• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Matisse (Ryzen 3000) overclocking/undervolting

Joined
Sep 9, 2009
Messages
1,050 (0.28/day)
Location
Austria, Europe
System Name Bang4Buck
Processor AMD Ryzen 9 3900x
Motherboard MSI PRESTIGE x570 CREATION
Cooling Fractal Design Celsius S36
Memory 32Gb 4400Mhz Patriot Viper Steel(Samsung B-die) @ 3800Mhz 16-16-16-32-48-1T @ 1.44v
Video Card(s) EVGA GeForce GTX 1080 Ti SC2 Gaming, watercooled with NZXT Kraken G12 + Thermaltake Water 3.0 140RGB
Storage Adata SX8200Pro M.2 Nvme/Samsung 840 EVO/6Tb Raid -HGST Enterprise/IronWolf 8Tb/BW-16D1H-U PRO
Display(s) Samsung UE49KS8002 4K HDR TV (US - 9 series)
Case Fractal Design Define R6 Black Usb-C
Audio Device(s) HDMI out to Denon X4400H reciever, 2x Dali Zensor 7, Dali Zensor Vokal, 2x Dali Zensor 1, Yamaha Sub
Power Supply Seasonic Prime Ultra 750W
Keyboard Logitech K520
Software Windows 10 Pro x64
Benchmark Scores http://www.3dmark.com/spy/10026304
Anyone want to guess what Silicon Lottery will post for golden chips? Suppose to show up at 10PM CST.
4.5Ghz allcore oc on 3800x @ 1.325v is my guess...
 
Joined
Nov 1, 2018
Messages
570 (1.26/day)
Suggest installing the cooler 1st and run some stuff you deem necessary to compare performance VS stock cooler and then swap the memory and repeat the tests: a bit more hassle but it gives us better idea of how much changing from stock cooler to a very good air cooler with everything else the same can do.
Okay, will do.
Cooler will arrive first it seems ( tomorrow ). Memory later in the week.
 
  • Like
Reactions: HTC
Joined
Mar 23, 2016
Messages
3,339 (2.38/day)
Processor Ryzen 5 3600
Motherboard MSI B450 Tomahawk ATX
Cooling Scythe Kotetsu with AM4 bracket
Memory PNY Anarchy-X XLR8 Red DDR4 3200 MHz C15-17-17-17-35
Video Card(s) MSI GeForce RTX 2060 GAMING Z 6G
Storage Samsung 970 EVO NVMe M.2 500 GB, SanDisk Ultra II 480 GB
Display(s) Samsung SyncMaster C27H711 OC refresh rate 110Hz
Case Phantek Eclipse P400S (PH-EC416PS)
Audio Device(s) EVGA NU Audio
Power Supply EVGA 850 BQ
Mouse SteelSeries Rival 310
Keyboard Logitech G G413 Silver
Software Windows 10 Professional 64-bit v1903
Joined
Oct 21, 2005
Messages
5,718 (1.10/day)
Location
USA
System Name Small ATX Desktop
Processor Intel i5 8600K @ 4.6 GHz Core & 4.3 GHz Uncore @ 1.168V
Motherboard Asrock Z370 Taichi
Cooling Phanteks PH-TC14PE, MonoPlus, Fans: 2xThermalright TY143 / 2xCorsair SP-120L / 2xYate Loon D14BH-12
Memory G-Skill TridentZ 2X8GB DDR4 3200 CL14 F4-3200C14D-16GTZ @ 3200 14-14-14-34 2T @ 1.35V
Video Card(s) Zotac 1060 6GB Mini ZT-P10600A-10L with Arctic MonoPlus and Yate Loon D14BH-12 Fan
Storage OS Samsung 970 Pro 512GB NVMe, Games Phison E12 NVMe 1TB, HDD WD Black 4TB WD4005FZBX, 2xWD10EZEX
Display(s) 2x Asus PB258Q 2560x1440 25" IPS
Case Lian Li PC A05NB (Inverted Mobo)
Audio Device(s) Audiotechnica ATH M50X, Antlion Mod Mic 4, SYBA SD-CM-UAUD, Acoustic Research 2Ch Speakers
Power Supply Seasonic SS-660XP2 660 Watt Platinum
Mouse Zowie EC2A Mouse, Razer Naga Chroma '14, Corsair MM600, Inateck 900x300 XL pad, Tiger Gaming Skates.
Keyboard Filco Majestouch II Ninja TKL, Goldtouch GTC 0033 Ten Key, PS3 Controller
Software Win 10 LTSC 1809
Actually the coolers between a 939 and an FX are different even though they may look/appear the same. Changes happened along the way and I'll say not to the good, the coolers got "cheaper" as time passed with the last relying on the heatpipes mostly.

