The 0.1% low, the 1% low fps all show Ryzen to have lower lows. See more benchmark you like to call invalid:

http://www.gamersnexus.net/hwreview...review-premiere-blender-fps-benchmarks/page-7

This means any smoothness is just the placebo effect your self-delusion induces.

So my claim about smoothness when gaming cannot be true because benchmark disagrees :D You really should learn difference between benchmarking and gaming.

Additionally your benchmarks are over 5 months old so they have been long time obsolete.
 
So my claim about smoothness when gaming cannot be true because benchmark disagrees :D You really should learn difference between benchmarking and gaming.

Additionally your benchmarks are over 5 months old so they have been long time obsolete.

5 months old does not make it irrelavant or invalid. How much has software changed in the 5 months? If there are new benches that shows something different present it already. Gaming is not benches, but benched results are a 1-to-1 mapping/proxy for gaming performance. All I hear from you is just denials without facts in any way or form. If it is smoother, show it. Smoothness should be quantifiable and demonstrable, and not just some subjective feeling from watching some youtube videos.
 
5 months old does not make it irrelavant or invalid. How much has software changed in the 5 months? If there are new benches that shows something different present it already. Gaming is not benches, but benched results are a 1-to-1 mapping/proxy for gaming performance. All I hear from you is just denials without facts in any way or form. If it is smoother, show it. Smoothness should be quantifiable and demonstrable, and not just some subjective feeling from watching some youtube videos.

How much changed? Quite lot in fact. Ryzen's AGESA is now much better and allows much higher memory clocks and/or tighter memory timings, so game performance is better.

Benched results cannot be 1-1 gaming performance, because benchmarks are missing user input, background software etc. User input is very important as benchmarks are totally missing it. If you can find proof that Ryzen does not offer smoother gaming experience, then show it. With benchmarks you can never do it, no matter how hard you try.
 
How much changed? Quite lot in fact. Ryzen's AGESA is now much better and allows much higher memory clocks and/or tighter memory timings, so game performance is better.

Benched results cannot be 1-1 gaming performance, because benchmarks are missing user input, background software etc. User input is very important as benchmarks are totally missing it. If you can find proof that Ryzen does not offer smoother gaming experience, then show it. With benchmarks you can never do it, no matter how hard you try.

I will actually back you up on this - almost every review of Ryzen I have ready (Both professional and otherwise) states that Ryzen systems feel quite a bit smoother.

I have a 6700K, and I don't seem to have as many of the slowdowns as my friends with i7's. However I have overclocked the RAM to 3600MHz with CAS-17-28 timings, overclocked the BCLK, and overclocked the cache. Combine that with making sure Chrome is closed while gaming, and there aren't any issues. Having said that most people don't get as lucky with their ram and cache as I did, and even my friend with a 20-thread Broadwell-E Xeon says it hiccups in BF1 sometimes.
 
I will actually back you up on this - almost every review of Ryzen I have ready (Both professional and otherwise) states that Ryzen systems feel quite a bit smoother.

....

Show the actual reviews with actual data? Failing that this is just subjective opinion. Smoothness something that can be measured. Demonstrate the measurements or stop with the false claims.
 
Great conclusion. You have to be stupid to buy anything but Ryzen at almost all the price points. For 300 bucks R7 1700 is amazing value for money and really all that you could need for a few years on.
I just hope that AMD will have the same bang for buck with TR line-up, because as it stands now you get ~ same performance as Intel but at half the price for the entire platform. From initial benchmarks, it seems that 1920X will match 7900X but costing less.

You have to be very stupid or very childish if you tell someone what to do without knowledge... Ryzen is very limited with memory, if you need to have more than 32gb memory will fall at 2133 or 2400mhz in lucky cases.
Overall the system is limited to 64(hypothetical)gb the x299 can handle 128gb without any problem.
Yes you can use threadripper but we are speaking about a system really expensive, a motherboard is more than 400$ an x299 is less than 200$.
There is people that use the pc to work or for heavy tasks in general, not only video games...
 
You have to be very stupid or very childish if you tell someone what to do without knowledge... Ryzen is very limited with memory, if you need to have more than 32gb memory will fall at 2133 or 2400mhz in lucky cases.
Overall the system is limited to 64(hypothetical)gb the x299 can handle 128gb without any problem.
Yes you can use threadripper but we are speaking about a system really expensive, a motherboard is more than 400$ an x299 is less than 200$.
There is people that use the pc to work or for heavy tasks in general, not only video games...
I'm not sure what you think you're looking at but there's no reason 64 gig can't be had easily. Or at faster speeds, quite a few 3kmhz boards from asrock, gigabyte etc at 64gig ram.
 
Back