Intel 10th gen beats Ryzen in gaming by 40-60 fps when using high speed ram...

[Space Lynx]

Why is no one talking about this? It's absolutely insane the gains...

@R-T-B

Am I missing something? Why are the gains so insane???? Ryzen is history... I mean yeah the ram costs $200 versus $100 for the same capacity, but small price to pay for 40-60 fps gains across the board...


timestamped:

gamersnexus review of high speed ram with intel i5-10600k beats ryzen 3700x by 60-70 fps in many games... and I can't believe no one is talking about that. its absolutely insane imo.
timestamped:

 

[EzioAs]

Maybe you linked the wrong source because I don't see how you reach the "high speed ram with intel i5-10600k beats ryzen 3700x by 60-70 fps in many games" conclusion. Emphasis on the 60-70 fps in many games.

The biggest gap I can see is with F1 2019 (1080p w/high settings) and even then, it's 51 fps at most. Enlighten me if I'm wrong.

 

[Space Lynx]

Maybe you linked the wrong source because I don't see how you reach the "high speed ram with intel i5-10600k beats ryzen 3700x by 60-70 fps in many games" conclusion. Emphasis on the 60-70 fps in many games.

The biggest gap I can see is with F1 2019 (1080p w/high settings) and even then, it's 51 fps at most. Enlighten me if I'm wrong.

i meant ryzen 3600 - for example shadow of tomb raider i5-10600k and high ram = 178 fps, on ryzen 3600 ddr4 3200 cas 14 its 40 fps difference. i guess i read it wrong, but still 40-50 fps doesn't change my point... insane gains...

i mean $100 more for 4000 cas 15 ram over cas 14 3200 ram, and $75 more for a 10600k vs a 3600, mobo costs are about same, MSI Z490 A Pro is $160 and has great vrm cooling.... we pay what $200-400 premium's on graphics cards for 15-30 fps gains...

 

[phanbuey]

People do know this, and have for a while - this has been true since the 8700K / 8600K launch. 10th Gen is basically exactly the same crap. High end/low latency ram on coffee lake would stomp ryzen up and down in academic gaming numbers. The main issue is ppl dont care as much when CPU bound vs GPU bound...

Take your shadow of the tomb raider example:
You would only notice this IF you had a 1080P 165hz+ screen and you were running a high end card, and you cared enough to notice those FPS.

If any of that scenario changes: i.e. If you're on a 1440P screen, or you're running a budget build with a lesser vcard or slower ram even on a 120hz 1080p monitor you don't care or feel that 40 FPS on the CPU at all. Those extra FPS never really amount to that much irl.

When it comes to GPU it's a different story, there you're paying for 15% FPS more in say something like 4K or more eyecandy, or whatever, so it's more noticeable than 40 fps when you're already pushing 140 because it allows you to experience more.

 

[EzioAs]

Should also mentioned that the i5-10600K is OC'ed to 5.1GHz. I mean, yeah it's nice to look at but it doesn't really change what a lot of us know about the current state of CPUs from both sides. If you focus primarily on gaming, the i5-10600K is the best gaming CPU period. That was the conclusion reached when it originally launched and tested by many reviewers. Overclock it and put higher speed memory (which I'm not sure at what speed are the 10th gen CPUs more comfortable at for daily use), it becomes even faster, no surprises there. However, like many reviewers have noted, if you do more than just gaming and want to focus more on productivity workloads, Ryzen 3000 CPUs are usually the better performer/value (outside of the XT variants of course).

I guess why most people here don't bother adding more to what's shown in the video above is because these aren't really that surprising, although it might be to some people. In the end, people buy what fits their need and budget, and if they're happy with what they choose, not much more to add is there?

 

[Kanan]

Intel has academically relevant FPS advantage in games, as in, barely relevant, while having a inferior architecture with security problems and no PCI-E 4.0 support which may very well be relevant soon, coming gen graphics cards. Plus, Intels CPUs are awfully inefficient, especially when overclocked. And 170 bucks extra, may not be much money for you, but for the most people it is rather significant.

 

[Space Lynx]

People do know this, and have for a while - this has been true since the 8700K / 8600K launch. 10th Gen is basically exactly the same crap. High end/low latency ram on coffee lake would stomp ryzen up and down in academic gaming numbers. The main issue is ppl dont care as much when CPU bound vs GPU bound...

Take your shadow of the tomb raider example:
You would only notice this IF you had a 1080P 165hz+ screen and you were running a high end card, and you cared enough to notice those FPS.

