MSI PRO Z690-A WIFI DDR4 test reports when running MSI BIOS 1.70 and Dasharo 1.0.0 with PL1/PL2 = 50W

Dasharo version Dasharo 1.0.0 for MSI PRO Z690-A WIFI DDR4

Hardware compatibility Same as https://github.com/Dasharo/dasharo-issues/issues/158 Intel Core i5 12600K 64 GiB - 4 * Kingston Fury Beast Black KF432C16BB1/16 @ 2400 MHz

Test[s] information

This is pretty much a benchmarking session between MSI BIOS 1.70 and Dasharo 1.0.0. My goal is to find performances differences between the propietary AMI and Coreboot Firmwares, but it is extremely hard to do so because the starting point is not the same for both. Dasharo uses Intel default values for everything whereas MSI Firmware default values benefit it in many ways.

The most interesing settings I found are the ones related to Loadline Calibration. Dasharo uses values that are equivalent to set on the MSI propietary CPU AC Loadline and CPU DC Loadline to 170 (Or using CPU Lite Load Mode 18, which is equivalent), whereas MSI default is 80. The lower values gives less extra Voltage under CPU load with incidentally reduces power consumption and gives Turbo the potential to read higher clock speeds before hitting Power Limits. For example, when initiating the Cinebench R23 Multi Core benchmark, Dasharo was peaking at a Package TDP of 133W (Above PL1) with maxed Turbo clocks, but after a minute (Tau is 28 seconds?) it reduced power consumption to 113W with lower clocks to compensate (Curiously, this is below the 125W PL1 limit). If I set both Loadline values to 170 in MSI, it had identical power consumption and clock speeds than Dasharofor a minute, albeit afterwards it went from 133W to 125W (Which makes sense) with higher clocks than Dasharo but slighty lower than when above PL1. For some reason Dasharo isn't reaching the full PL1 power budget. Meanwhile, using MSI Loadline default values of 80, Package TDP reached 111W with maxed Turbo clocks for the entire duration of the benchmark, giving it both a power consumption and performance advantage. Note that I'm not uploading these benchmarks because they lack polishing and I most likely would redo them.

I decided to test with PL1 and PL2 set to 50W (Which is one of the few options that I can modify in Dasharo from within Windows with Intel XTU, which is otherwise not unlocked for overclock), and I found quite interesing priorities about who gets the power budget, which points out to a totally different algorithm. For example, there is a benchmark where Dasharo tries to max out P-Cores clock speeds with E-Cores remaining @ 800 MHz whereas MSI instead favours E-Cores and drops a single P-Core clock speed to almost 1 GHz lower than the other 5 P-Cores.

FIRMWARE CONFIGURATION (MSI 1.70 uses Microcode 0x22 whereas Dasharo Microcode is 0x1F) See photo: firmwareconfiguration

Note than these values are SPECIFICALLY for the 12600K. I have another system nearby with a MSI PRO Z690-A DDR4 (Non-WiFi) and a Core i5 12400, and the default MSI values are totally different.

For simplification purposes, these settings are shared by all attached benchmarks: PL1 50W (On Dasharo this is changed using Intel XTU, on MSI, from within Firmware) PL2 50W (On Dasharo this is changed using Intel XTU, on MSI, from within Firmware) Tau 28s ICCMAX 175A LCC Auto (Unknow on Dasharo) High Performance (Windows Power Profile)

These are the changes which I'm benchmarking: CPU Lite Load Control Advanced AC_LL 80 (MSI default) or 170 (MSI & Dasharo) DC_LL 80 (MSI default) or 170 (MSI & Dasharo) (These two values are equivalent to use CPU Lite Load Normal with Mode 9 (Auto) and Mode 18, respectively) Turbo On or Off (Changed from within Windows with ThrottleStop)

AC_LL and DC_LL seems to be readable in Windows by using HWinfo64 (It is Domain Loadline (AC/DC) in mOhm), whereas I don't know if LCC can be checked via Software. Nor I know what MSI default for it is, and whenever it changes based on the previously mentioned two values. Also, ThrottleStop reports than PL4 on MSI is 0, whereas it is 241W on Dasharo.

