Gaming: is a 1600 an upgrade from a 3770k?

[Gdourado]

Hello, how are you? 

 

With games becoming more and more multi threaded, I am wondering if for a gaming rig, a ryzen 1600 is an upgrade from a 3770k? 

 

From what I see on reviews, the 1600 can oc to around 4.0 and use ddr4 2933 or even 3200 if one is lucky. 

 

My 3770k runs at 4.6 with ddr3 running at 2400.

 

How much of an upgrade would it be? 

I game at 1080p 144hz. 

 

Would the 1600 increase both the average fps as well as the minimums? 

Or for 2017, 2018, the 8 threaded i7 with a higher clock speed would still be better? 

 

Cheers and thanks 

[DocSwag]

Not really any bit of an upgrade a 1600 isn't much better than a 3770k.

[ARikozuM]

Keep your CPU until the next gen Intel architecture is released.

[DexterSmythe]

5 minutes ago, DocSwag said:

Not really any bit of an upgrade a 1600 isn't much better than a 3770k.

 

2 minutes ago, ARikozuM said:

Keep your CPU until the next gen Intel architecture is released.

The 1600 would only be 13% faster on average performance according to UserBenchmark. So I agree with you both.

[ARikozuM]

1 minute ago, CapedCrusader21 said:

 

The 1600 would only be 13% faster on average performance according to UserBenchmark. So I agree with you both.

Take UBM as a grain of salt.

[EPPHO]

In gaming it's not an upgrade

[pmos]

if you want more FPS, get a gtx 1080

[Nup]

Bit of an aside, OP, but i got some faster ram recently and have been having trouble keeping my 3770k stable with it. 

 

Did you find anything useful for stabilising your OC with the ram (changing timings, voltages, ect)?

[RAM555789]

25 minutes ago, Gdourado said:

Hello, how are you? 

 

With games becoming more and more multi threaded, I am wondering if for a gaming rig, a ryzen 1600 is an upgrade from a 3770k? 

 

From what I see on reviews, the 1600 can oc to around 4.0 and use ddr4 2933 or even 3200 if one is lucky. 

 

My 3770k runs at 4.6 with ddr3 running at 2400.

 

How much of an upgrade would it be? 

I game at 1080p 144hz. 

 

Would the 1600 increase both the average fps as well as the minimums? 

Or for 2017, 2018, the 8 threaded i7 with a higher clock speed would still be better? 

 

Cheers and thanks 

If you plan to record or stream, yes, just game no.

[paddy-stone]

For gaming it's not worth the upgrade.. if you do any encoding or other CPU intensive work, then yes it might be worth the upgrade, but would be better to grab a 1700 for the extra $70 or so than the 1600 IMO.

[DocSwag]

40 minutes ago, CapedCrusader21 said:

 

The 1600 would only be 13% faster on average performance according to UserBenchmark. So I agree with you both.

Userbenchmark is trash, in reality the gaming performance difference is like 0%

[Gdourado]

2 hours ago, Nup said:

Bit of an aside, OP, but i got some faster ram recently and have been having trouble keeping my 3770k stable with it. 

 

Did you find anything useful for stabilising your OC with the ram (changing timings, voltages, ect)?

Just started with some gskill ripjaws X 2133.

I manually set the voltage to 1.65 on the memory. 

Then manually set ddr3 2400 and the timmings CL10-12-12-31.

 

That's all it took. 

Been rock solid at these settings with the cpu at 4.6 1.297v.

 

Cheers 

[Gdourado]

This machine does nothing else besides gaming. 

That's why I was asking for my use case. 

Just gaming at 1080p 144hz. 

 

Cheers 

[i_build_nanosuits]

2 hours ago, DocSwag said:

Userbenchmark is trash, in reality the gaming performance difference is like 0%

it's trash if you can't read it...

if you actually take your time to read the information that is presented to you and which come from thousands of user submitted results...it's actually stupid good and very informative.