If any of that scenario changes: i.e. If you're on a 1440P screen, or you're running a budget build with a lesser vcard or slower ram even on a 120hz 1080p monitor you don't care or feel that 40 FPS on the CPU at all. Those extra FPS never really amount to that much irl.

When it comes to GPU it's a different story, there you're paying for 15% FPS more in say something like 4K or more eyecandy, or whatever, so it's more noticeable than 40 fps when you're already pushing 140 because it allows you to experience more.

this makes sense, plus @Mussels pointed out to me the test was done with 2080 ti which skews everything a bit more since its an unrealistic gpu for most people. and that is true, I would have like to see the tests done with a 2070 super stock or something to see if the gains are as wide, and more 1440p tests, cause if you have that rig and you are gaming at 1080p you are doing it wrong, lol

still I find it impressive that ram can make so much in gains... its really impressive how far ram has come, i remember in ddr3 days when I built my first PC from ground up, my buddy just told me ram is ram try not to worry about it, and I think that was true back then, not so much anymore.

 

[Kanan]

i remember in ddr3 days when I built my first PC from ground up, my buddy just told me ram is ram try not to worry about it, and I think that was true back then, not so much anymore.

The first days I remember Ram being more significant, was with DDR (1), where I actually bought my first high quality kit for my back then Athlon 64 3800+. It had to be high quality, because via FSB overclocking (multi was locked on the regular CPUs), the Ram as well was overclocked by 20% to 240 MHz. The CPU then ran on 2400 instead of 2000 MHz. The Ram even did as much as 280 MHz without problems. Another good kit I bought was then later for the Phenom II 940, a DDR2 kit with 800 MHz and CL9 timings.

This in contrast to the first days I had a PC, where I would find spare EDO Ram pieces laying around in my fathers office and simply take and insert it in the empty slots of my, back then, Asus Socket 7 board, which had only 2 of 6 slots populated. Was a nice increase from 32 MB to 196 MB back then.

 

[Space Lynx]

The first days I remember Ram being more significant, was with DDR (1), where I actually bought my first high quality kit for my back then Athlon 64 3800+. It had to be high quality, because via FSB overclocking (multi was locked on the regular CPUs), the Ram as well was overclocked by 20% to 240 MHz. The CPU then ran on 2400 instead of 2000 MHz. The Ram even did as much as 280 MHz without problems. Another good kit I bought was then later for the Phenom II 940, a DDR2 kit with 800 MHz and CL9 timings.

This in contrast to the first days I had a PC, where I would find spare EDO Ram pieces lying around in my fathers office and simply take and insert it in the empty slots of my, back then, Asus Socket 7 board, which had only 2 of 6 slots populated. Was a nice increase from 32 MB to 196 MB back then.

Did that actually increase FPS in games by large margins though? That is my point here.

 

[Kanan]

Did that actually increase FPS in games by large margins though? That is my point here.

Back then every bit helped. Since 2008 or so, CPU performance is in excess because of the advent of Quad Cores and games actually using more than just one core. 2006 or earlier every bit helped, but reviews with actual data about that, or people generally caring about CAS timings were rather rare. You only wanted higher quality kits for the better OCs, which helped of course. For the Phenom II 940, which was the AM2+ version of a CPU more widely used with DDR3 on AM3 boards, it helped a lot as well, cause you had to counteract the loss of bandwidth with sharp timings, which was the only upside of DDR2 compared to DDR3, aside from the actually payable prices back then.

 

[Space Lynx]

does anyone know how much FPS I would gain at 1440p with everything the same on a more realistic system, say a 2070 super and ryzen 3700x at 1440p high refresh on a realistic setup vs the 2080 ti at 1080p gamersnexus benches show?

and the only variable change would be ram - say realistic ram - 3200 cas 16-18-18 vs 3600 cas 16-16-16

 

[Chomiq]

Why is no one talking about this? It's absolutely insane the gains...

@R-T-B

Am I missing something? Why are the gains so insane???? Ryzen is history... I mean yeah the ram costs $200 versus $100 for the same capacity, but small price to pay for 40-60 fps gains across the board...


timestamped:

gamersnexus review of high speed ram with intel i5-10600k beats ryzen 3700x by 60-70 fps in many games... and I can't believe no one is talking about that. its absolutely insane imo.
timestamped:

You're missing the bit about cpus not running at identical clock speeds. Increased clock speed, higher ram speed, reduced latency and there you have it. Clear advantage in cpu performance.