I got the idea of disabling Turbo to see if I can finally get to compare both Firmwares with maxed clocks, which seems impossible with Dasharo due to slighty exceeding the PL1 power budget for the 12600K. That will be my next comparison.

Have fun with the data dump (Individual XTU logs for every benchmark run and Screenshots). Ignore ThrottleStop monitoring and the right panel on Intel XTU screenshots since these represent the moment where I took the screenshot, use the Intel XTU left panel (The graph) and the CSV files to correctly monitors values. Need to figure out a better way to present the results...

OS & DRIVERS Windows 11 Pro

Manually installed Intel i225-V NIC for basic network connectivity (Wired_driver_27.4_x64.zip), then Windows 11 downloaded and installed every other Driver by itself. So I have whatever latest it decided to automatically get.

TWEAKS Set Intel High Definition Audio Controller to MSI mode By default, the Intel High Definition Audio Controller (00:1f.3) uses as Interrupt mode legacy IRQ instead of MSI (Message Signaled Interrupt). This increases latency Check if on Device Manager the IRQ is positive (Legacy IRQ) or negative (MSI) by doing as follows: Open Device Manager Menu View -> Resources by type (PC Name) -> Interrupt request (IRQ) -> High Definition Audio Controller should have a negative number (Also check for any other PCI Device that should use MSI) If not using MSI, open regedit and manually set the value of MSISupported property to 1 in the following location: Computer\HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\PCI\VEN_8086&DEV_7AD0&SUBSYS_9D251462&REV_11\3&11583659&0&FB\Device Parameters\Interrupt Management\MessageSignaledInterruptProperties MSISupported 1

TOOLS Intel XTU 7.8.0.21 Changing Intel Processor settings (Tweaking/Overclocking) on Windows, and also for Monitoring. Also has a Benchmark (Never used it). ThrottleStop 9.5 Changes some settings than Intel XTU can't, also for Monitoring (But worse than XTU). Also has a Benchmark. Cinebench R23 Standard MultiThreading and Single Threading Benchmark.

THROTTLESTOP Open ThrottleStop and Intel XTU. On ThrottleStop, select the intended Power Profile (High Performance, Balanced and Power Saver. They match Windows 11 default 3 Power Profiles, and actually tells Windows to set them). On ThrottleStop, press the TS Bench Button, configure with Priority Normal, Threads 16 (For 12600K), Size 7680M, MHz Fixed. Each run is considered a single benchmark. When ready to benchmark, enable File Logging on XTU before starting it. Remember to stop File Logging on XTU after benchmark is finished.

MSI BIOS 1.70 AC_LL/DC_LL 80 Turbo Yes Fixed MHz 569426 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hpthrottlestop-tsbench.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hpthrottlestop-tsbench.png 1 P-Core - 2.7 GHZ @ 0.94V 5 P-Cores - 3.6 GHz @ 0.94V 4 E-Cores - 2.8 GHz @ 0.94V Cache - 2.8 GHz @ 0.91V Consistent across the entire run.

MSI BIOS 1.70 AC_LL/DC_LL 170 Turbo Yes Fixed MHz 612681 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hpthrottlestop-tsbench.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hpthrottlestop-tsbench.png 1 P-Core - 2.7 GHZ @ 0.94V 5 P-Cores - 3.45 GHz @ 0.94V 4 E-Cores - 2.8 GHz @ 0.94V Cache - 1.2 GHz @ 0.91V Consistent across the entire run. Pretty much identical results to Turbo No.

