| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Community
Forum: Hardware Chat Room
Thread: Ryzen performance |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 66
|
|
| Author |
|
|
geoff.andreychuk
Cruncher Joined: Nov 2, 2015 Post Count: 42 Status: Offline Project Badges: 
|
Hello everyone, just wanting to give a bit of a performance breakdown on my Ryzen 1700 vs some of my other machines. Currently I have three machines all doing SCC work units to give some balance to these results. These are also an avg over a couple of days
----------------------------------------Ryzen 1700 @3.2Ghz, Points 68,000 Units Returned 108 (6.75 Units returned per core) Intel 6800k @3.4Ghz, Points 44,000, Units Returned 77 (6.41 Units returned per core) Intel 4790k @4Ghz, Points 34,500, Units Returned 60 (7.5 Units returned per core) Anyone else who has Ryzen please post some results so we can see just how good Ryzen is!!!  |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7745 Status: Offline Project Badges: 
|
What OSs are you using and what is the mix of projects for each machine ?
----------------------------------------Thanks
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
duanebong
Advanced Cruncher Singapore Joined: Apr 25, 2009 Post Count: 134 Status: Offline Project Badges:
|
Thanks for sharing, those are solid performance figures indeed. At 3.2GHz Ryzen is probably consuming less than 60 watts... any reason why you're not clocking the Ryzen higher?
----------------------------------------The boost clock of the 1700 is 3.7GHz and the stock Wraith cooler is rated for 95W - you have 300-400MHz untapped headroom which is still within your setup's official rated specs. Would you mind also sharing details of the motherboard and what DDR4 memory speed you manage to set on your Ryzen? With 16 threads I imagine performance would be affected by memory bandwidth.  |
||
|
|
geoff.andreychuk
Cruncher Joined: Nov 2, 2015 Post Count: 42 Status: Offline Project Badges:
|
@Sgt.Joe I am using windows 10 on all systems, and I listed above that all machines are currently doing SCC only
----------------------------------------@duanebong I have OC'ed my systems to see what the max performance is, however I also have found that the increased power demand is far greater than the minimal performance increase. The Power saved by not OC allowed me to add in my 1700 and still save a bit of money on my hydro bill. My specs are as follows Case: CoolMaster Kendomen (has 2x front, 2x top, 1x rear fans.. all stock) Motherboard: x370-pro RAM: 16Gb Corsair Vengence 3200 CPU: Ryzen 1700 Cooler: Wraith Spire Cooler HD: 1Tb WB Black Video: EVGA 8400 (only to allow post and remote access) Power: Thermaltake 750W bronze Cost for everything except power was $1150 CAD (including tax) |
||
|
|
VietOZ
Senior Cruncher United States Joined: Apr 8, 2007 Post Count: 205 Status: Offline Project Badges: 
|
Here's the quick compare between the 1700 and the dual E5-2695 V4 ES (56 threads)
----------------------------------------Due to the fluctuation of points with the short WUs of FAAH (which the Ryzen hasn't got 1 yet) I'll pick the 4 days where both machines are running without a short WUs. 18th - 21st. The Ryzen 1700 @ 3.7ghz Ubuntu 17.04, power draws 165W at the wall Average for the 4 days: WCG points = 131,832 PPW : WCG points = 799 Dual E5-2695 ES V4 (56T) @ 2.8ghz Linux Mint 17.2 , power draws 305W at the wall Average for the 4 days: WCG points = 290,946 PPW : WCG points = 954 So looks like Dual xeon is still dual xeon. But the Ryzen still every competitive. (compare to an I7-5960X which is also an 8C/16T $1100 cpu, it crushed the 5960X in power consumption) Ryzen 1700 Dual E5-2695 ES V4 |
||
|
|
enels
Senior Cruncher Joined: Apr 25, 2008 Post Count: 286 Status: Offline Project Badges:
|
Does anyone want to try undervolting/underclocking?
I'm hoping the Ryzen 1700 can be made to use less than 100W at the wall without loosing too much performance. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Do you observe power draw difference between running under W10 v Linux at same clock speed? Just curious, because my dual boot ancient runs quite a few degrees hotter under Linux, w/kernel 4.10 v W10 (bld 15063 now and previous Windows)
|
||
|
|
VietOZ
Senior Cruncher United States Joined: Apr 8, 2007 Post Count: 205 Status: Offline Project Badges:
|
I noticed the power draw went up about 3W when I first installed ubuntu 17.04. But after a couple of updates it is dead lock at 165W, which is the same with W10 pro.
----------------------------------------I'm also using kernel 4.10. |
||
|
|
duanebong
Advanced Cruncher Singapore Joined: Apr 25, 2009 Post Count: 134 Status: Offline Project Badges:
|
@VietOZ... thanks for the comparison.
----------------------------------------Your Ryzen @ 3.7GHz getting an average of ~131K points per day is incredible. That's slightly more than what I can achieve with my entire farm (4770K desktop + 2600K desktop + i5 4200U Brix HTPC + i5 3317U ultrabook + Atom Z3735F tablet + Helio X25 Android phone)! |
||
|
|
duanebong
Advanced Cruncher Singapore Joined: Apr 25, 2009 Post Count: 134 Status: Offline Project Badges:
|
@geoff.andreychuk... thanks for elaborating on your hardware configuration. Does your Corsair memory run at 3200MHz? I read many people have to run their memory slower than their rated specs because of the current immaturity of the BIOS firmware.
---------------------------------------- |
||
|
|
|