CPU Ugrade

[Cammaron]

I’m looking to upgrade my sons cpu in his gaming rig and looking for advice.

 

How much of a performance upgrade would he get when upgrading from a Ryzen 3 1200 to a Ryzen 5 3600x.

 

I’ve already upgraded his GPU to a 1080GTX G1 and increased the ram to 16gb of 2600. Also planning on upgrading his drive from a 7200rpm HDD to a NVME this week.

 

Before anyone gives me crap I built the computer 4 years ago when he was 10 and he is now 14 and his gaming requirements have definitely increased. Just not ready to pull the trigger on a whole new motherboard yet.

[Skiiwee29]

Would be a decent upgrade to go from a 1200 to 3600x. The 1200 is likely bottlenecking the 1080 and this should allow it to stretch its legs properly now. 

[tkitch]

As long as you don't get fleeced buying a 3600?  That's a really significant upgrade.  2 generations worth of IPC upgrades, as well as several extra cored.

[Cammaron]

1 minute ago, Skiiwee29 said:

Would be a decent upgrade to go from a 1200 to 3600x. The 1200 is likely bottlenecking the 1080 and this should allow it to stretch its legs properly now. 

Yeah the CPU and drive are both the bottleneck atm. I’m just trying to decide wether to buy that or set aside a little more money and go a little better. Wanted to see what the consensus was on the interweb of things was. Lol

[CT854]

Any reason you're going after a 3600X instead of a 3600? I'm skeptical the extra money for the X SKU justifies a small performance increase.

[Cammaron]

2 minutes ago, tkitch said:

As long as you don't get fleeced buying a 3600?  That's a really significant upgrade.  2 generations worth of IPC upgrades, as well as several extra cored.

Yeah I’ll probably buy off New Egg. After all this my son will have a better rig than mine….. might have to build me a new one next lol

[Cammaron]

1 minute ago, CT854 said:

Any reason you're going after a 3600X instead of a 3600? I'm skeptical the extra money for the X SKU justifies a small performance increase.

Well that’s the other thing I’m debating. $80 price difference roughly for 200mhz more performance.

[CT854]

Just now, Cammaron said:

Well that’s the other thing I’m debating. $80 price difference roughly for 200mhz more performance.

I'd ditch the X just for the 3600. Look through different reviews from when the 3600X came out as many of them make direct comparisons with the base 3600 for a second opinion but I think the better value is just in the 3600.

 

Also, what's your board model? If you have a bottom of the barrel A320/B350 board I'd imagine there'd be some consideration as to whether power delivery is strong enough to max out a 3600 on ambient cooling. Happy to get a second opinion on this, though.

 

 

[SirWendall]

Or go with a 5600x? I know its a $100-$120 increase from the 3600 (non x), but I'm liking my performance gains on it. Also depends on what games are going to be played too.

[GhostRoadieBL]

https://pc-builds.com/calculator/Ryzen_3_1200/GeForce_GTX_1080/0MR0Uu8A/16/100/

 

CPU is a pretty stiff bottleneck but still playable for most games

 

https://pc-builds.com/calculator/Ryzen_5_3600/GeForce_GTX_1080/0U90Uu8A/16/100

3600 would be a perfect upgrade, even a 2600X is a step up but if you can get a good price for a 3600 that's what I would go to

[Lurick]

7 minutes ago, SirWendall said:

Or go with a 5600x? I know its a $100-$120 increase from the 3600 (non x), but I'm liking my performance gains on it. Also depends on what games are going to be played too.

If it's a 300 series board (which I suspect) then you're not getting a 5000 series CPU on it unless you happen to find and be okay with a super beta buggy BIOS for the few boards that have them.

[Fasauceome]

12 minutes ago, GhostRoadieBL said:

https://pc-builds.com/calculator/Ryzen_3_1200/GeForce_GTX_1080/0MR0Uu8A/16/100/

 

CPU is a pretty stiff bottleneck but still playable for most games

 

https://pc-builds.com/calculator/Ryzen_5_3600/GeForce_GTX_1080/0U90Uu8A/16/100

3600 would be a perfect upgrade, even a 2600X is a step up but if you can get a good price for a 3600 that's what I would go to

Bottleneck calculators don't really mean anything

[GhostRoadieBL]

6 minutes ago, Fasauceome said:

Bottleneck calculators don't really mean anything

you have anything to support that?

site provides a quick view of fps with thousands of combinations of parts to compare. The percentage numbers are calculated from those results, how is this worse than digging through reviews with completely different parts and parameters?

 

It means a lot more than a bunch of random anecdotal opinions on a forum lol

[Fasauceome]

37 minutes ago, GhostRoadieBL said:

you have anything to support that?

site provides a quick view of fps with thousands of combinations of parts to compare. The percentage numbers are calculated from those results, how is this worse than digging through reviews with completely different parts and parameters?

 

It means a lot more than a bunch of random anecdotal opinions on a forum lol

testing methodology not clear, the result given as a percentage means nothing in terms of real world performance. Where is this data even coming from? Their home page has some news articles, no explanation of their procedures. When you use a tool like their comparison index, they give BS like a "future proof score" which is nonsense. They have all these games listed in the comparison "data" but do not share the system configuration, operating system version, game patch version, and other relevant data that a comprehensive comparison would provide.

 

it means significantly less than anecdotal experience. You would obtain an equally useful number by using a random number generator.