Dasharo 1.0.0 AC_LL/DC_LL 170 Turbo Yes Fixed MHz 755285 XTU MONITORING dasharo-1-0-0turbo-yes_hpthrottlestop-tsbench.csv SCREENSHOT dasharo-1-0-0turbo-yes_hpthrottlestop-tsbench.png 6 P-Cores - 3.6 GHz @ 0.96V 4 E-Cores - 0.8 GHz @ 0.96V Cache - 2.9 GHz @ 0.91V Starts to drop CPU Utilization and Active Cores around 2/3 of the benchmark (9 minutes mark).

MSI BIOS 1.70 AC_LL/DC_LL 80 Turbo No Fixed MHz 606957 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hpthrottlestop-tsbench.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hpthrottlestop-tsbench.png 1 P-Core - 2.7 GHZ @ 0.94V 5 P-Cores - 3.6 GHz @ 0.94V 4 E-Cores - 2.8 GHz @ 0.94V Cache - 2.8 GHz @ 0.91V Consistent across the entire run. Benchmark score is significantly lower than Turbo Yes even though clock speeds seems to be the same.

MSI BIOS 1.70 AC_LL/DC_LL 170 Turbo No Fixed MHz 610862 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hpthrottlestop-tsbench.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hpthrottlestop-tsbench.png 1 P-Core - 2.7 GHZ @ 0.94V 5 P-Cores - 3.45 GHz @ 0.94V 4 E-Cores - 2.8 GHz @ 0.94V Cache - 1.2 GHz @ 0.91V Consistent across the entire run. Pretty much identical results to Yes Turbo.

Dasharo 1.0.0 AC_LL/DC_LL 170 Turbo No Fixed MHz 709705 XTU MONITORING dasharo-1-0-0turbo-yes_nothrottlestop-tsbench.csv SCREENSHOT dasharo-1-0-0turbo-yes_nothrottlestop-tsbench.png 6 P-Cores - 3.6 GHz @ 0.95V 4 E-Cores - 0.8 GHz @ 0.95V Cache - 2.9 GHz @ 0.91V Starts to drop CPU Utilization and Active Cores around 2/3 of the benchmark (9 minutes mark). For some reason, significantly better score than with Turbo.

CINEBENCH R23 Open Cinebench R23, ThrottleStop and Intel XTU. On ThrottleStop, select the intended Power Profile (High Performance, Balanced and Power Saver. They match Windows 11 default 3 Power Profiles, and actually tells Windows to set them). On Cinebench R23, press the CPU (Multi Core) Button for benchmark, and, after it finishes, the CPU (Single Core) Button. Each run is considered a single benchmark. When ready to benchmark, enable File Logging on XTU before starting it. Remember to stop File Logging on XTU after benchmark is finished.

MSI BIOS 1.70 AC_LL/DC_LL 80 Turbo Yes MT 12392 ST 1867 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hpcinebench-r23-mt.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hpcinebench-r23-mt.png 1 P-Core - 2.7 GHZ @ 0.93V 5 P-Cores - 3.4 GHz @ 0.93V 4 E-Cores - 2.8 GHz @ 0.93V Cache - 1.2 GHz @ 0.91V Consistent across the entire run.

XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hpcinebench-r23-st.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hpcinebench-r23-st.png Max Core Frequency - 4.8 GHz @ 1.2V Cache - 4.5 GHz @ 0.91V Mostly consistent across the entire run.

SCREENSHOT msi-1-70__pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-yes_hp__cinebench-r23.png

MSI BIOS 1.70 AC_LL/DC_LL 170 Turbo Yes MT 11478 ST 1833 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hpcinebench-r23-mt.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hpcinebench-r23-mt.png 1 P-Core - 2.7 GHZ @ 0.94V 5 P-Cores - 3.25 GHz @ 0.94V 4 E-Cores - 2.7 GHz @ 0.94V Cache - 1.0 GHz @ 0.91V Consistent across the entire run.

XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hpcinebench-r23-st.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hpcinebench-r23-st.png Max Core Frequency - 4.75 GHz @ 1.25V Cache - 3.6 GHz @ 0.91V Mostly consistent across the entire run.

