22 hours ago, brucewtb said:

Thanks for the links.  Hardware Unboxed seem to have opened something of a Pandora’s Box (no pun intended). Unsure what to make of it all with so many conflicting opinions.  Was about to order a Hero now not so sure.

The Maximus XI Hero has a pretty sturdy looking 8-phase regulator, uninformed opinions from "Hardware Unhinged" notwithstanding.  I like the motherboard so far with the hexacore 8086K on-board.  Temps look really good across the board when scanned with a Fluke IR thermometer.  I'm into my second hour of Prime95 with the 8086K at 5.2GHz, 1.35 Vcore, temps barely peeking above 70 deg C (HT Off).  With HT On, so far I've taken it to 5.1 at 1.31v, temps peaked at 74 deg C in Prime95 (no AVX).  Stock thermal paste, too...guess I'm going to punt on delidding the chip with these numbers.

I'm fighting the urge to just bag the overclocking test runs and load P3D.  I know better, so I wont.  But I REALLY want to.  😉

Cheers

14 hours ago, Rob Ainscough said:

. … I know reviewers like to be "first" but there is obviously lots of inconsistent data across the reviewers.

Cheers, Rob.

 

 

 

 

It's been pretty consistent re 9900K, Most are seeing high temps. It was mostly Linus Tech Tips and Tom Logan that saw (unrealistic) very low temps. That was because they were using cheaper motherboards with a meagre 4 phase VRM, nowhere near capable enough of extracting the best from the chip, hence it was limiting the clock speed. Whereas the other reviews were using high spec motherboards. 

 

 

 

Edited October 24, 20187 yr by martin-w

My teammate tested a quite good 9900k (no Delidd) on a XI Gene with 4400mhz mems

cooling AIO X62 no problem 5.4ghz HT off Cinebenc R-15 , temps ok 87-90c

its not setup for flightsim  ,think its capable to do 5.4 with a customloop +5.5 with Chiller in P3D (HT-OFF)

 

I would like to see some tests run for P3D that are more realistic for the vast majority of users.  Running super high end custom cooling and a chiller with the most expensive MBs doesnt tell the picture for many who would consider upgrading but certainly not at that price point with everything combined. 

Anyone doing a setup with say a very good mid range board but still has 10-12 phase power, with a 9700k and a mainstream AIO? 

I'm going to need a bigger radiator ....

@westman  How come you didn't tell me <g> 

So yesterday I dumped the 24/7 5.0 de8beaur cpu bios setting base i've used since I built my rig, and pulled up the 4133 anchor. Looking for a new direction I switched to Slamm's 5.1 setup ...WOW! what a difference. I'm smashing all my previous bench records, by a good margin. And the most relevant and important note ....The LAX approach feels like the SFO approach now ...pictures are coming.  Interesting thing here is the 52/4700 4133  de8beur 5.0 base I used prior doesn't even close to the performance I'm getting with Slamms 5.1 base at 52/4800 3866 c16 ..... It's clear to me that Slamm magic comes from the VRM page in his bios setting .....the guy appears to be the electricity wizzard at the ROG lab.

  

No time for picts ...will post later:

highlights: 5.2/4800 3866 c16

Reabench: 3 run average of 170  175 hi, just ate up the encoding bench ...292 image score --- crazy and repeatable scores

232 cinebech  ...does 238 at 5.3 can do 5.3 as long as there's no avx involved  ...

mem scores still show low 39's latency

50,700 3dmark 06.

Oh, did I mention P3d is a whole different experience? 🙂

Edited October 24, 20187 yr by FunknNasty

29 minutes ago, FunknNasty said:

No time for picts ...will post later:

highlights: 5.2/4800 3866 c16

Reabench: 3 run average of 170  175 hi, just ate up the encoding bench ...292 image score --- crazy and repeatable scores

232 cinebech  ...does 238 at 5.3 can do 5.3 as long as there's no avx involved  ...

mem scores still show low 39's latency

50,700 3dmark 06.