The old 939's had a base with a bit of substance to them compared to the later ones, out of all from 939 to FX with this design they are still the best.

I would dare say they are at least just as good as the Wraith coolers too.
Yeah they had a big copper plate on them. Had one of them back in the day on top of my Opteron.
 
Joined
Dec 14, 2013
Messages
1,316 (0.59/day)
Location
Alabama
Processor Ryzen 2700X
Motherboard X470 Tachi Ultimate
Cooling Scythe Big Shuriken 3
Memory C.R.S.
Video Card(s) Radeon VII
Software Win 7
Benchmark Scores Never high enough
The copper plate is what makes the difference, more mass to absorb heat and move it along. Those that came later with the direct heatpipe contact are..... OK but not as good, even though they too have a plate but it does little in comparison to the 939 type.

I do not like direct heatpipe contact coolers, they are just a cheaper form of cooler.
 
Joined
Nov 11, 2004
Messages
4,747 (0.85/day)
Location
Formosa
System Name Overlord Mk MXVI
Processor AMD Ryzen 7 3800X
Motherboard Gigabyte X570 Aorus Master
Cooling Corsair H115i Pro
Memory 32GB Viper Steel 3600 DDR4 @ 3800MHz 16-19-16-19-36
Video Card(s) Gigabyte RTX 2080 Gaming OC 8G
Storage 1TB WD Black NVMe (2018), 2TB Viper VPN100, 1TB WD Blue 3D NAND
Display(s) Asus PG27AQ
Case Corsair Carbide 275Q
Audio Device(s) Corsair Virtuoso SE
Power Supply Corsair RM750
Mouse Logitech G500s
Keyboard Wooting Two
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/ztiub6
Anyone want to guess what Silicon Lottery will post for golden chips? Suppose to show up at 10PM CST.
That must be a joke no? That looks very mediocre at best and not worth paying extra for.
I guess some people snagged some cheap CPU's though...
 
Last edited:
Joined
Nov 1, 2018
Messages
570 (1.26/day)
That website/company is a scam and anyone with half a functioning brain knows it.

Unfortunately there are plenty of people with less than half of a brain functioning... which is why they are still in business.
 
Joined
Dec 14, 2013
Messages
1,316 (0.59/day)
Location
Alabama
Processor Ryzen 2700X
Motherboard X470 Tachi Ultimate
Cooling Scythe Big Shuriken 3
Memory C.R.S.
Video Card(s) Radeon VII
Software Win 7
Benchmark Scores Never high enough
That website/company is a scam and anyone with half a functioning brain knows it.

Unfortunately there are plenty of people with less than half of a brain functioning... which is why they are still in business.
Truth!
No way I'd pay the extra they are charging for these.
 
Joined
Nov 1, 2018
Messages
570 (1.26/day)
You can thank amd for the results.
Actually, I think you can thank TSMC for not being able to hit AMD's expectations.

I really believe AMD designed this core to boost up to 5 Ghz on 7nm, but failed to do so for reasons we will probably never know.
It's possible that "A0" silicon targeted those clocks, but proved to be unstable and causing errors, no matter what, and they altered some circuits to make the cores run well and error free and had to lower the clocks to make the changes to the current "B0" silicon work.

This is obviously a wild guess, but it might explain why there's no headroom whatsoever on these chips.

