*Official Benchmark Thread* - Post it here or it didn't happen :D

[Mr. Fox]

7 hours ago, electrosoft said:

Monstrous value here used but like new sold and shipped from Asrock 6950XT for $420 via Amazon

 

Glad to see Asrock and Amazon worked out their differences as their store is nearly fully stocked with their products.
🙂

That is pretty cheap.

 

It's nice to see Linus presenting a rational explanation and it really fits with the fact that manual overclockers are generally having none of the issues. Mostly the gamerboyz that try to use the bugged mobo manufacturer "features" and eTVB garbage and misguided sloppy LLC undervolting mistakes. (I don't agree with the advice at the end to install the latest BIOS update. The better solution would be to suggest manual tuning.)

 

 

[Papusan]

3 hours ago, Mr. Fox said:

Speaking of lies and missrepresentations, for those who have drank the Kool-Aid and joined the muppet show of Winduhz 11 apostles, here is some fuel for the fire for those of us that despise the new cancer OS. Windows 11 offers no benefit and is actually harmful to performance. The only thing Winduhz 11 offers is a new level of ugliness for those that prefer a less-useful and tackier-looking GUI. The joke is on Windoze 11 users, but it's not a funny one. Utterly worthless OS downgrade. But, everyone is too busy hating on Intel to talk about this right now.

 

 

Expect unexpected firmware updates for your Intel powered Pc forwards. Because Microsoft do whatever they want. Microsoft will gladly help their partners to screw up your PC or render it bricked. Because they are damn good at this. And no one can do nothing about it. Because you all have accepted their EULA. Microsoft can't be accountable for their partners firmware, software and driver bugs. They just push it quietly on your PC. Same as with their own patches.  Good luck to you all. So.......

 

Microsoft quietly installing KB5001716 Windows 11/10 update for the good of your PC

Neowin · 19 hours ago · Hot! 0

It is that time of the year again when Microsoft has begun installing the KB5001716 update and this time it is for both Windows 10 as well as 11 systems. The update is meant to be good for your PC.

Windows 11 market share on Steam drops below 46%

Neowin · Aug 2, 2024

Valve has published the latest results of its monthly Software & Hardware Survey. According to the findings, Windows 11's market share dropped a bit in July 2024, going below the 46% mark.

 

1 hour ago, Mr. Fox said:

It's nice to see Linus presenting a rational explanation and it really fits with the fact that manual overclockers are generally having none of the issues.

 

No-one want to talk about AMD and their problems.... At least Puget have popped up now to tell the truth about Ryzen chips. And Puget Systems ain't a small Pc-building company.

 

High-end PC builder Puget Systems released records showing that its current failure rate of 13th- and 14th-generation Intel processors is a little over 2%, while systems it has shipped with AMD Ryzen 5000 and 7000 chips have a little over 4% failure rate

 

Puget says its Intel CPU failure rate is lower than AMD Ryzen failures — retailer releases failure rate data, cites conservative power settings

 

Puget claims AMD Ryzen 5000 and 7000 chips have a higher failure rate than Intel 13th- and 14th-gen processors, so far.

[Mr. Fox]

7 hours ago, Papusan said:

No-one want to talk about AMD and their problems.... At least Puget have popped up now to tell the truth about Ryzen chips. And Puget Systems ain't a small Pc-building company.

 

High-end PC builder Puget Systems released records showing that its current failure rate of 13th- and 14th-generation Intel processors is a little over 2%, while systems it has shipped with AMD Ryzen 5000 and 7000 chips have a little over 4% failure rate

 

Puget says its Intel CPU failure rate is lower than AMD Ryzen failures — retailer releases failure rate data, cites conservative power settings

 

Puget claims AMD Ryzen 5000 and 7000 chips have a higher failure rate than Intel 13th- and 14th-gen processors, so far.

Yes, I already knew this and I suspect you did as well, but speaking about it is unpopular. AMD is often protected as the meek and lowly underdog, but the fanboys rush to crucify the big, mean evil Intel and NVIDIA when they make mistakes. One of my sons prefers AMD and he has had two Ryzen CPUs die in the past two years. He does not overclock. He sets PBO and plays games. I have had Intel CPUs die as well, but only one that I have no idea why. The others were executed courtesy of the incompetent idiots at ASUS. The way I look at it is this: pick your poison. None are perfect or free of fault. Decide what you want and buy it. Then deal with whatever happens. Expect the worst. All of them suck and all of them are dishonest. None are honorable or trustworthy. And none of them care about you, your PC, or how it affects you. It's not their problem. We should not expect more from Intel or hold them to a higher standard simply because their failure rate is lower or because their faults are more rare. In the end, what we expect from any of them does not matter. We do not matter. They only care about money, market status and shareholders. They all operate on the same business model.

