Question I currently have a Ryzen 1600 msi b350 carbon GTX strix 1070 would it be worth to stck in a ryzen 3500 n just upgrade the bios?

mchiani123

Commendable
Sep 29, 2017
4
0
1,510
My ryzen 5 1600 is working fine with my gtx 1070 16gb 3200 ram, i was just wondering if taking advantage of the holiday sales that i could get me a brand new cpu andn just plug it into my current mobo msi b350 carbon I kno i gotta upgrade the bios no biggie im just curious if it would work and if its worth doing naturally if there are problems then ill just save for couple months and get atleast a b450 board and suggestions?
 
Ryzen 3500 is ONLY available in prebuilt OEM machines. It is not available to the general public. The lowest current Ryzen 5 3000 series is the R5 3600. That, would definitely be a major upgrade from your R5 1600.

To the tune of about a 54% increase in single core performance and about a 61% increase in multithreaded performance. Should work fine with your B350 Gaming Pro carbon but you WILL need to have the latest BIOS version or at least one new enough to support the 3000 series SKUs.
 
  • Like
Reactions: mitch074

InvalidError

Titan
Moderator
The 3600 would only be about 20% faster, won't necessarily be noticeably better unless your 1600 is "just barely not quite cutting it" in some cases. I'd say either wait for the 4600 to launch and upgrade to that. By then, a significantly discounted 3700X may also be an option.
 
Not from what I see. I'm not sure where you see that it's only 20% faster but all of the comparisons I've seen show exactly or close to the numbers I posted above. 3600 has much better IPC and single core performance than that 1st Gen Ryzen.

I realize this isn't a perfect real world comparison, but it's usually pretty accurate in general terms, comparatively speaking. As accurate as anything else out there and likely a lot more accurate than userbenchmark.

https://www.cpubenchmark.net/compare/AMD-Ryzen-5-1600-vs-AMD-Ryzen-5-3600/2984vs3481

Based on those numbers it's a a 54% increase in single core performance and about a 61% increase in multithreaded performance.
 

InvalidError

Titan
Moderator
Not from what I see. I'm not sure where you see that it's only 20% faster but all of the comparisons I've seen show exactly or close to the numbers I posted above. 3600 has much better IPC and single core performance than that 1st Gen Ryzen.
Much better, sure. But not 60% better. You only get that sort of figure from benchmarks that greatly over-emphasize Zen 2's advantages in areas where Zen 1 was particularly weak. If you look at games which represent a broader load mix, 15-25% is more typical.
 
You're right, those are somewhat misleading numbers. I don't think these are though.


Cinebench R20 multi - 44% increase

Cinebench R20 singe - 42% increase

Winrar - 104% increase

7zip - 40% increase

Premier Pro - 33% increase

Vray - 32% increase

Blender - 32% increase

Games ( Using 1% low min FPS, not max FPS)

Assassin's creed odyssey 1080p - 35% increase

Battlefield V - 15% increase

Shadow of the tomb raider - 40% increase

The division 2 - 35% increase

Far cry new dawn - 21% increase

Hitman 2 - 36% increase

Total war 3 kingdoms - 50% increase

Average increase: 39%

Source: https://www.techspot.com/review/1871-amd-ryzen-3600/



At 1080p, the R5 3600’s stock AVG FPS shows a ~7% improvement over the 2600 overclocked to 4.2GHz and ~33% over the stock 1600, beating out the full range of older 6C/12T AMD parts, overclocked or not.

https://www.gamersnexus.net/hwreviews/3489-amd-ryzen-5-3600-cpu-review-benchmarks-vs-intel

Even a 30% increase is pretty phenomenal for a two generation increase. That's 15% per generation. I can't recall the last time Intel had two 15% increases in performance in a row.
 
This video shows a comparison of the 1600, 2600 and 3600. It's done on different motherboards though, but same memory and RTX 2080.
View: https://www.youtube.com/watch?v=wJguDHmytaE
I see about a 15-25% difference in fps at what appears to be 1080p, which for some of the games tested means 30-40+fps increase on 3600 over the 1600. If you are playing a Battlefield 5 and getting say 130fps on the 1600, you may get 170-180fps on the 3600.

Whether the upgrade is worth it to you depends on the games you play. Look for benchmarks or fps data. Youtube usually has at least some videos of fps results during game play.
 
The difference will be COMPLETELY different, depending on what type of game it is. Games that are very CPU intensive are obviously going to see much larger increases in performance going from the 1600 to the 3600. Games that are mostly GPU bound will see MUCH smaller changes. There are some games in some of the comparisons I looked at that had less than a 10% difference. Obviously those are games where the CPU is not the limiting factor. In games where the 1600 IS the limiting factor, the increases are going to be much larger.

But it is also worth keeping in mind that most of these gaming and performance related reviews are not done on somebody's daily driver, real world type system. They are done on machines with the bare essentials required to run the testing and produce the results on. For systems where people are potentially running a higher degree of background processes from installed applications, do not have any of the many windows services disabled because they are not needed on a test machine, are not also streaming, or browsing, or recording, or any of a variety of other tasks, then there could be some additional gains in performance there where normally a faster CPU might not show much in the way of performance increase but does because it is more easily able to handle those additional background or multitasking processes.

In other words, lab machine versus YOUR machine, does not usually or necessarily result in the same type of results.
 

InvalidError

Titan
Moderator
Waiting another four to six months, or longer, knowing how these hardware launches are generally delayed, probably isn't an option or desirable, for many people. If you need an upgrade now, I say anything from the 3600 to the 3700x is a good choice.
It really does boil down to how "desperate" for an upgrade you are. In OP's case though, the upgrade is mostly a got-spare-change 'itch' thing, not a 'need' thing.