 

For example, take this: http://cpu.userbenchmark.com/Compare/AMD-Ryzen-5-1600-vs-Intel-Core-i7-3770K/3919vs1317

 

it says: Ryzen, 6% faster OVERALL...it's a start

Gaming 81% vs 77%...similar

Ryzen, much faster multi-core speed...how much...58% of it...interesting

Core i7 3770K ... 5% more single-thread performance at stock settings...also interesting.

overclocked benchmarks, now 3770K reach 9% better single-threaded performance.

multi-thread the 1600 ryzen still lead by a whooping 50%.

 

All the rest is the information you should not even read...except CPUPro's comments at the bottom for each chips, this guy is a good and he is always SPOT on in the comments there.

 

SO...Ryzen 1600 vs 3770K? for gaming? it will depend on the games...some games love more CPU threads (although 8 is enough) and some  (more...most) games will react better with extra per-core performance...so, in gaming, these chips perform more or less the same, they trade blows, and perform within margin of error from each other...and who told us that? the 20 000 users who ran the tool and benched and submitted there performance numbers AKA userbenchmark...which is FANTASTIC btw.

[DocSwag]

1 minute ago, i_build_nanosuits said:

it's trash if you can't read it...

if you actually take your time to read the information that is presented to you and which come from thousands of user submitted results...it's actually stupid good and very informative.

 

For example, take this: http://cpu.userbenchmark.com/Compare/AMD-Ryzen-5-1600-vs-Intel-Core-i7-3770K/3919vs1317

 

it says: Ryzen, 6% faster OVERALL...it's a start

Gaming 81% vs 77%...similar

Ryzen, much faster multi-core speed...how much...58% of it...interesting

Core i7 3770K ... 5% more single-thread performance at stock settings...also interesting.

overclocked benchmarks, now 3770K reach 9% better single-threaded performance.

multi-thread the 1600 ryzen still lead by a whooping 50%.

 

All the rest is the information you should not even read...except CPUPro's comments at the bottom for each chips, this guy is a good and he is always SPOT on in the comments there.

 

SO...Ryzen 1600 vs 3770K? for gaming? it will depend on the games...some games love more CPU threads (although 8 is enough) and some  (more...most) games will react better with extra per-core performance...so, in gaming, these chips perform more or less the same, they trade blows, and perform within margin of error from each other...and who told us that? the 20 000 users who ran the tool and benched and submitted there performance numbers AKA userbenchmark...which is FANTASTIC btw.

It only tells you performance in some single threaded bench, 4 core bench, and multi threaded bench. It doesn't actually give any real performance benchmarks in say games, editing, etc. It literally just tells you Ryzen is slightly worse in single threaded performance but faster in multi threaded, which we already know.

 

The information is pretty much useless IMO as it doesn't give you real numbers that actually happen in real world scenarios where stuff could be shifting between using multiple threads a lot.

[i_build_nanosuits]

1 minute ago, DocSwag said:

It only tells you performance in some single threaded bench, 4 core bench, and multi threaded bench. It doesn't actually give any real performance benchmarks in say games, editing, etc. It literally just tells you Ryzen is slightly worse in single threaded performance but faster in multi threaded, which we already know.

 

The information is pretty much useless IMO as it doesn't give you real numbers that actually happen in real world scenarios where stuff could be shifting between using multiple threads a lot.

the fuck you talking about? if i tell you the single-thread performance AND multi-thread performance of a chip both at stock AND peak overclocked...you need NOTHING else to be able to make a conclusion of ''real world'' performance...unless you're a monkey or somehting?

[ARikozuM]

Just now, i_build_nanosuits said:

the fuck you talking about? if i tell you the single-thread performance AND multi-thread performance of a chip both at stock AND peak overclocked...you need NOTHING else to be able to make a conclusion of ''real world'' performance...unless you're a monkey or somehting?

The problem that I have with UBM is that they only do their series of benches. Having one source doesn't mean a lot if there's no real-world stats along with it. Giving a user a UBM page for two chips isn't going to help  when the disparity in sample sizes is close to 20K. Give the OP some reviews of games, rendering potential, and web browsing to complete the overall picture. 

[DocSwag]

5 minutes ago, i_build_nanosuits said:

the fuck you talking about? if i tell you the single-thread performance AND multi-thread performance of a chip both at stock AND peak overclocked...you need NOTHING else to be able to make a conclusion of ''real world'' performance...unless you're a monkey or somehting?