[tps3443]

4 hours ago, Mr. Fox said:

I saw my posts and forum name and avatar several times at both EVGA and OC.net forums. 🤣

 

No matter what degree of merit any of this new round of psychotic pablum has, all of the negative publicity is going to be harmful to Intel and it could ultimately be more harmful to overclocking enthusiasts if they start locking things down and interfering with user tuning controls. The media is a two-edged sword and they have the power to create more issues than they solve. GN is certainly a trusted source and I do not doubt the accuracy (very well supported by facts) but I do question whether or not any of it is going to benefit anyone and may end up being more harmful than helpful in the long run.

 

One of my biggest take-aways from all of the hype (and it surfaced a few times in the GN video) is the hidden nugget (unintended and inconvenient truth) that refusing BIOS and OS updates when my systems are functioning the way I want them to is an extremely wise personal best practice that I have abided by for more than a decade and I shall continue doing so indefinitely because it has exempted me, almost entirely, from the stupid nonsense that update evangelists have suffered from. If it ain't broke, don't fix it. If you do, it will eventually get broken.

 

Nothing I have seen so far changes my fondness for my nicely binned 13th and 14th Gen i9 CPUs and none of it makes me pause to consider the possibility of AMD as a viable option. All of these big tech companies are equally dishonest scumbags that are rotten from head to toe and never let an opportunity to cover their own asses with lies and misrepresentations to slip by them. None of them give a rat's butt about their employees or the people that buy their products. We cannot trust any of them to be honest about anything... ever. The current mess with Intel causes too many of us to forget the past messes, missteps lies and scams from AMD and NVIDIA. Some fairly recent.

---

Speaking of lies and missrepresentations, for those who have drank the Kool-Aid and joined the muppet show of Winduhz 11 apostles, here is some fuel for the fire for those of us that despise the new cancer OS. Windows 11 offers no benefit and is actually harmful to performance. The only thing Winduhz 11 offers is a new level of ugliness for those that prefer a less-useful and tackier-looking GUI. The joke is on Windoze 11 users, but it's not a funny one. Utterly worthless OS downgrade. But, everyone is too busy hating on Intel to talk about this right now.

 

I recently needed to make a decision for buying a yard tool, and this tool was something I needed now and would also use for years down the road. A “Chainsaw” lol. Very simple item. But most are consumer garbage and disposable just like a lot of the PC’s and laptops available today, you wouldn’t believe most chain saws have a plastic crank, and even plastic crank case housing (Don’t drop it). And just like in the tech PC world, in the chainsaw world their are chainsaw enthusiasts 😂 Yep, and they are overclocking and tweaking and building some crazy high performance chainsaws 😂 haha anyways, I wanted something that would last me, something you don’t toss in the bin when it breaks, something that can be fixed if required, so I started looking at Pro saws for clearing my lot now and in the future for years to come. Good chainsaws are pretty expensive. But I ended up settling on a model.  And within most of these models they have added the usual “Updated fancy tech electronic features” to these chainsaws, This feature is called “Mtronic” this is like an ECU for a chainsaw. It controls fuel, spark, if it’s rich or lean etc. I couldn’t help but think why the heck do I need this on a chainsaw? Fortunately they sale both, and either option is the same price. Moral of the story, I seriously could not decide which one to buy, is newer better? I just thought which one you’d probably buy. And since they still sale the same one they designed without all of these electronics and epa regulations/ and emission restrictions attached, right beside the new fancy pants one, and for the same price, that’s the one I went with. A darn good ole chainsaw with a regular carburetor, no electronics on it, updates were disabled for 10 years on this one (Just like your Windows) 🤣 It’ll always work and it will always run! Newer is not always better. Another thing I am not a fan of is these newer Z790 Apex bios by the way. I’m running 9901 currently, this is the only bios available that has “XMP Tweaked” for the white Apex, and this is a nice bios. Now, time to go knock out some of those darn trees! 😎

[Mr. Fox]

10 minutes ago, tps3443 said:

