AMD Ryzen 7 1700 Overclocking - Best Ryzen Processor?

[WhoBeDaPlaya]

I absolutely love my 1700. Got the Asus CH6, H110i, and Corsair Vengence 16GB LEDs.

I got it stable at 4066Mhz with 1.41v & RAM at 3285Mhz.

Try running some benches. The Infinity Fabric runs at a multiple of the RAM speed, which means that the review benches running the RAM at 2133, 2400, etc.
are severely limiting the fabric bandwidth.

Guys educate yourself more on this issue. Some of you are acting like you can't game on amd. Some 4K results show Ryzen being faster than all Intel cpus. It performs bad in some games at 1080p. pc gamer has a review up. On over half of the games it performs just fine. Few games it loses bad for whatever reason. Get a through look at reviews with more games.

Same drama queens who think that +20% absolute CPU performance (SB -> Sky/Kabylake) makes all non-Sky/Kaby chips absolutely useless for gaming, like we're rednecks doing the tech equivalent of running a Ferrari with a golf cart engine.

 

[-PK-]

SO an 'X' processor will overclock itself to and extent based on how cool it is, and a non 'X' processor will have you do the hardwork?
I think i would just for for the 'X' and call it a day
(If I ever clean out debts, which seems like a far away affair right now.

The X processors have a XFR extended boost of 100MHz above the listed boost clock speed if you have the thermal headroom. The non-x 1700 currently acts the same, except it only has an additional 50MHz XFR if you have the thermal headroom. Neither XFR matters if you're manually overclocking these because the AMD overclock utility disables XFR.

 

[psyclist]

Considering Ryzen is more energy efficient on GF first iteration of 14nm FINFet that Intels 3-4th Iteration, id say its a win for sure. These chips and all its variants are gonna bring a ton of market share back to AMD. Looking like Naples is is gonna be a monster on the server side as well. If i were building it would probably be the 1700 on the X370 to keep future compatibility rather than Intel's HEDT. Then I get a multitasking monster and with the money saved by going AMD, I can get a high end GPU and then have the best of both worlds.

 

[Elios]

fact that my 4790K AT STOCK is still better for gaming is sad... was hoping it would at at lest as good and thats with Ryzen OC'ed at 4Ghz..

 

[sirmonkey1985]

fact that my 4790K AT STOCK is still better for gaming is sad... was hoping it would at at lest as good and thats with Ryzen OC'ed at 4Ghz..

yeah it did surprise me but lets see what these game patches do for performance once they happen, it'll be interesting to see if it does anything or if it's a total waste of time for the developers implementing the changes. could also use it as an excuse for finally buying a 4k display to get around the performance difference as well.

 

[bwang]

SO an 'X' processor will overclock itself to and extent based on how cool it is, and a non 'X' processor will have you do the hardwork?
I think i would just for for the 'X' and call it a day
(If I ever clean out debts, which seems like a far away affair right now.

Initial reviews show XFR to be fairly ineffective; the 1800X ships very close to the maximum achievable speed right now so XFR basically does nothing on them.

 

[ole-m]

Initial reviews show XFR to be fairly ineffective; the 1800X ships very close to the maximum achievable speed right now so XFR basically does nothing on them.

X1800X can however be used on the cheapest of cheapest motherboards and still benefit from it's performance.

1700 may need an X370 to achieve it's 1800X speeds of 4ghz.
however B350 does do 3.8ghz with ease on a 1700 on a lowprofile cooler (I can confirm.) on the b350m-a from asus with added vrm heatsinks just in case.
total cost: motherboard and cpu is less than 450 usd.

 

[Motley]

Ok for anybody calling a Ryzen gaming is poor.

Here's some poor smuck with a 7700k running BF1 multiplayer w/64 players.

And here's me with My 1700 @4000Mhz Playing BF1 multiplayer w/64 players. Smooth as duckbutter!

 

[WhoBeDaPlaya]

Motley, could I ask a favor?
Could you see what fps you get with the x264 stability test v2? (no need to run the whole thing, just avg after a few mins)
https://mega.nz/#!ywAFDQQQ!hEQCeRXDKpHoeRYEaspux3ZA9Smx6tp8h0leb7ZHdJo

^ Not a funky binary. From the OCN Haswell Guide
and my go-to tool for stressing Intel >= Haswell gen CPUs.