Also, it is possible there will be a refresh down the line that DOES boost higher ( by 200-300 Mhz, potentially hitting that magic 5Ghz on single thread ). All core... no way, but single thread definitely possible if they find a fix to what's preventing the chips to go there. The process itself should definitely be capable !
 
Joined
Dec 31, 2009
Messages
16,358 (4.45/day)
I think this has more to do with the uarch than it does with the process. Zen, and Zen+ behave similarly. No headroom. It's AMD getting the most out of these CPUs out of the box. The process has little to do with it.

So, SL is just doing what they are doing... hardly a scam, but clearly not worth it to this subset of ENTHUSIASTS... whereas those who may not be as well versed thinks it is worth the time and effort. But this is hardly a scam and a very myopic look at the whole thing IMO.
 
Joined
Oct 21, 2005
Messages
5,718 (1.10/day)
Location
USA
System Name Small ATX Desktop
Processor Intel i5 8600K @ 4.6 GHz Core & 4.3 GHz Uncore @ 1.168V
Motherboard Asrock Z370 Taichi
Cooling Phanteks PH-TC14PE, MonoPlus, Fans: 2xThermalright TY143 / 2xCorsair SP-120L / 2xYate Loon D14BH-12
Memory G-Skill TridentZ 2X8GB DDR4 3200 CL14 F4-3200C14D-16GTZ @ 3200 14-14-14-34 2T @ 1.35V
Video Card(s) Zotac 1060 6GB Mini ZT-P10600A-10L with Arctic MonoPlus and Yate Loon D14BH-12 Fan
Storage OS Samsung 970 Pro 512GB NVMe, Games Phison E12 NVMe 1TB, HDD WD Black 4TB WD4005FZBX, 2xWD10EZEX
Display(s) 2x Asus PB258Q 2560x1440 25" IPS
Case Lian Li PC A05NB (Inverted Mobo)
Audio Device(s) Audiotechnica ATH M50X, Antlion Mod Mic 4, SYBA SD-CM-UAUD, Acoustic Research 2Ch Speakers
Power Supply Seasonic SS-660XP2 660 Watt Platinum
Mouse Zowie EC2A Mouse, Razer Naga Chroma '14, Corsair MM600, Inateck 900x300 XL pad, Tiger Gaming Skates.
Keyboard Filco Majestouch II Ninja TKL, Goldtouch GTC 0033 Ten Key, PS3 Controller
Software Win 10 LTSC 1809
That website/company is a scam and anyone with half a functioning brain knows it.

Unfortunately there are plenty of people with less than half of a brain functioning... which is why they are still in business.
Yeah, the CoffeeLake binning didn't seem legit at all based on what my sample does and what people were reporting on Overclock.net. Also the pricing for 5GHz 5.1, 5.2 etc CPUs is ridiculous given most CoffeeLake can achieve that easily with enough voltage and a solid board.

Furthermore, the binning they do is very simplified and not very exact. The voltages they use for 5.1+ are way too high and that tells me their top bins are nothing special. A high bin 8700K for example would be 5.1 @ 1.25, not 5.1 @ 1.4v. 5.1 @ 1.4V is average and quite mundane for an 8700K especially with a -2 offset. Your typical retail CPU can do that with a midrange Z370.

A lot of their bins rely on AVX offset which, when subtracted from the "binned" frequency shows they are nothing special. I am also very skeptical of the binning statistics.
 
Last edited:
Joined
Dec 31, 2009
Messages
16,358 (4.45/day)
What were people reporting?

What do you mean by the binning they do is simplified and not 'exact'? They needed the voltage they for stability in how they test.

Many users run an AVX offset in the first place as with typical CPUs at 5GHz+ with AVX, most coolers can't keep up anyway.

I strongly disagree with the assertion that SL is a scam. It may not be the best deal, but for those who do not want to gamble, it can be worth the premium. It may not be worth it to you, but you're also more of an enthusiast than most and take the chance and time to manually overclock your own.
 