SCREENSHOT msi-1-70__pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-yes_hp__cinebench-r23.png

Dasharo 1.0.0 AC_LL/DC_LL 170 Turbo Yes MT 10581 ST 1658 XTU MONITORING dasharo-1-0-0turbo-yes_hpcinebench-r23-mt.csv SCREENSHOT dasharo-1-0-0turbo-yes_hpcinebench-r23-mt.png 6 P-Cores - 3.4 GHz @ 0.94V 4 E-Cores - 0.8 GHz @ 0.94V Cache - 1.2 GHz @ 0.91V Consistent across the entire run. E-Cores at 0.8 MHz most likely a measuring error and is actually 800 MHz.

XTU MONITORING dasharo-1-0-0turbo-yes_hpcinebench-r23-st.csv SCREENSHOT dasharo-1-0-0turbo-yes_hpcinebench-r23-st.png Max Core Frequency - 4.75 GHz @ 1.25V Cache - 4.3 GHz @ 0.91V Mostly consistent across the entire run. For some reason it hits Power Limit Throttling with a Package TDP of just 30W instead of the whole 50W, which may explain the 2000 ST score difference.

SCREENSHOT dasharo-1-0-0__turbo-yes_hp__cinebench-r23.png

MSI BIOS 1.70 AC_LL/DC_LL 80 Turbo No MT 12369 ST 1444 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hpcinebench-r23-mt.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hpcinebench-r23-mt.png 1 P-Core - 2.7 GHZ @ 0.93V 5 P-Cores - 3.4 GHz @ 0.93V 4 E-Cores - 2.8 GHz @ 0.93V Cache - 1.2 GHz @ 0.91V Consistent across the entire run. Same score than with Turbo Yes.

XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hpcinebench-r23-st.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hpcinebench-r23-st.png Max Core Frequency - 3.7 GHz @ 1.05V Cache - 3.5 GHz @ 0.91V Consistent across the entire run.

SCREENSHOT msi-1-70__pl1-50w_pl2-50w_acdc-80_lcc-auto_turbo-no_hp__cinebench-r23.png

MSI BIOS 1.70 AC_LL/DC_LL 170 Turbo No MT 11510 ST 1446 XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hpcinebench-r23-mt.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hpcinebench-r23-mt.png 1 P-Core - 2.7 GHZ @ 0.94V 5 P-Cores - 3.25 GHz @ 0.94V 4 E-Cores - 2.75 GHz @ 0.94V Cache - 1.0 GHz @ 0.91V Consistent across the entire run. Slighty higher score than with Turbo Yes.

XTU MONITORING msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hpcinebench-r23-st.csv SCREENSHOT msi-1-70pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hpcinebench-r23-st.png Max Core Frequency - 3.7 GHz @ 1.08V Cache - 3.5 GHz @ 0.91V Consistent across the entire run. Same score than with AC_LL/DC_LL 80/80.

SCREENSHOT msi-1-70__pl1-50w_pl2-50w_acdc-170_lcc-auto_turbo-no_hp__cinebench-r23.png

Dasharo 1.0.0 AC_LL/DC_LL 170 Turbo No MT 10605 ST 1303 XTU MONITORING dasharo-1-0-0turbo-no_hpcinebench-r23-mt.csv SCREENSHOT dasharo-1-0-0turbo-no_hpcinebench-r23-mt.png 6 P-Cores - 3.4 GHz @ 0.94V 4 E-Cores - 0.8 GHz @ 0.94V Cache - 1.2 GHz @ 0.91V Consistent across the entire run. E-Cores at 0.8 MHz most likely a measuring error and is actually 800 MHz. Roughly the same clock speeds and MT scores than with Turbo Yes.

XTU MONITORING dasharo-1-0-0turbo-no_hpcinebench-r23-st.csv SCREENSHOT dasharo-1-0-0turbo-no_hpcinebench-r23-st.png Max Core Frequency - 3.7 GHz @ 1.04V Cache - 3.3 GHz @ 0.91V Consistent across the entire run. I'm not sure if the 200 MHz lower Cache can account for an almost 150 ST points difference.