[GhostRoadieBL]

1 minute ago, Fasauceome said:

testing methodology not clear, the result given as a percentage means nothing in terms of real world performance. Where is this data even coming from? Their home page has some news articles, no explanation of their procedures. When you use a tool like their comparison index, they give BS like a "future proof score" which is nonsense. They have all these games listed in the comparison "data" but do not share the system configuration, operating system version, game patch version, and other relevant data that a comprehensive comparison would provide.

 

it means significantly less than anecdotal experience. You would obtain an equally useful number by using a random number generator.

so anecdotal experience, have you personally owned a 1080 and r3 1200 and then upgraded to a 3600? or have a system running a 3600 and 1080? what other parts are in your system which effect the performance, no two systems in the forum are identical so rough numbers from a convenient website where everything is laid out in fps and you can easily punch in parts to see roughly what the performance will be is far more than any individual or group on a forum can provide.

The layer of assumption that comes from distrusting a website without "all" the information cluttering it's pages should be equal or greater for those on a forum who don't own or list their system specs and update version etc etc etc. If you don't personally own the hardware, you are making an educated estimate at the performance based on reading about it or third party information on how it performs compared to x, y , or z other hardware. This is the same as using a bottleneck calculator (slightly less so since at least the calculator puts a number value to "it'll work great in your system")

[Fasauceome]

5 minutes ago, GhostRoadieBL said:

The layer of assumption that comes from distrusting a website without "all" the information cluttering it's pages

I would rather have all the information than not have any.

 

Again, this website's testing methodology is non existent and their data is therefore worthless.

[GhostRoadieBL]

1 minute ago, Fasauceome said:

I would rather have all the information than not have any.

 

Again, this website's testing methodology is non existent and their data is therefore worthless.

neither is anyone's here on the forum, thus their opinions on the hardware are therefore worthless too?

[Fasauceome]

Just now, GhostRoadieBL said:

neither is anyone's here on the forum, thus their opinions on the hardware are therefore worthless too?

You always take the risk of getting advice from someone on a forum who proports to know more than they really do. However, the forum offers something that these worthless calculators do not:

 

A chance to explain your reasoning. Any one of us users can dig into the meat of the question, ask about the software being used and provide an explanation for hardware suggestions. Is the CPU underpowered? Why? What would help? Generalist recommendations are not a problem because the advice isn't about the exact system performance down to the last pixel calculation. A percentage attached to a bottleneck just isn't a useful figure compared to someone saying why a CPU with more cores or faster clocks might be better in a specific case.

[SirWendall]

24 minutes ago, Fasauceome said:

You always take the risk of getting advice from someone on a forum who proports to know more than they really do. However, the forum offers something that these worthless calculators do not:

 

A chance to explain your reasoning. Any one of us users can dig into the meat of the question, ask about the software being used and provide an explanation for hardware suggestions. Is the CPU underpowered? Why? What would help? Generalist recommendations are not a problem because the advice isn't about the exact system performance down to the last pixel calculation. A percentage attached to a bottleneck just isn't a useful figure compared to someone saying why a CPU with more cores or faster clocks might be better in a specific case.

Exactly. No one on here is going out of there way to cite bottleneck calculator sites. Its like Wikipedia in a sense. @GhostRoadieBL

[GhostRoadieBL]

36 minutes ago, Fasauceome said:

A chance to explain your reasoning. Any one of us users can dig into the meat of the question, ask about the software being used and provide an explanation for hardware suggestions. Is the CPU underpowered? Why? What would help? Generalist recommendations are not a problem because the advice isn't about the exact system performance down to the last pixel calculation. A percentage attached to a bottleneck just isn't a useful figure compared to someone saying why a CPU with more cores or faster clocks might be better in a specific case.

 

10 minutes ago, SirWendall said:

Exactly. No one on here is going out of there way to cite bottleneck calculator sites. Its like Wikipedia in a sense. @GhostRoadieBL

given the premise you seem to have [where "bottleneck calculators" or sites which host comparison data between systems without every nit picked piece of data about testing for fps] are terrible and provide zero information for a system builder, or should not be trusted to compare part performance accurately, where do you source your information?

My premise is these sites provide as accurate information as individuals who cannot have first hand information on every configuration, especially to someone asking a "this vs that part" question. Obviously the information didn't come from thin air so the volume of information is going to be more accurate, even more so for fps results. Using the site as a tool coupled with the educated opinion of an enthusiast is far better than taking an average of opinions across any number of people with zero back-up information.

 

I'd like to break down those questions to sources,
How do you know the CPU is under powered - a hunch? estimated off your personal experience? based on reviews others have done with vastly different systems than the OP?

Why - Rhetoric about the company or generation of the part? personal experience with the part? anecdotes from other users making up a fraction of a percent of the owners?

What would help - general assumption of what the OP "needs" or "should have" or "fits the budget"? opinion on "best" hardware for the price?

 

what I'm getting at is a lack of information vs information on a comparison site is on equal level.

The psychology behind your opinion of these sites is one source is from an individual you can interact with while the other is from a faceless webpage you can't ask more information from. People are hard wired to trust presented information from individuals and distrust convenient sources.

@SirWendallfor a general overview of information for the uneducated individual asking a question, is wikipedia the wrong way to provide a low detail overview? I'm sure we can agree no one sources much of anything on a forum, even less when it's a "this specific combo vs that combo" type questions. Putting low detail data behind an opinion is still better than simply stating "this is what I think. however I'm not telling you my credentials to why you should trust me"

 

 

at the end of the day have you tested your personal system's performance vs the site? I have, it's within a few percent in every system I have ever run. (going on 15+ now) no one should be chasing the "under 1% bottleneck" on the site but anything under 10 you're going to be just fine. Well within margin of error on even GN's fps charts.