Overclockers Australia Forums

OCAU News - Wiki - QuickLinks - Pix - Sponsors  

Go Back   Overclockers Australia Forums > Manufacturer-specific Forums > AMD x86 CPUs and chipsets

Notices


Sign up for a free OCAU account and this ad will go away!
Search our forums with Google:
Reply
 
Thread Tools
Old 15th March 2017, 5:54 PM   #91
lowdog
Member
 
lowdog's Avatar
 
Join Date: Feb 2003
Location: Dreamland.
Posts: 4,144
Default

Ummmm what DF mean


Also what res was the Resident Evil game run at?

Want to see more game results please.
__________________

Last edited by lowdog; 15th March 2017 at 5:57 PM.
lowdog is offline   Reply With Quote

Join OCAU to remove this ad!
Old 15th March 2017, 6:00 PM   #92
mAJORD Thread Starter
Member
 
mAJORD's Avatar
 
Join Date: Jun 2002
Location: Griffin , Brisbane
Posts: 8,678
Default

Data fabric. Maybe it should read / instead of and
__________________
On the Bench: Ryzen 7 1700 - ASUS Prime X370 Pro - 16GB Corsair Vengance DDR4-3200
Main:
Ryzen 7 1800x - Crosshair VI Hero - 16GB G.SKill DDR4-3333 - ASUS 1080Ti FE - Custom water - Be-Quiet Silent Base
Ryzen OC / Owners Thread
mAJORD is offline   Reply With Quote
Old 15th March 2017, 8:24 PM   #93
aussie-revhead
Member
 
aussie-revhead's Avatar
 
Join Date: Feb 2005
Location: St Clair 2759
Posts: 18,521
Default

mAJORD - re the score in your chart showing 20k fs physics, what cpu clock was that? Do we have fps data for BF1 comparing Ryzen with Intel products?

__________________
Bench 1- Asus Rampage 5 Extreme - Intel 5960x - 32gb of 3200c16 DDR4- 1x Galax GTX1080 Ti HOF - Samsung 960 Evo + Intel 600P m.2 512GB - Enermax Revo1250 - water cooled tech bench - QNIX 2560x1440 screen @ 100Hz - Aten 120Hz KVM
Netbox - Dell 990 SFF - 2500 - 16GB ram - gtx750Ti - 240GB SSD - 2TB green
Current vehicles- Black 2007 ZX14 - VE V8 Calais (wifes car) - VN SS (for sale)
aussie-revhead is offline   Reply With Quote
Old 15th March 2017, 9:03 PM   #94
mAJORD Thread Starter
Member
 
mAJORD's Avatar
 
Join Date: Jun 2002
Location: Griffin , Brisbane
Posts: 8,678
Default

Sorry, forgot to add those details , 3.9Ghz for all tests

There are a few tests around for bf1, one of the few games where Ryzen does well, with SMT on to boot
__________________
On the Bench: Ryzen 7 1700 - ASUS Prime X370 Pro - 16GB Corsair Vengance DDR4-3200
Main:
Ryzen 7 1800x - Crosshair VI Hero - 16GB G.SKill DDR4-3333 - ASUS 1080Ti FE - Custom water - Be-Quiet Silent Base
Ryzen OC / Owners Thread
mAJORD is offline   Reply With Quote
Old 15th March 2017, 11:55 PM   #95
Kelvin
Member
 
Join Date: Jul 2001
Posts: 2,316
Default

hmmmm..

Just putting a together a 1700 and Prime B350M-A in a silverstone SG11..

A few issues along the way.. most minor..

The Standoff layout on the Asus board is slightly non standard.. (no holes for the top of the board)

The Standoffs in the Silverstone SG11 are mostly fixed and cannot be removed....

Had to use a bit of masking tap other bits to insulate the board electrically from the stand offs.

The front header layout.. is poorly marked on the board (not typical of asus)... had to consult the manual on several occasions..

Can't figure out the CPU fan LED connector thing (will sort that out later)

Installing windows 7 on Ryzen.. is interesting...
http://www.anandtech.com/show/11182/...-windows-7-x64

had to dig out a PS2 keyboard and board.. (the first one I had.. had a faulty spacebar and faulty row of keys (5tgb).. installed via DVD ROM..