SCREENSHOT dasharo-1-0-0__turbo-no_hp__cinebench-r23.png

I decided to investigate why Cinebench R23 was giving consistently lower Single Core scores on Dasharo. Scores were substantially lower with Turbo on, but even without Turbo, it is common than MSI scores around 1420-1450 and Dasharo could be anywhere from 1200-1300, so this couldn't be power related.

The most obvious things which I noticed is that the Processor Cache (Also known as Uncore) clocks up to 3.5 GHz on MSI whereas on Dasharo it never went above 3.3 GHz, so I decide to lower the Multiplier to 33x on MSI to match Dasharo. I have no idea which of those two is supposed to be the correct Intel default. Worth noting than this setting is much preferable to be changed from within MSI BIOS than runtime with Intel XTU, since there are several reports about issues changing the Processor Cache Ratio, including BSODs, which happened to me two times (And I was underclocking it, not overclocking), and that the Cache clock speed can get stuck and never downclock again even after a full power cycle, requiring to reflash the Firmware: https://community.intel.com/t5/Processors/Intel-XTU-Alder-Lake-Ring-issue/m-p/1346576/highlight/true#M55474 This also happened to me and had to use flashrom to flash the backup MSI image again because Cache clock speed was fixed to 3.3 GHz after trying to set it with Intel XTU and BSODing. Note than ThrottleStop can also change Cache ratio, but didn't tested if it has the same issues than XTU.

Also, I decided to run with Intel standard PL1/PL2 125W/150W and no Turbo, since what I want to see is if both Firmwares perform the same at the same clock speeds, and the 50W PL1/PL2 never let them clock to maximum. There is a lot of power headroom without Turbo and the default Power Limiters, and this is seen by Loadline values now making no difference in performance, albeit they do on power usage.

After some benchmarking I noticed that MT scores were almost evenly matched at just above 14000, with ST always looking inconsistent. Do note that I didn't tested multiple runs but Dasharo seems to be 1% lower on MT than MSI, and this seems to go beyond individual run variance, cause it is always slighty lower, but by a small margin. Also, Dasharo on warm reboots scores 2-3% lower than if doing a full power cycle (13700 vs 14000), but I didn't benchmarked this, just noticed that behavior because I did a warm reboot and scores that were near even now showed a bigger difference.

The Cinebench R23 ST score issue ended up being related to Core affinity. When running the ST benchmark, Intel XTU showed than there was ocassional load in E Cores which didn't happened on the MSI, where only P Cores were loaded, albeit Windows CPU Scheduler seems to like to ocassionally jump the Thread around anyways. Manually setting Processor affinity on Windows Task Manager for the Cinebench process didn't work as expected since Cinebench resets affinity when you tell it to do the benchmark, so you have to set it AFTER the benchmark initiates (This seems to also be the case with CPU-Z benchmark). I ended up using an application, Process Lasso, that has an option, Forced Mode, which continously reapplies the affinity settings in case the application resets them. With that, now I could finally get near identical scores in Cinebench ST benchmark to MSI propietary Firmware. So the difference must be some Firmware level hint about how Windows should load the Cores.

CINEBENCH R23 Open Cinebench R23, ThrottleStop, Intel XTU and Process Lasso. On ThrottleStop, select the intended Power Profile (High Performance), and tick Disable Turbo (For some reason Intel XTU does NOT let me to do so, the option is grayed out, but it works on ThrottleStop). On Process Lasso, Right Click on the Cinebench R23 process, open the menu at CPU Affinity -> Always -> Select CPU affinity, then select the desired CPUs (On Windows 11, first comes the Physical Core then its SMT/Hyper Threading Logical Core, so CPU0 and CPU1 are the same Core. E Cores are at the end, and identified by an extra E). On Cinebench R23, you can initiate the MT benchmark by pressing the CPU (Multi Core) Button, and, after it finishes, the CPU (Single Core) Button. Each run is considered a single benchmark. You can also go to File -> Preferences and use a custom amount of Threads, which I did (4, 12 and 16). When ready to benchmark, enable File Logging on XTU before starting it. Remember to stop File Logging on XTU after each benchmark is finished.