Joined
Oct 21, 2005
Messages
5,718 (1.10/day)
Location
USA
System Name Small ATX Desktop
Processor Intel i5 8600K @ 4.6 GHz Core & 4.3 GHz Uncore @ 1.168V
Motherboard Asrock Z370 Taichi
Cooling Phanteks PH-TC14PE, MonoPlus, Fans: 2xThermalright TY143 / 2xCorsair SP-120L / 2xYate Loon D14BH-12
Memory G-Skill TridentZ 2X8GB DDR4 3200 CL14 F4-3200C14D-16GTZ @ 3200 14-14-14-34 2T @ 1.35V
Video Card(s) Zotac 1060 6GB Mini ZT-P10600A-10L with Arctic MonoPlus and Yate Loon D14BH-12 Fan
Storage OS Samsung 970 Pro 512GB NVMe, Games Phison E12 NVMe 1TB, HDD WD Black 4TB WD4005FZBX, 2xWD10EZEX
Display(s) 2x Asus PB258Q 2560x1440 25" IPS
Case Lian Li PC A05NB (Inverted Mobo)
Audio Device(s) Audiotechnica ATH M50X, Antlion Mod Mic 4, SYBA SD-CM-UAUD, Acoustic Research 2Ch Speakers
Power Supply Seasonic SS-660XP2 660 Watt Platinum
Mouse Zowie EC2A Mouse, Razer Naga Chroma '14, Corsair MM600, Inateck 900x300 XL pad, Tiger Gaming Skates.
Keyboard Filco Majestouch II Ninja TKL, Goldtouch GTC 0033 Ten Key, PS3 Controller
Software Win 10 LTSC 1809
Nice, perfect. Now you will probably be in the 60s.
 
Joined
Nov 1, 2018
Messages
570 (1.26/day)
Nice, perfect. Now you will probably be in the 60s.
Extremely unlikely, there's over 30 degrees in this room...



They weren't kidding in the reviews that installation is a breeze.

It literally took me only 5 minutes to mount it, and another 10 to admire the craftsmanship...

But now my computer is all over the place in colors ... oh well ;)

First observation:
I can't hear it !
The noise outside in the street is stronger than the fan at auto-speed.
Even at full speed (~2000 rpm it's barely audible inside the case)
WOW !

---- First quick round of benchmarks (Single thread) ---
Note: It's VERY hot here, I don't have AC and it's afternoon. Room thermometer shows around 30 degrees Celsius ambient ( 86 F )

The percentages are related to this post:
https://www.techpowerup.com/forums/threads/matisse-ryzen-3000-overclocking-undervolting.257505/post-4084338 (The "Default" 100% reference)

wPrime 32M - 29.0 (101.7%)
SuperPI 4M - 45.8 (101.5%)
Cine R10 - 10156 (100.7%)
CpuZ - 517 (100.2%)
3DMark2003 CPU T1 - 58.8 (102.8%)
CPU-M 1.6 - 20310 (101.2%)

Extra: 7zip single threaded (32MB dic): 7014 MIPS, 5227 KB/s Comp, 90546 KB/s Decomp
Also 7zip but 16 threads: 73860 MIPS, 51410 KB/s Comp, 1000594 KB/s Decomp

One observation about clocks, during all the single threaded tests it was consistently boosting between 4.3 and 4.4, with tiny spikes to 4.45 ( I have PBO+100 enabled )
Max temp during 7zip ran for 5 mins ... ~ 81 degrees.
Cinebench R20 went to 85 ( no longer hitting 90+, so that's progress). Will retry some testing when it's colder (night).

Faster? Yes. But 1-2% is nothing.
Did it worth the 95 euro cost for the cooler ? If you look at performance... no.
But... I spend 8-10 hours per day working on my computer ( job ), so for NOISE... it's a definite YES. Even at 100% it has a pleasant "wind like" sound, not a harsh "turbine" sound. I LOVE IT.
 