None of the USB2 ports worked until the drivers were installed..
(copied the drivers onto an SSD and plugged the SSD into the sata port)
Kelvin is offline   Reply With Quote
Old 16th March 2017, 2:39 AM   #96
cbjaust
Member
 
cbjaust's Avatar
 
Join Date: Sep 2003
Location: Melbourne
Posts: 942
Default

Quote:
Originally Posted by mAJORD View Post
finally got around to collecting Some DF and RAM scaling data.

I plan to do a lot more game related testing than this. But i graphed the couple that showed the trend most obviously - that is, Memory/DF clocks trump timings

image

image

and particuarly interesting is the skydiver combined test. you can see what appears to be n very little memory bw bottleneck on the individual tests, but a clear scaling when you combine physics with graphics.

this trend is not seen on firestrike, possibly because it's more bound to both raw rendering and computational power on both cpu and gpu wheras the lighter skydiver test , memory subsystem becomes the bottleneck.

I'd love to to test this on 4+0 config, and also on a system with fixed L3 / ringbus clock (something thats not possible on ryzen - so bulldozer or intel system) while varying ram speed. and see if the combined test still see'
s an uplift.


image
I think you should chart the different latencies individually and use the correct scale for the frequency like this for example (using your data):



It shows that least in this benchmark that the lower latency gives a small but consistent increase in score.
__________________
MAIN AMD FX-8350 @ 4.78GHz | Asus Crosshair V Formula-Z | 16GB G.Skill TridentX | Radeon R9 280X 3GB
HTPC AMD A8-7650K @ 3.8GHz | ASRock FM2A88X-ITX+ | 8GB G.Skill Ares | Radeon R7 Graphics on-board
SERVER AMD Phenom II X4 940 @ 3.4GHz | DFI DK 790FXB-M2RSH | 8GB G.Skill PI | Radeon HD5670 1GB

Last edited by cbjaust; 16th March 2017 at 2:53 AM.
cbjaust is offline   Reply With Quote
Old 16th March 2017, 12:53 PM   #97
2_stroke
Member
 
Join Date: Aug 2008
Location: cranbourne 3977
Posts: 1,155
Default

somebody get a x1800 please
__________________
http://bfbc2.elxx.net/sig/clean1/pc/2stroke.png
Successful trades
Sold to: chopstick9558, jackdaw, davomaster, haksam, jomo, kingpeck
Bought from: h$v, epo299, tangcla, Gillyske, winterstide, jboles, nooobs,Sainter4Life, auron1
stop listing years ago just way to many trades lol
2_stroke is offline   Reply With Quote
Old 16th March 2017, 12:57 PM   #98
Bertross
Member
 
Bertross's Avatar
 
Join Date: Feb 2009
Location: Melb, CBD
Posts: 12,731
Default

Quote:
Originally Posted by 2_stroke View Post
somebody get a x1800 please
OJR should be setup tonight with the 1800x.
__________________

EVGA GTX 1080TI FTW3 | RYZEN 1700+H100I | ASUS ROG BIG HERO 6+SAMSUNG 950 PRO M.2 | 32GB TRIDENT Z RGB 3000 | PHANTEKS ENTHOO EVOLV TEMPERED | DELL U3415W 34"
Bertross is offline   Reply With Quote
Old 16th March 2017, 1:02 PM   #99
mAJORD Thread Starter
Member
 
mAJORD's Avatar
 
Join Date: Jun 2002
Location: Griffin , Brisbane
Posts: 8,678
Default

yes most certainly tighter timing does give an increase. Mmain reason I graphed in that fashion was was to illustrate a greater importance for memory clock under such workloads especially at lower clocks.. Ultimately you'd need to map minimum stable timings @ the different memory clocks for your own system, but for the most part, it seems chase memory clock first - at least up to 2666 - 2933 which should be achievable fairly easily, then by all means tighten timings as much as possible.