I know that when I was pulling my hair out on my Core2 because I was doing stuff that needed 20GB of RAM on 8GB of RAM, upgrading to toss 32GB in my system (16GB initially and an extra 16GB shortly thereafter when I realized 16GB wasn't enough to get me the total swapping relief I craved at the time), nobody would have been able to convince me to upgrade later :)
 
It really does boil down to how "desperate" for an upgrade you are. In OP's case though, the upgrade is mostly a got-spare-change 'itch' thing, not a 'need' thing.

I know that when I was pulling my hair out on my Core2 because I was doing stuff that needed 20GB of RAM on 8GB of RAM, upgrading to toss 32GB in my system (16GB initially and an extra 16GB shortly thereafter when I realized 16GB wasn't enough to get me the total swapping relief I craved at the time), nobody would have been able to convince me to upgrade later :)

Very true.
 

thekillerx10

Honorable
Apr 12, 2018
538
38
10,990
My ryzen 5 1600 is working fine with my gtx 1070 16gb 3200 ram, i was just wondering if taking advantage of the holiday sales that i could get me a brand new cpu andn just plug it into my current mobo msi b350 carbon I kno i gotta upgrade the bios no biggie im just curious if it would work and if its worth doing naturally if there are problems then ill just save for couple months and get atleast a b450 board and suggestions?
upgrading to 3600 isn't worth it wait for 4th gen ryzen
 
That too is subjective. If somebody is running a 1080p high refresh rate configuration they might necessarily want a CPU that can provide as many FPS as possible, but not have any need for a higher tiered graphics card because they are already doing Ultra setting at that resolution with the graphics card they have, in this case, the 1070.
 
That too is subjective. If somebody is running a 1080p high refresh rate configuration they might necessarily want a CPU that can provide as many FPS as possible, but not have any need for a higher tiered graphics card because they are already doing Ultra setting at that resolution with the graphics card they have, in this case, the 1070.
A CPU upgrade won't help much when the only thing that will increase fps is a faster GPU, which is the case for quite a lot of newer games. I think you are over estimating your own percentage numbers you posted above. I'm seeing way lower numbers than you posted for the 1600 vs the 3600 for just the games at 1080p high/ultra settings.

Assassins Creed Odyssey 26%
Battlefield V 14%
Shadow of the Tomb Raider 27%
The Division 2 25%
Far Cry New Dawn 15%
Hitman 2 23%
Total War Three Kingdoms 33%

Edit - Changed the numbers to reflect the 1% averages when I previously had average frame rate only.
Edit 2 - My math is wrong here because I wasn't paying attention while tired, but I'm leaving it because it's my own shame to bear.


Using my own Ryzen 5 2600 and GTX 1070 playing New Dawn with ultra settings at 1080p, I saw around 70-80fps during most of my playtime in the game with lows of mid 50s and highs over 100. That is only a 10fps difference compared to my i7 2600k and i7-4770s. Upgrading to a Ryzen 5 3600 would mean closer to a 20fps difference over my i7-2600k. This is why I previously said that people need to look at the benchmark and fps data for the games they will be playing before making any decision on an upgrade. Upgrading one part won't always be the only thing they need to upgrade. If you already have a good CPU to start with, a high-end graphics card is usually a better upgrade path since it can be moved to a new platform in the future.
 
Last edited:
Your math is flawed. But whatever. Your reasoning seems flawed as well. At 1080p there is little chance any game is going to be GPU limited with a GTX 1070. That's traditionally been a 1440p card and maxes out games at Ultra settings on 1080p.

CPU limits FPS, not GPU, unless you are terribly underpowered graphically. That's pretty much across the board for all scenarios.

It won't matter if you have a 2080 ti, if you have an R5 1600 you're going to have a hard time with any high refresh rate/high FPS gaming unless it's a potato type game. GPU is normally the limiting factor for quality settings, not FPS.
 
Your math is flawed. But whatever. Your reasoning seems flawed as well. At 1080p there is little chance any game is going to be GPU limited with a GTX 1070. That's traditionally been a 1440p card and maxes out games at Ultra settings on 1080p.

CPU limits FPS, not GPU, unless you are terribly underpowered graphically. That's pretty much across the board for all scenarios.

It won't matter if you have a 2080 ti, if you have an R5 1600 you're going to have a hard time with any high refresh rate/high FPS gaming unless it's a potato type game. GPU is normally the limiting factor for quality settings, not FPS.
Yeah my math is wrong, you were right before. I was just being my own big dumb self doing math when tired. I was using the higher number instead of lower when dividing. Assassins Creed should have been 20 divided by 57 and so on... 0.3508
 

DSzymborski

Titan
Moderator
Your math is flawed. But whatever. Your reasoning seems flawed as well. At 1080p there is little chance any game is going to be GPU limited with a GTX 1070. That's traditionally been a 1440p card and maxes out games at Ultra settings on 1080p.

CPU limits FPS, not GPU, unless you are terribly underpowered graphically. That's pretty much across the board for all scenarios.

It won't matter if you have a 2080 ti, if you have an R5 1600 you're going to have a hard time with any high refresh rate/high FPS gaming unless it's a potato type game. GPU is normally the limiting factor for quality settings, not FPS.

In addition to what DB said, there are other things besides gaming. I played Borderlands 3 today, but I also did some nonparametric modeling using a very large dataset.
 
I've said my opinions already and even managed to make a fool of myself with easy math. Whether anyone agrees with my opinions is up to them. I wrote a few paragraphs of opinion and personal observations I almost posted, but I think I'm just gonna let other take over and just go to bed to refresh myself.