I recently needed to make a decision for buying a yard tool, and this tool was something I needed now and would also use for years down the road. A “Chainsaw” lol. Very simple item. But most are consumer garbage and disposable just like a lot of the PC’s and laptops available today, you wouldn’t believe most chain saws have a plastic crank, and even plastic crank case housing (Don’t drop it). And just like in the tech PC world, in the chainsaw world their are chainsaw enthusiasts 😂 Yep, and they are overclocking and tweaking and building some crazy high performance chainsaws 😂 haha anyways, I wanted something that would last me, something you don’t toss in the bin when it breaks, something that can be fixed if required, so I started looking at Pro saws for clearing my lot now and in the future for years to come. Good chainsaws are pretty expensive. But I ended up settling on a model.  And within most of these models they have added the usual “Updated fancy tech electronic features” to these chainsaws, This feature is called “Mtronic” this is like an ECU for a chainsaw. It controls fuel, spark, if it’s rich or lean etc. I couldn’t help but think why the heck do I need this on a chainsaw? Fortunately they sale both, and either option is the same price. Moral of the story, I seriously could not decide which one to buy, is newer better? I just thought which one you’d probably buy. And since they still sale the same one they designed without all of these electronics and epa regulations/ and emission restrictions attached, right beside the new fancy pants one, and for the same price, that’s the one I went with. A darn good ole chainsaw with a regular carburetor, no electronics on it, updates were disabled for 10 years on this one (Just like your Windows) 🤣 It’ll always work and it will always run! Newer is not always better. Another thing I am not a fan of is these newer Z790 Apex bios by the way. I’m running 9901 currently, this is the only bios available that has “XMP Tweaked” for the white Apex, and this is a nice bios. Now, time to go knock out some of those darn trees! 😎

You made the right decision. 😆 Newer is always newer.

 

I was still a mechanic when electronic ignition surfaced. I always used Mallory dual-point distributors in my street rods. I had some friends that "upgraded" to MSD electronic ignition kits. The claimed "they never wear out" and "zero maintenance" and I was happy to replace my points and condensor when they got worn out. Very easy and affordable. I laughed at my friends when their ECM failed and they had to save up or work overtime to buy a new one. They were expensive. I reminded them "no moving parts" and "never wear out" and "zero maintenance" was the reason they felt newer was better. 

[Papusan]

36 minutes ago, tps3443 said:

I recently needed to make a decision for buying a yard tool, and this tool was something I needed now and would also use for years down the road. A “Chainsaw” lol. Very simple item. But most are consumer garbage and disposable just like a lot of the PC’s and laptops available today, you wouldn’t believe most chain saws have a plastic crank, and even plastic crank case housing (Don’t drop it). And just like in the tech PC world, in the chainsaw world their are chainsaw enthusiasts 😂

 

What brand did you choose? Here home I put my money on Stihl - Wikipedia products. A German company that always offer quality. They have a lot different tools for the garden and the forrest. And I have always been very pleased with this brand. And they don't change products for the sake of newer has to be better. They rather make revision changes if needed. 

 

 

BTW. Also Nvidia follow the new tech trend........

Nvidia Blackwell GPUs allegedly delayed due to design flaws

 

Firm says Nvidia's skyrocketing AI valuation is in a 'bubble' and 'overhyped' — Elliott says AI apps are not viable

 

Elliott's commentary comes amid a rally in AI-related stocks, driven by investor enthusiasm for generative AI, the FT says. Meanwhile, Nvidia has lost nearly $600 billion in market capitalization since early July, and that says something.

[Mr. Fox]

31 minutes ago, Papusan said:

 

What brand did you choose? Here home I put my money on Stihl - Wikipedia products. A German company that always offer quality. They have a lot different tools for the garden and the forrest. And I have always been very pleased with this brand. And they don't change products for the sake of newer has to be better. They rather make revision changes if needed. 

 

 

BTW. Also Nvidia follow the new tech trend........

Nvidia Blackwell GPUs allegedly delayed due to design flaws

 

Firm says Nvidia's skyrocketing AI valuation is in a 'bubble' and 'overhyped' — Elliott says AI apps are not viable

 

Elliott's commentary comes amid a rally in AI-related stocks, driven by investor enthusiasm for generative AI, the FT says. Meanwhile, Nvidia has lost nearly $600 billion in market capitalization since early July, and that says something.

Stihl are excellent. chainsaws. I agree. That was the brand we always bought back in the day that I actually had a use for them. I also like Husqvarna and always loved their amazing dirt bikes.

 

Here is another amazing repair by Tony at Northwest Repair. I posted a comment suggesting that he reattach the I/O bracket reinforcement to the heat pipes using JB Weld. I am 99.9% certain that will be more than strong enough to hold it forever.