Last edited:
Joined
Jun 2, 2017
Messages
1,898 (1.96/day)
System Name Best AMD Computer
Processor AMD TR4 1920X
Motherboard MSI X399 SLI Plus
Cooling Alphacool Eisbaer 420 x2 Noctua NHU-14S TR4
Memory Gskill RIpjaws 4 3000MHZ 48GB
Video Card(s) Sapphire Vega 64 Nitro, Gigabyte Vega 64 Gaming OC
Storage 6 x NVME 480 GB, 2 x SSD 2TB, 5TB HDD, 2 TB HDD, 2x 2TB SSHD
Display(s) Acer 49BQ0k 4K monitor
Case Thermaltake Core X9
Audio Device(s) Corsair Void Pro, Logitch Z523 5.1
Power Supply Corsair HX1200!
Mouse Logitech g7 gaming mouse
Keyboard Logitech G510
Software Windows 10 Pro 64 Steam. GOG, Uplay, Origin
Benchmark Scores Firestrike: 24955 Time Spy: 13500
Extremely unlikely, there's over 30 degrees in this room...



They weren't kidding in the reviews that installation is a breeze.

It literally took me only 5 minutes to mount it, and another 10 to admire the craftsmanship...

But now my computer is all over the place in colors ... oh well ;)

First observation:
I can't hear it !
The noise outside in the street is stronger than the fan at auto-speed.
Even at full speed (~2000 rpm it's barely audible inside the case)
WOW !

---- First quick round of benchmarks (Single thread) ---
Note: It's VERY hot here, I don't have AC and it's afternoon. Room thermometer shows around 30 degrees Celsius ambient ( 86 F )

The percentages are related to this post:
https://www.techpowerup.com/forums/threads/matisse-ryzen-3000-overclocking-undervolting.257505/post-4084338 (The "Default" 100% reference)

wPrime 32M - 29.0 (101.7%)
SuperPI 4M - 45.8 (101.5%)
Cine R10 - 10156 (100.7%)
CpuZ - 517 (100.2%)
3DMark2003 CPU T1 - 58.8 (102.8%)
CPU-M 1.6 - 20310 (101.2%)

Extra: 7zip single threaded (32MB dic): 7014 MIPS, 5227 KB/s Comp, 90546 KB/s Decomp
Also 7zip but 16 threads: 73860 MIPS, 51410 KB/s Comp, 1000594 KB/s Decomp

One observation about clocks, during all the single threaded tests it was consistently boosting between 4.3 and 4.4, with tiny spikes to 4.45 ( I have PBO+100 enabled )
Max temp during 7zip ran for 5 mins ... ~ 81 degrees.
Cinebench R20 went to 85 ( no longer hitting 90+, so that's progress). Will retry some testing when it's colder (night).

Faster? Yes. But 1-2% is nothing.
Did it worth the 95 euro cost for the cooler ? If you look at performance... no.
But... I spend 8-10 hours per day working on my computer ( job ), so for NOISE... it's a definite YES. Even at 100% it has a pleasant "wind like" sound, not a harsh "turbine" sound. I LOVE IT.
Excellent, based on my own testing these Noctua coolers are indeed golden. I agree totally with the noise argument. I have one on my TR 4 build and it keeps the 1900x in the 30s while idle and a max of 67 C when gaming.
 

HTC

Joined
Apr 1, 2008
Messages
3,804 (0.88/day)
Location
Portugal
System Name HTC's System
Processor Ryzen 5 2600X
Motherboard Asrock Taichi X370
Cooling NH-C14, with the AM4 mounting kit
Memory G.Skill Kit 16GB DDR4 F4 - 3200 C16D - 16 GTZB
Video Card(s) Sapphire Nitro+ Radeon RX 480 OC 4 GB
Storage 1 Samsung NVMe 960 EVO 250 GB + 1 3.5" Seagate IronWolf Pro 6TB 7200RPM 256MB SATA III
Display(s) LG 27UD58
Case Fractal Design Define R6 USB-C
Audio Device(s) Onboard
Power Supply Corsair TX 850M 80+ Gold
Mouse Razer Deathadder Elite
Software Ubuntu 19.04 LTS
Okay, will do.
Cooler will arrive first it seems ( tomorrow ). Memory later in the week.
Don't forget to also test that "high tech extra wind source" :laugh:
 