Watch Dogs 2 is a heavily multi threaded game, yet you see Ryzen trailing heavily behind a 7700k.

 

Userbenchmark doesn't give any representation of that.

[i_build_nanosuits]

1 minute ago, ARikozuM said:

The problem that I have with UBM is that they only do their series of benches. Having one source doesn't mean a lot if there's no real-world stats along with it. Giving a user a UBM page for two chips isn't going to help  when the disparity in sample sizes is close to 20K. Give the OP some reviews of games, rendering potential, and web browsing to complete the overall picture. 

so, if your point is that ''less advanced'' users can't directly draw a conclusion on gaming performance if given the informations about single and multi-core performance along with overclocking potential...then YES, there you have a point...but that does NOT equate to userbenchmark being ''trash'' or ''useless'' or whatever...it comes down to the user's habiltity to do what has to be done with the data they are being presented with...i persoally have a VERY good idea of how a CPU with 50% less cores/threads but roughly 10% higher per-thread performance once overclocked will perform in games vs the other one...maybe some other person will not, and i get that.

 

Sry for the confusion...

[i_build_nanosuits]

1 minute ago, DocSwag said:

Watch Dogs 2 is a heavily multi threaded game, yet you see Ryzen trailing heavily behind a 7700k.

 

Userbenchmark doesn't give any representation of that.

ryzen trail being a 7700K in EVERY games...because the 7700K still has the 8 fondamental threads going on and it's 34% faster PER CORE once OVERCLOCKED...

Userbenchmark told me that:

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-5-1600-vs-Intel-Core-i7-7700K/3919vs3647

 a quick glance at this page, one experienced user can tell within 6 seconds that the i7-7700K is a MUCH MORE capable CPU when it comes to gaming, especially if overclocked.

[smokefest]

3 hours ago, Gdourado said:

Hello, how are you? 

 

With games becoming more and more multi threaded, I am wondering if for a gaming rig, a ryzen 1600 is an upgrade from a 3770k? 

 

From what I see on reviews, the 1600 can oc to around 4.0 and use ddr4 2933 or even 3200 if one is lucky. 

 

My 3770k runs at 4.6 with ddr3 running at 2400.

 

How much of an upgrade would it be? 

I game at 1080p 144hz. 

 

Would the 1600 increase both the average fps as well as the minimums? 

Or for 2017, 2018, the 8 threaded i7 with a higher clock speed would still be better? 

 

Cheers and thanks 

not an upgrade. keep the i7 for sure.

[DocSwag]

1 hour ago, i_build_nanosuits said:

ryzen trail being a 7700K in EVERY games...because the 7700K still has the 8 fondamental threads going on and it's 34% faster PER CORE once OVERCLOCKED...

Userbenchmark told me that:

http://cpu.userbenchmark.com/Compare/AMD-Ryzen-5-1600-vs-Intel-Core-i7-7700K/3919vs3647

 a quick glance at this page, one experienced user can tell within 6 seconds that the i7-7700K is a MUCH MORE capable CPU when it comes to gaming, especially if overclocked.

Except Watch Dogs 2 is a heavily multi threaded game; there's a reason why the 7700k beats the 6900k in other games but loses in this one.

 

Which would make you think Ryzen would be very close to a 7700k. Instead it's 25% behind.

 

Multi threaded performance wouldn't have predicted this.

[i_build_nanosuits]

1 hour ago, DocSwag said:

Multi threaded performance wouldn't have predicted this.

no, but single-thread performance wouldn...like i said...if you read the informations that are presented to you.

[DocSwag]

1 hour ago, i_build_nanosuits said:

no, but single-thread performance wouldn...like i said...if you read the informations that are presented to you.

That's not my point!!!

 

Watch Dogs 2 is a HEAVILY multi threaded game, therefore single threaded performance doesn't apply to the game!!!

 

Gaming performance isn't as simple as single threaded and multi threaded performance. Games see a HUGE benefit from overclocking memory on Ryzen, whereas something like Cinebench only sees a 1% boost, max, in both single threaded and multi threaded tests when OCing memory even from 2133 MHz to 3200 MHz.

[ARikozuM]

Move on you two. @DocSwag @i_build_nanosuits