The other thing that needs testing is single vs Dual rank. The later will yield higher memory clocks, but does this offset any potential performance deficit vs Dual Rank, i'm not sure. It does have some effect on Bristol Ridge
__________________
On the Bench: Ryzen 7 1700 - ASUS Prime X370 Pro - 16GB Corsair Vengance DDR4-3200
Main:
Ryzen 7 1800x - Crosshair VI Hero - 16GB G.SKill DDR4-3333 - ASUS 1080Ti FE - Custom water - Be-Quiet Silent Base
Ryzen OC / Owners Thread
mAJORD is offline   Reply With Quote
Old 16th March 2017, 1:21 PM   #100
TaroT
Member
 
TaroT's Avatar
 
Join Date: Jan 2002
Location: Hazelbrook nsw 2779
Posts: 8,452
Default

i have a 1800x...what do you want to know
have not added it in yet still tinkering.
__________________
AMD Threadripper 1950X@stock:corsair H115i cooler: Asrock Taichi X399Gskilll Flarex 3200 ddr4 @3200 cas14XFX Vega 64 8 gig
TaroT is offline   Reply With Quote
Old 16th March 2017, 2:44 PM   #101
lowdog
Member
 
lowdog's Avatar
 
Join Date: Feb 2003
Location: Dreamland.
Posts: 4,144
Default

Quote:
Originally Posted by TaroT View Post
i have a 1800x...what do you want to know
have not added it in yet still tinkering.

F6 to optimised defaults in bios and what volts does your cpu vcore read please while in bios
__________________
lowdog is offline   Reply With Quote
Old 16th March 2017, 3:16 PM   #102
TaroT
Member
 
TaroT's Avatar
 
Join Date: Jan 2002
Location: Hazelbrook nsw 2779
Posts: 8,452
Default

oh but but means i have to reboot again...ok brb

ok 1.460 vcore at default settings (loads up to 1.504 at times in bios)
hence why i manually newtered it
__________________
AMD Threadripper 1950X@stock:corsair H115i cooler: Asrock Taichi X399Gskilll Flarex 3200 ddr4 @3200 cas14XFX Vega 64 8 gig

Last edited by TaroT; 16th March 2017 at 3:29 PM.
TaroT is offline   Reply With Quote
Old 16th March 2017, 3:42 PM   #103
lowdog
Member
 
lowdog's Avatar
 
Join Date: Feb 2003
Location: Dreamland.
Posts: 4,144
Default

Quote:
Originally Posted by TaroT View Post
oh but but means i have to reboot again...ok brb

ok 1.460 vcore at default settings (loads up to 1.504 at times in bios)
hence why i manually newtered it

1.46 wow. Obviously different kettle of fish to my 1700. Mine boosts to 3.7GHz with default vcore 1.19v and I can set the speed in bios to 3700 or 3.7Ghz and leave vcore on auto and in windows when loaded it never goes over 1.19v.

Can do 3.9GHz with 1.325v and LLC - 4 and under load it's 1.320v max.

You should be able to get similar settings for 3.9GHz......the extra volts for 4GHz and the increased power draw just don't seem worth it for the extra 100MHz.

wonder what you max volts will bump to in windows under load if you set your cpu speed in bios to fixed 3700/3.7GHz....be interested to see if it's similar to mine of under 1.2V while running prime full load.
__________________
lowdog is offline   Reply With Quote
Old 16th March 2017, 3:57 PM   #104
TaroT
Member
 
TaroT's Avatar
 
Join Date: Jan 2002
Location: Hazelbrook nsw 2779
Posts: 8,452
Default

on stock at 3.8 it was pushing 1.5 auto seems to think the more volts the merrier but then the 1700 is a 65w part mine is a 95w so i expected some hulk like power

i could drop it down to 1.35 but i think it will fall over so i, m happy where i am at the moment till the bioses even out
__________________
AMD Threadripper 1950X@stock:corsair H115i cooler: Asrock Taichi X399Gskilll Flarex 3200 ddr4 @3200 cas14XFX Vega 64 8 gig
TaroT is offline   Reply With Quote
Old 16th March 2017, 4:03 PM   #105
lowdog
Member
 
lowdog's Avatar
 
Join Date: Feb 2003
Location: Dreamland.
Posts: 4,144
Default

Quote:
Originally Posted by TaroT View Post
on stock at 3.8 it was pushing 1.5 auto seems to think the more volts the merrier but then the 1700 is a 65w part mine is a 95w so i expected some hulk like power

i could drop it down to 1.35 but i think it will fall over so i, m happy where i am at the moment till the bioses even out

Lol they both perform the same at a given overclock frequency eg: 4GHz but may require different volts to get there
__________________
lowdog is offline   Reply With Quote
Reply

Bookmarks

Sign up for a free OCAU account and this ad will go away!

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time now is 3:17 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
OCAU is not responsible for the content of individual messages posted by others.
Other content copyright Overclockers Australia.
OCAU is hosted by Micron21!