 

[Assimilator]

I wish you would stop posting these BS clickbait "Intel is bettar than Ryzen OMG WTF LOL" threads. Or at least, read further than the end of your own nose before people point out to you that you should have read further than the end of your nose, before posting said threads.

I would've thought you would've learned from the previous one, but as they say, the definition of insanity...

 

[Space Lynx]

I wish you would stop posting these BS clickbait "Intel is bettar than Ryzen OMG WTF LOL" threads. Or at least, read further than the end of your own nose before people point out to you that you should have read further than the end of your nose, before posting said threads.

I would've thought you would've learned from the previous one, but as they say, the definition of insanity...

huh, did you not read any of my posts? I didn't realize the 2080 ti was the skewing factor and I am now trying to find more realistic results... maybe you should not post on my threads unless you have something valuable to add? welcome to my ignore list. you are of no help.

 

[Kanan]

huh, did you not read any of my posts? I didn't realize the 2080 ti was the skewing factor and I am now trying to find more realistic results... maybe you should not post on my threads unless you have something valuable to add? welcome to my ignore list. you are of no help.

What do all these Youtubers have in common? They all recommend Ryzen, why do you think that is? Simply watch videos to the end, they always add a disclaimer and tell which CPUs are the best, regarding value and performance.

That said, a lot of GamersNexus videos are highly made for entertainment. They do a lot of freakish benchmark runs, that are barely realistic.

 

[Space Lynx]

What do all these Youtubers have in common? They all recommend Ryzen, why do you think that is? Simply watch videos to the end, they always add a disclaimer and tell which CPUs are the best, regarding value and performance.

That said, a lot of GamersNexus videos are highly made for entertainment. They do a lot of freakish benchmark runs, that are barely realistic.

gamersnexus is an an enthusiast (some people enjoy watching that) and trustworthy imo. he has a wide variety of videos catering to many budgets.

 

[Kanan]

gamersnexus is an an enthusiast (some people enjoy watching that) and trustworthy imo. he has a wide variety of videos catering to many budgets.

Didn't say anything to the contrary, I'm just saying, not all his videos are to be taken for realistic world usage, unless he adds his disclaimer at the end, as per usual, that these combinations are highly unlikely to happen. Super expensive Ram and a 1200$ card with a midrange CPU - I don't think so.

 

[Space Lynx]

Didn't say anything to the contrary, I'm just saying, not all his videos are to be taken for realistic world usage, unless he adds his disclaimer at the end, as per usual, that these combinations are highly unlikely to happen. Super expensive Ram and a 1200$ card with a midrange CPU - I don't think so.

I agree, but I didn't realize this until Mussels pointed it out to me, which is why I mentioned it an earlier post (making your comment redundant)

I'm now curious what realistic builds on mid-range budgets look like if you just take say a 3600 stock and then some high end ram paired with mid-range gpu, as mentioned in previous post, I can not find any benches for this and I think it would be interesting to see the results. Could be some interesting gains for only $100 more, or maybe it all levels off because it really was the 2080 ti doing the skewing too much? we won't know without a thorough and controlled study that only people like w1zz and gamersnexus are capable of. also i would love to see some older games tested again for this stuff like witcher 3, why everyone is using F1 racing a game 3% of games play is beyond me.

 

[kayjay010101]

Didn't say anything to the contrary, I'm just saying, not all his videos are to be taken for realistic world usage, unless he adds his disclaimer at the end, as per usual, that these combinations are highly unlikely to happen. Super expensive Ram and a 1200$ card with a midrange CPU - I don't think so.

You really don't understand how benchmarking works, do you? You need to eliminate potential bottlenecks when testing a specific part so that you're actually testing the part. If they built realistic builds that were balanced, they'd 9/10 times just be testing the GPUs when they're intending to test CPUs, since most games are GPU limited. By going with a standardized build that just swaps CPUs you eliminate those variables. Their videos aren't to be taken as purchasing advice for builds, they never advertise themselves as that. Their videos are benchmarks and reviews of single products, They compare single products to other singular products. They aren't there to make build guides for you.

Also, Steve is not the only one at GamersNexus. They have a whole team, Steve is just the face of the channel and the main writer.

 

[Space Lynx]

You really don't understand how benchmarking works, do you? You need to eliminate potential bottlenecks when testing a specific part so that you're actually testing the part. If they built realistic builds that were balanced, they'd 9/10 times just be testing the GPUs when they're intending to test CPUs, since most games are GPU limited. By going with a standardized build that just swaps CPUs you eliminate those variables.