Joined
Dec 10, 2014
Messages
437 (0.23/day)
Location
Nowy Warsaw
System Name SYBARIS
Processor AMD Ryzen™ 5 3600
Motherboard MSI Arsenal Gaming B450 Tomahawk
Cooling Cryorig H7 Quad Lumi
Memory Team T-Force Delta RGB 2x8GB 3200CL16
Video Card(s) Colorful GeForce RTX 2060 6G V2
Storage WD Black WD1003FZEX 1TB + Crucial MX500 500GB
Display(s) LG 22MP68VQ-P 22" 75hz IPS
Case In Win Mana 136
Audio Device(s) HyperX Cloud X | iVOOMi iVO-169SUFBT 2.1
Power Supply Cooler Master G550M
Mouse Logitech G102 Prodigy | Logitech G402 Hyperion Fury
Keyboard Fantech MK871 RGB TKL Outemu Blue mechanical keyboard
Software Windows 10 Education 1909 x64
@Wavetrex how hot is there in NL? It's 33c here during the day.
 
Joined
Dec 31, 2009
Messages
16,358 (4.45/day)
Bad time to test new computer hardware which changes speeds based on temperature ;)
Which is why most people normalize temps and go from there. So, essentially, you pick a temp to start with, say 22C, and whatever your ambient is, add/subtract that from the results. So if my room was 26C at the time of testing, I would subtract 4C from all my results.

Martin/Skinee labs used this method years ago when they were around. Or...... you can use a delta which temps shouldn't matter anyway. ;)
 
Joined
Nov 1, 2018
Messages
570 (1.26/day)
Yes but the problem with Zen 2 core... it boosts depending on ABSOLUTE temperature, not the relative over ambient.

I'm not a rich youtuber that can control the temp with LN2 *cough*Steve , or run AC at 100% to keep ambient at 20 at all times *cough*Jay ...
And most of us aren't either.

So we deal with real-life situations, when we can't "normalize" anything... If the room has 33 degrees like @Khonjel 's... then that's his starting temperature and his potential Zen 2 cores will run slower than for someone that burns 1500W of electricity per hour to cool the room to 20.
 
Joined
Nov 18, 2010
Messages
4,826 (1.44/day)
Location
Rīga, Latvia
System Name HELLSTAR
Processor Intel 5960X @ 4.4GHz
Motherboard Gigabyte GA-X99-UD3
Cooling Custom Loop. 360+240 rads. 5x Nidec Servo Gentle Typhoons. AC CUPLEX KRYOS NEXT.
Memory 4x8GB Corsair Vengeance LPX 3000MHz 15-17-17-36 CR1
Video Card(s) ASUS 1080 Ti FE + water block
Storage Optane 900P + Samsung PM981 NVMe 1TB + 750 EVO 500GB
Display(s) Philips PHL BDM3270
Case Phanteks Enthoo Evolv ATX Tempered Glass
Audio Device(s) Sound Blaster ZxR
Power Supply Fractal Design Newton R3 1000W
Mouse Razer Basilisk
Keyboard Razer Deathstalker
Software Windows 10 insider
Yes but the problem with Zen 2 core... it boosts depending on ABSOLUTE temperature, not the relative over ambient.

I'm not a rich youtuber that can control the temp with LN2 *cough*Steve , or run AC at 100% to keep ambient at 20 at all times *cough*Jay ...
And most of us aren't either.

So we deal with real-life situations, when we can't "normalize" anything... If the room has 33 degrees like @Khonjel 's... then that's his starting temperature and his potential Zen 2 cores will run slower than for someone that burns 1500W of electricity per hour to cool the room to 20.
Get a ticket to Alaska or Siberia? :D Case solved.

It is +18C outside for me now too thou lol :D
 
Joined
Nov 1, 2018
Messages
570 (1.26/day)
New GN video about this exact subject...

New toy to play with... (probably a waste of money, just like for those that bought 3600X and 3800X ... ^.^)

127568




127571
127570
 
Top