 

[os2wiz]

I ended up ordering a 1700x to play with and a B350 board since the X370 boards are limited stock and I couldn't find one in stock. Will post results Monday evening when I get my shit.

The X370 boards are short in supply because they are upgrading the bios to improve the user experience. Some of the problems with memory bandwidth are bios related though not all. I personally think they should have waited a couple of weeks to release Ryzen the bios issues would have been fixed and a few more games would have been optimized.

 

[Robergal]

There is nothing wrong with the R7 Ryzen chip. Its how windows is using the SMP. With SMP disabled the prcecessor is doing very well in 1080p. Windows is switching cores instead of locking in.

 

[WhoBeDaPlaya]

There is nothing wrong with the R7 Ryzen chip. Its how windows is using the SMP. With SMP disabled the prcecessor is doing very well in 1080p. Windows is switching cores instead of locking in.

At least on Win10. Win7 gets the same performance w/ and w/o SMT disabled on Ryzen.

 

[Trimlock]

There is nothing wrong with the R7 Ryzen chip. Its how windows is using the SMP. With SMP disabled the prcecessor is doing very well in 1080p. Windows is switching cores instead of locking in.

There's no issues with AMDs SMT on windows.

 

[FrgMstr]

There is nothing wrong with the R7 Ryzen chip. Its how windows is using the SMP. With SMP disabled the prcecessor is doing very well in 1080p. Windows is switching cores instead of locking in.

I did a bit of 3D benchmark testing yesterday and it did not make a difference in titles we have seen scale well with cores in the past.

 

[FrgMstr]

Of course not Kyle , it would help with those apps that do not scale well, that is the whole point. May I bring Ashes of Singularity to your attention. Scales very well with more cores and threads. So why did it do well a month ago and not this past week. Simple. Stardock did a major update of the program in the past week and it broke the old AMD optimizations. That is why Brad Wardell is working with AMD feverishly to make the new fixes. Wardell started out on the os/2 platform tears ago. He is a major advocate for multithreading and multicore support. His strategy games run like large databases optimized for the multicre environment. More of his kind will turn the gaming workld on its head and bring gaming software into the 21st century. You will see in another week the amazing superiority of Ryzen in that game. I have bought almost everrpy Stardock game since the os/2 days and they all smack of quality programming as well as great design. This is the tip of the iceberg Ryzen will show its worth even in gaming. Keep your head and chin up. Take care.

Time will tell. All we can do it tell you what is happening right now with performance. We don't make a lot of statements about what "could'a should'a would'a" happened when it comes to perf. When/if that changes, certainly we will report on it.

 

[FrozenSun]

Looking forward to seeing how well the 1700 OC's when you get it on the bench. Also waiting to see how some of the more mid range boards overclock. I have a feeling that Ryzen overclocking is more dependent on high end mobos than Intel's.

 

[FrameBuffer]

Motley, could I ask a favor?
Could you see what fps you get with the x264 stability test v2? (no need to run the whole thing, just avg after a few mins)
https://mega.nz/#!ywAFDQQQ!hEQCeRXDKpHoeRYEaspux3ZA9Smx6tp8h0leb7ZHdJo

^ Not a funky binary. From the OCN Haswell Guide
and my go-to tool for stressing Intel >= Haswell gen CPUs.

================================================================
x264-64 Stability test
================================================================
x264 0.148.2597 e86f3a1
(libswscale 3.0.0)
(libavformat 56.21.0)
built on Aug 19 2015, gcc: 4.9.2
x264 configuration: --bit-depth=8 --chroma-format=all
libx264 configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat license: LGPL version 2.1 or later
==== Configuration =============================================
Log name = x264-log_test.rtf
Loops = 2
Threads = 16
Priority = high
==== Results ===================================================
Start: 22:14:18.91 Sat 03/04/2017
Loop 1: 22:14:18.92
encoded 2121 frames, 7.06 fps, 36016.71 kb/s
Loop 2: 0:19:19.86
encoded 2121 frames, 7.09 fps, 36016.71 kb/s
Finish: 0:24:19.24 Sun 03/05/2017