Oh, did I mention P3d is a whole different experience? 🙂

 

This is for you....

 

15 hours ago, w6kd said:

The Maximus XI Hero has a pretty sturdy looking 8-phase regulator, uninformed opinions from "Hardware Unhinged" notwithstanding. 

 

Asus forum thread on this here...

https://rog.asus.com/forum/showthread.php?105731-Can-someone-clear-up-misinformation-on-the-Z390-XI-Hero-s-power-phasing

 

It's limited to 95 watts though still. Hence the issue. Suspected that Linus and Tom Logan failed to increase the power limits in the BIOS. Hence the 95 watt limit. 

 

Quote

 

No idea why this below has appeared. Ignore.

 

Edited October 24, 20187 yr by martin-w

2 hours ago, FunknNasty said:

I'm going to need a bigger radiator ....

@westman  How come you didn't tell me <g> 

So yesterday I dumped the 24/7 5.0 de8beaur cpu bios setting base i've used since I built my rig, and pulled up the 4133 anchor. Looking for a new direction I switched to Slamm's 5.1 setup ...WOW! what a difference. I'm smashing all my previous bench records, by a good margin. And the most relevant and important note ....The LAX approach feels like the SFO approach now ...pictures are coming.  Interesting thing here is the 52/4700 4133  de8beur 5.0 base I used prior doesn't even close to the performance I'm getting with Slamms 5.1 base at 52/4800 3866 c16 ..... It's clear to me that Slamm magic comes from the VRM page in his bios setting .....the guy appears to be the electricity wizzard at the ROG lab.

I don’t consider myself stupid because I can fly Concorde and handle the fuel management as well but when I read the above I haven’t a clue what it all means.

Do we need a separate sub forum for really clever people and you have to pass an exam before you’re allowed in? 🤣

4 hours ago, Ray Proudfoot said:

I don’t consider myself stupid because I can fly Concorde and handle the fuel management as well but when I read the above I haven’t a clue what it all means.

Do we need a separate sub forum for really clever people and you have to pass an exam before you’re allowed in? 🤣

 

6 hours ago, FunknNasty said:

I'm going to need a bigger radiator ....

@westman  How come you didn't tell me <g> 

So yesterday I dumped the 24/7 5.0 de8beaur cpu bios setting base i've used since I built my rig, and pulled up the 4133 anchor. Looking for a new direction I switched to Slamm's 5.1 setup ...WOW! what a difference. I'm smashing all my previous bench records, by a good margin. And the most relevant and important note ....The LAX approach feels like the SFO approach now ...pictures are coming.  Interesting thing here is the 52/4700 4133  de8beur 5.0 base I used prior doesn't even close to the performance I'm getting with Slamms 5.1 base at 52/4800 3866 c16 ..... It's clear to me that Slamm magic comes from the VRM page in his bios setting .....the guy appears to be the electricity wizzard at the ROG lab.

  

Could you please expand? What BIOS setting did you change. You overclock your RAM to 4133 and say a huge gain the P3D? Please detail before/after BIOS settings. Thanks

5 hours ago, Ray Proudfoot said:

I don’t consider myself stupid because I can fly Concorde and handle the fuel management as well but when I read the above I haven’t a clue what it all means.

Do we need a separate sub forum for really clever people and you have to pass an exam before you’re allowed in? 🤣

Yes, an eye exam is required. Lol

Just looked at a bench mark single core performance is only up 6% compared to a 8700k not worth spending the money to upgrade,since that boost if probably from the 5.0 vs 4.7ghz turbo boost. I'll stick with my 8700k I can hit 4.9ghz on 1.33v

12 minutes ago, z06z33 said:

Just looked at a bench mark single core performance is only up 6% compared to a 8700k

From the test results I've seen that are more closely associated with how P3D works in terms of single main sync thread is Civ VI as it's heavier on the CPU side that also utilizes a main sync thread ... it shows about a 12% increase over a 8700K.  (I am a give Civ VI fan BTW, but that's not a BIAS)

Multi-core is important for P3D but Cinebench R15 single/multithread isn't a great emulation of how P3D works … what cause long frames (aka stutters) isn't entirely CPU based, it's overall system throughput from CPU to RAM to GPU to VRAM.  It's always about balance.

