German tom's hardware guide just got the 5700 xt within an inch of 2080 performance by boosting it to 2.276 GHz @1.25V with an EK waterblock.
Using 250 watts.
Google translated article:
https://translate.google.com/translate?sl=auto&tl=en&u=https%3A%2F%2Fwww.tomshw.de%2F2019%2F07%2F11%2Fungefesselt-radeon-rx-5700-xt-auf-ueber-22-ghz-uebertaktet-break-the-limits-mit-den-neuen-softpowerplaytables-fuer-die-rx-5700-und-rx-5700-xt%2F igor'sLAB
Unrestrained: Radeon RX 5700 XT overclocked to over 2.2GHz. Break the limits with the new SoftPowerPlayTables for the RX 5700 and RX 5700 XT
By: Igor Wallossek on 07/11/2019 at 9:41 am posted in graphics cards
In contrast to Nvidia's restrictions, you can virtually override and change the default settings of the Radeon cards - be it clock caps, power limit or even the voltage. Always provided you have the right profile at hand. The whole thing is called SoftPowerPlayTable and, for the end user quite simple, can also be quickly implemented in the registry. As with the Radeon VII and earlier at Vega, our forum user hellm has once again done a great job here and left me the new tables to experiment. Yes, the RTG (Radeon Tweaker Group) is alive!
Important preliminary remark
But I have to preface this with an important preliminary remark. The here linked in the article SoftPowerPlayTables (short SPPT) are an experimental intervention in the system and certainly not suitable for 24/7 use in production systems! Especially the now possible, higher voltages are a big unknown in the usual overclocker-Milchmädchen bill. Because if you overdo it, you might save the money for the next faster card at the moment, but you risk losing your hardware and thus the entire investment for a long time. The article should show what we could do with it and where there are possible limits is nothing more than a short description of what we have done. No liability is assumed for your hardware and the consequences, because you are completely responsible for that!
Test the limits!
It was a tedious approach, because there was no experience. By the way, you do not have to use a chiller to get similar values, simple water cooling is fine if you do not overdo it, so you do not even have to use a chiller. The delta between the water temperature and the GPU diode was in extreme cases about 35 degrees, with only medium overclocking then much lower. As long as you remain significantly below 50 ° C GPU temperature, even with the maximum clock actually no major hurdles. Below about 45 ° C is ideal, but requires a water temperature of 30 ° C maximum. You can do it that way, but not with air coolers.
AMD's Power Tune on the Navi cards is a pretty interesting thing, because the clock control is usually something of Chill, the power-saving feature of the AMD driver. The maximum clock (including voltage) is therefore not always on, but only if the graphics load requires it. This style of play is quite smart, of course leaves no constant tact to specify, but a good timing, which adjusts the game load dynamically. If you like it consistently, Witcher 3 invites you to stop. Then the stove beats as if Hansel and Gretel had to burn 100 witches at once.
In the WattMan you can, depending on the implemented SPPT, adjust the settings of the card (chip quality) and cooling:
For my test with the 2.2 GHz + peak clock I have chosen a setting with 2276 MHz for the clock and the full 1.25 Volt. The curve can be left as it is and later flattened if necessary. For the power consumption that plays almost no role. The setting for the power I have set to 95%. we'll see on the next page that this does not end in extravagant stream orgies, but it may be important for a smoother frametimes.
That's exactly what I noticed when I logged the frequencies. In the maximum profile, we see a slightly softer course with fewer drops in fluctuating loads. I have again tested Shadow of the Tomb Raider in Full HD, but all settings on Ultra. And no, it does NOT limit the CPU, as many suspect, because even an RTX 2080 Ti OC scales well with shaders and clock, which I've tried in a plausibility test, of course, before times. Only the GPU is limited here.
These high peak clock rates are quite remarkable for this card because whenever they occur, the graphics load is especially high.
I'll show you on the next page, that the performance has risen again neatly and the power consumption is NOT extremely exploded. The increase in tension at this point had no dramatic consequences. But, I must emphasize once again, the high voltages cause even with perfect cooling in the long run a creeping death. And that's why it's not for everyday use. You have to pay attention!
Test system and measurement methods
I have already described the new test system and the methodology in the basic article " How we test graphics cards , as of February 2017 " (English: " How We Test Graphics Cards ") in great detail and therefore reference only this detailed description for the sake of simplicity , So if you want to read everything exactly again, is invited to do so.
Those who are interested in the summary in tabular form will quickly get a quick overview:
Test systems and measuring rooms
Hardware:
Intel Core i7-8700K @ 5 GHz
MSI Z370 Gaming Pro Carbon AC
16GB KFA2 DDR4 4000 Hall Of Fame
1x 1 TB Toshiba OCZ RD400 (M.2, System SSD)
2x 960 GB Toshiba OCZ TR150 (Storage, Images)
Be Quiet Dark Power Pro 11, 850-watt power supply
Cooling:
Alphacool Ice Block XPX + EKWB Vector Radeon RX 5700
8x Be Quiet! Silent Wings 3 PWM, 2x 480mm Alphacool Nexxos UT60 480mm
Thermal Grizzly Kryonaut (for radiator change)
Casing:
Lian Li PC-T70 with expansion kit and modifications
Modes: Open Benchtable, Closed Case
Monitor: Eizo EV3237-BK
power consumption:
Non-contact DC measurement on the PCIe slot (riser card)
Non-contact DC measurement on the external PCIe power supply
Direct voltage measurement on the respective feeders and on the power supply
2x Rohde & Schwarz HMO 3054, 500 MHz multichannel oscilloscope with memory function
4x Rohde & Schwarz HZO50, current clamp adapter (1 mA to 30 A, 100 KHz, DC)
4x Rohde & Schwarz HZ355, probe divider (10: 1, 500 MHz)
1x Rohde & Schwarz HMC 8012, digital multimeter with memory function
thermography:
Optris PI640, infrared camera
PI Connect evaluation software with profiles
acoustics:
NTI Audio M2211 (with calibration file)
Steinberg UR12 (with phantom power for the microphones)
Creative X7, Smaart v.7
own anechoic measuring room, 3.5 x 1.8 x 2.2 m (LxTxH)
Axial measurements, perpendicular to the center of the sound source (s), measuring distance 50 cm
Noise development in dBA (slow) as RTA measurement
Frequency spectrum as graphic
operating system Windows 10 Pro (1903, all updates)
contents
Introduction and Settings
Gaming performance and power consumption
SoftPowerPlayTables - Application and Download
f