 

[WhoBeDaPlaya]

Thanks! What R7 and speed were you running?
Here are some of my numbers (using x264 v2.00) that I've been collecting over the years :

PII X6 1045t                 1.72fps
i5 2500K @ 3.6GHz            2.12fps
i5 2500K @ 4.5GHz            2.58fps
FX 8320                      2.75fps
i7 4770K @ 4.5GHz            3.60fps
i7 4810MQ                    2.66fps        [Clevo w230ss - Stock settings ~3.2GHz under full load]
i7 4700MQ                    2.45fps        [ThinkPad T540p - Stock settings ~2.9GHz under full load due to TDP limit]
i7 3770 @ 4.1GHz             3.10fps
i5 3570K @ 4.4GHz            2.78fps
2x E5620                     3.08fps
2x X5675                     6.09fps
2x X5660                     5.56fps
i7 3610QM                    2.35fps        [Alienware m17x R4 - Stock settings ~3.1GHz under full load]

So yup, a single R7 would probably trade blows with my current Z800 (dual X5690) in x264 encoding

 

[bwang]

not really Ryzen related but I ran the benchmark on my quad E5-4650 for shits and giggles and got:

================================================================

                     x264-64 Stability test

================================================================


x264 0.148.2597 e86f3a1
(libswscale 3.0.0)
(libavformat 56.21.0)
built on Aug 19 2015, gcc: 4.9.2
x264 configuration: --bit-depth=8 --chroma-format=all
libx264 configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat license: LGPL version 2.1 or later
 

==== Configuration =============================================


Log name = x264-log_x264.txt.rtf

Loops    = 10

Threads  = 64

Priority = normal


==== Results ===================================================

Start:   3:10:47.37 Sun 03/05/2017

Loop 1:  3:10:47.38
encoded 2121 frames, 9.00 fps, 36113.53 kb/s

doesn't seem to want to scale to 64 threads

 

[WhoBeDaPlaya]

^ Is it I/O constrained? eg. to feed my 12c/24t in FLAC encoding, the source files have to be on a SSD.

 

[FrameBuffer]

not really Ryzen related but I ran the benchmark on my quad E5-4650 for shits and giggles and got:

================================================================

                     x264-64 Stability test

================================================================


x264 0.148.2597 e86f3a1
(libswscale 3.0.0)
(libavformat 56.21.0)
built on Aug 19 2015, gcc: 4.9.2
x264 configuration: --bit-depth=8 --chroma-format=all
libx264 configuration: --bit-depth=8 --chroma-format=all
x264 license: GPL version 2 or later
libswscale/libavformat license: LGPL version 2.1 or later
 

==== Configuration =============================================


Log name = x264-log_x264.txt.rtf

Loops    = 10

Threads  = 64

Priority = normal


==== Results ===================================================

Start:   3:10:47.37 Sun 03/05/2017

Loop 1:  3:10:47.38
encoded 2121 frames, 9.00 fps, 36113.53 kb/s

doesn't seem to want to scale to 64 threads

R7 1700@ 4.0Ghz

Thanks! What R7 and speed were you running?
Here are some of my numbers (using x264 v2.00) that I've been collecting over the years :

PII X6 1045t                 1.72fps
i5 2500K @ 3.6GHz            2.12fps
i5 2500K @ 4.5GHz            2.58fps
FX 8320                      2.75fps
i7 4770K @ 4.5GHz            3.60fps
i7 4810MQ                    2.66fps        [Clevo w230ss - Stock settings ~3.2GHz under full load]
i7 4700MQ                    2.45fps        [ThinkPad T540p - Stock settings ~2.9GHz under full load due to TDP limit]
i7 3770 @ 4.1GHz             3.10fps
i5 3570K @ 4.4GHz            2.78fps
2x E5620                     3.08fps
2x X5675                     6.09fps
2x X5660                     5.56fps
i7 3610QM                    2.35fps        [Alienware m17x R4 - Stock settings ~3.1GHz under full load]

So yup, a single R7 would probably trade blows with my current Z800 (dual X5690) in x264 encoding

1700 @ 4Ghz