I benchmarked the following on MSI 1.70 with both 80 and 170 Loadline ACDC, and Dasharo 1.0.0: 16 Threads MT default CPU affinity 16 Threads ST default CPU affinity (This is where Dasharo scores much lower) 12 Threads MT P-Core only affinity 12 Threads ST P-Core only affinity (This is where it matches MSI) 4 Threads MT E-Core only affinity 4 Threads MT E-Core only affinity

All had consistent 6 P-Cores 3.7 GHz, 4 E-Cores 2.8 GHz and Cache 3.3 GHz clock speeds, difference being usually Voltage and power consumption.

MSI BIOS 1.70 --- AC_LL/DC_LL 80 - Turbo No

16 Threads all Cores --- MT 14119 @ Package TDP 69W / ST 1443 @ Package TDP 15W SCORES SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_scores.png MT XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.png MT XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.csv ST XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.png ST XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.csv

12 Threads P-Cores only --- MT 11442 @ Package TDP 60W / ST 1427 @ Package TDP 15W SCORES SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_scores.png MT XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_mt.png MT XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_mt.csv ST XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_st.png ST XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_st.csv

4 Threads E-Cores only --- MT 2699 @ Package TDP 17W / ST 751 @ Package TDP 9W SCORES SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_scores.png MT XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_mt.png MT XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_mt.csv ST XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_st.png ST XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_st.csv

MSI BIOS 1.70 --- AC_LL/DC_LL 170 - Turbo No

16 Threads all Cores --- MT 14075 @ Package TDP 77W / ST 1443 @ Package TDP 16W SCORES SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_scores.png MT XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.png MT XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.csv ST XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.png ST XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.csv

12 Threads P-Cores only --- MT 11446 @ Package TDP 68W / ST 1440 @ Package TDP 16W SCORES SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_scores.png MT XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_mt.png MT XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_mt.csv ST XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_st.png ST XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_st.csv

4 Threads E-Cores only --- MT 2758 @ Package TDP 18W / ST 768 @ Package TDP 10W SCORES SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_scores.png MT XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_mt.png MT XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_mt.csv ST XTU SCREENSHOT msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_st.png ST XTU MONITORING msi-1-70pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_st.csv

Dasharo 1.0.0 --- AC_LL/DC_LL 170 - Turbo No

16 Threads all Cores --- MT 14026 @ Package TDP 76W / ST 1313 @ Package TDP 16W SCORES SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_scores.png MT XTU SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.png MT XTU MONITORING dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.csv ST XTU SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.png ST XTU MONITORING dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.csv

12 Threads P-Cores only --- MT 10988 @ Package TDP 67W / ST 1415 @ Package TDP 15W SCORES SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_scores.png MT XTU SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_mt.png MT XTU MONITORING dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_mt.csv ST XTU SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_st.png ST XTU MONITORING dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_12threads-pcores_st.csv

4 Threads E-Cores only --- MT 2725 @ Package TDP 17W / ST 759 @ Package TDP 11W SCORES SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_scores.png MT XTU SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_mt.png MT XTU MONITORING dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_mt.csv ST XTU SCREENSHOT dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_st.png ST XTU MONITORING dasharo-1-0-0pl1-125w_pl2-150w_acdc-170_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_4threads-ecores_st.csv

msi-1-70__pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hp__cinebench-r23_16threads-allcores_scores msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_mt.csv msi-1-70pl1-125w_pl2-150w_acdc-80_lcc-auto_turbo-no_cache-33ghz_hpcinebench-r23_16threads-allcores_st.csv