I see the 7900X has dropped in price $896 … but I'm hoping the "man from the west" is correct and I can get 5.5Ghz HT off on my setup … and I'm assuming that's 5.5, 5.5, 5.3, 5.3, 5.2, 5.2, 5.2, 5.2  where my current 7900X on water only is 5.0, 5.0, 4.8, 4.9, 5.0, 4.8, 4.8, 4.9, 5.0, 4.9.

Cheers, Rob.

9 hours ago, FunknNasty said:

So yesterday I dumped the 24/7 5.0 de8beaur cpu bios setting base i've used since I built my rig, and pulled up the 4133 anchor. Looking for a new direction I switched to Slamm's 5.1 setup ...WOW! what a difference. I'm smashing all my previous bench records, by a good margin. And the most relevant and important note ....The LAX approach feels like the SFO approach now ...pictures are coming.  Interesting thing here is the 52/4700 4133  de8beur 5.0 base I used prior doesn't even close to the performance I'm getting with Slamms 5.1 base at 52/4800 3866 c16 ..... It's clear to me that Slamm magic comes from the VRM page in his bios setting .....the guy appears to be the electricity wizzard at the ROG lab.

//universal translator on//

So yesterday I dumped the 24 hr 7 day a week 5.0GHz BIOS preset (made by de8auer, a well-known German competitive benchmarking/overclocking champ), and got rid of my 4133 MHz RAM configuration.  Looking to try something new, I switched to Slamm's (a well-known Russian competitive benchmarking/overclocking champ) 5.1 GHz preset (these presets come programmed in the BIOS of the Asus Maximus Apex motherboards).  I'm smashing all my previous bench records, by a good margin. And the most relevant and important note ....The LAX approach feels like the SFO approach now ...pictures are coming.  Interesting thing here is the 5.2 GHz (52 core multiplier and 4700 MHz Cache/Uncore) config based on the de8auer 5.0 preset I used prior doesn't even close to the performance I'm getting at 5.2 GHz with Slamm's 5.1 GHz preset with a core mult of 52, cache speed 4800 MHz with RAM running at 3866 MHz and CAS 16 timings.  It's clear to me that Slamm magic comes from the VRM page in his bios setting .....the guy appears to be the electricity wizzard at the ROG lab.

//universal translator off//

🙂

 

1 hour ago, w6kd said:

//universal translator on//

So yesterday I dumped the 24 hr 7 day a week 5.0GHz BIOS preset (made by de8auer, a well-known German competitive benchmarking/overclocking champ), and got rid of my 4133 MHz RAM configuration.  Looking to try something new, I switched to Slamm's (a well-known Russian competitive benchmarking/overclocking champ) 5.1 GHz preset (these presets come programmed in the BIOS of the Asus Maximus Apex motherboards).  I'm smashing all my previous bench records, by a good margin. And the most relevant and important note ....The LAX approach feels like the SFO approach now ...pictures are coming.  Interesting thing here is the 5.2 GHz (52 core multiplier and 4700 MHz Cache/Uncore) config based on the de8auer 5.0 preset I used prior doesn't even close to the performance I'm getting at 5.2 GHz with Slamm's 5.1 GHz preset with a core mult of 52, cache speed 4800 MHz with RAM running at 3866 MHz and CAS 16 timings.  It's clear to me that Slamm magic comes from the VRM page in his bios setting .....the guy appears to be the electricity wizzard at the ROG lab.

//universal translator off//

🙂

 

rofl

You and Martin can jump in a lake.  🙂

 

...somebody hack my account?

Edited October 25, 20187 yr by FunknNasty