Also, Steve is not the only one at GamersNexus. They have a whole team, Steve is just the face of the channel and the main writer.

ram is a different ball game these days though, are you not curious at all what 3800 cas 15/16 ram can do with a ryzen 3600 and rtx 2070 super at 1440p? i mean this ram is very affordable now, $95 for 3200 cas 14 and then oc it with ryzen calculator. i am really interested to see how much it gains over someone who just went full budget with $60 ram at 3200 cas 16

 

[Kanan]

You really don't understand how benchmarking works, do you? You need to eliminate potential bottlenecks when testing a specific part so that you're actually testing the part. If they built realistic builds that were balanced, they'd 9/10 times just be testing the GPUs when they're intending to test CPUs, since most games are GPU limited. By going with a standardized build that just swaps CPUs you eliminate those variables.

Also, Steve is not the only one at GamersNexus. They have a whole team, Steve is just the face of the channel and the main writer.

I do understand it, don't get angry. My point just is, that nobody will sport this combination (in this topics video), unless he does it for fun. A lot of his videos are interesting and realistic. This one isn't. Also where exactly did I say, he is one guy? I follow this channel regularly, I know that he isn't alone. Nor can he be, with so much work at hand.

 

[kayjay010101]

ram is a different ball game these days though, are you not curious at all what 3800 cas 15/16 ram can do with a ryzen 3600 and rtx 2070 super at 1440p? i mean this ram is very affordable now, $95 for 3200 cas 14 and then oc it with ryzen calculator. i am really interested to see how much it gains over someone who just went full budget with $60 ram at 3200 cas 16

I have a 3700X with 3600CL14-14-14-28 timings. Before that I had 3200 CL16 with the same setup. This was at 1440p. At most I've seen a 3-4 FPS difference between the two, the real difference was in the AIDA64 memory benchmark.

 

[Space Lynx]

I have a 3700X with 3600CL14-14-14-28 timings. Before that I had 3200 CL16 with the same setup. This was at 1440p. At most I've seen a 3-4 FPS difference between the two, the real difference was in the AIDA64 memory benchmark.

see I find those results very interesting, it is only one person though. i would expect 10 fps increase, especially seeing as how it scales to 40-50 fps when you bring in the 2080 ti at 1080p, or maybe 1440p is what is stopping the scaling of the ram performance gains? this is why i wish benchmarkers did budget build tests. something doesn't make sense here it should be at least 10 fps, i know it doesn't make sense to match the 40-50 fps, but i would just think it would be more than that scaled down even

 

[kayjay010101]

I do understand it, don't get angry. My point just is, that nobody will sport this combination (in this topics video), unless he does it for fun.

My point is your point is irrelevant. GN isn't making a suggestion, they're just doing benchmarks to test.

A lot of his videos are interesting and realistic. This one isn't.

I would argue most of their videos aren't realistic. They're supposed to eliminate variables that in the real world would be present.

Also where exactly did I say, he is one guy? I follow this channel regularly, I know that he isn't alone. Nor can he be, with so much work at hand.

You keep referring to their channel as his.

see I find those results very interesting, it is only one person though. i would expect 10 fps increase, especially seeing as how it scales to 40-50 fps when you bring in the 2080 ti at 1080p, or maybe 1440p is what is stopping the scaling of the ram performance gains? this is why i wish benchmarkers did budget build tests. something doesn't make sense here it should be at least 10 fps, i know it doesn't make sense to match the 40-50 fps, but i would just think it would be more than that scaled down even

I don't know if those numbers are right btw, I haven't really done any actual testing. It's just my observation that I haven't noticed any higher FPS numbers after upgrading the RAM, but without doing actual benchmarks my words are naught.

 

[Kanan]

Mt point is your point is irrelevant. GN isn't making a suggestion, they're just doing benchmarks to test.



I would argue most of their videos aren't realistic. They're supposed to eliminate variables that in the real world would be present.


You keep referring to their channel as his.

Calm down. First of all, GN is testing and recommending a lot. And they also do a lot of enthusiast videos for entertainment. Both. My point isn't irrelevant, because the OP thought Ryzen is "so much" inferior, while it really isn't. Real world usage, with normal resolutions, RAM sticks and everyman's GPU's, the difference is abysmal between the CPUs. And then, you can switch from Zen 2 to Zen 3, while with Intel you have a dead end ahead.