 

[tps3443]

1 hour ago, Papusan said:

 

What brand did you choose? Here home I put my money on Stihl - Wikipedia products. A German company that always offer quality. They have a lot different tools for the garden and the forrest. And I have always been very pleased with this brand. And they don't change products for the sake of newer has to be better. They rather make revision changes if needed. 

 

 

BTW. Also Nvidia follow the new tech trend........

Nvidia Blackwell GPUs allegedly delayed due to design flaws

 

Firm says Nvidia's skyrocketing AI valuation is in a 'bubble' and 'overhyped' — Elliott says AI apps are not viable

 

Elliott's commentary comes amid a rally in AI-related stocks, driven by investor enthusiasm for generative AI, the FT says. Meanwhile, Nvidia has lost nearly $600 billion in market capitalization since early July, and that says something.

 

1 hour ago, Mr. Fox said:

Stihl are excellent. chainsaws. I agree. That was the brand we always bought back in the day that I actually had a use for them. I also like Husqvarna and always loved their amazing dirt bikes.

 

Here is another amazing repair by Tony at Northwest Repair. I posted a comment suggesting that he reattach the I/O bracket reinforcement to the heat pipes using JB Weld. I am 99.9% certain that will be more than strong enough to hold it forever.

 

 

I bought a new Stihl 261 with 20” bar. Yep they are German made pro saws. I love it! Mounted a hour/tach meter to it as well. Cut a few trees so far and it works amazing. 
 

[Clamibot]

The TEC in my MasterLiquid ML360 Sub Zero cooler seems to have died. I can't get it to come back up anymore and it looks like there's a hardware error code being displayed by the cooler. Does anyone have any tips for reviving a thermoelectric element or is it completely toast?

 

If the TEC is irrecoverable, I heard that the Lian Li Galahad II Trinity Performance 360 is one of the best AIOs on the market. What do you guys think of it?

[electrosoft]

14 minutes ago, Clamibot said:

The TEC in my MasterLiquid ML360 Sub Zero cooler seems to have died. I can't get it to come back up anymore and it looks like there's a hardware error code being displayed by the cooler. Does anyone have any tips for reviving a thermoelectric element or is it completely toast?

 

If the TEC is irrecoverable, I heard that the Lian Li Galahad II Trinity Performance 360 is one of the best AIOs on the market. What do you guys think of it?

 

I'm using the Lian Li Galahad II Trinity Performance 360 right now in my test rig to replace my old faithful EVGA CLC 360 since getting another Socket 1700 mounting bracket would be $35-40 so I just decided to grab what has been rated the best for awhile now in AIOs. While I acknowledge I am using good silicon with this SP109 14900KS, it is handling everything I throw at it up to 5.9 all core. It is pretty beastly and built like a tank.

 

I've never used an AIO with a mounting block so thick and dense before. It also has a very high flow rate if you desire (which I do, so I have the pump running full tilt).

 

You can use Lian Li's software for more exacting controls via USB connection, but it also lets you hook it straight to the MB for direct control which is how i'm controlling it on auto.

 

It is a lot noisier under load than my Arctic Cooler Liquid Freezer II 420 in my main rig so if noise isn't your jam, you might want to look elsewhere or just replace the fans.

 

It is also dirt cheap atm on Amazon for $110 (free shipping if Prime member). All models shipping now are late revision V2 so the pump issues are a thing of the past (in theory).

 

 

Currently in my test case....

[tps3443]

18 hours ago, electrosoft said:

SP109 14900KS + TG 8200 + Asrock Z790i update:

------------------------------------------------------------------

 

So I've been testing UE5 shader compiles with Fortnite (clear cache, relaunch, repeat) with the 14900KS and whew, watching this suck down 300w+ and all cores ablaze on launch for a bit is something I've never seen in a game before but I never played some of the major heavy hitters like this before either. I don't remember CP2077 doing this (but that was on my 12900k). My 7900XTX will not fit in my test case since the 360mm AIO is mounted to the front so I slapped in my 3060ti from my backup 12400 system for testing. When I switch over to the Nucleus 240mm for phase 2 testing, I'll slap in the 7900XTX and swap back and forth with the 4090.

 

I think I just need to pick up an open bench and keep it for testing hardware. With modern hardware, this Corsair 540 is showing its volume age.

 

I haven't really used the Epic Game store and it is very unpolished with issues compared to Steam. Not a fan.

 

As for the Asrock Z790i Lightning, I like it, but I do have some pet peeves with it and little bugs that are annoying.

---

#1. The CMOS battery is glued to the side of the rear I/O with a 2 position connector wire that plugs in that is just screaming it will break eventually as everything you have to yank the wire to do a hard reset.

 

#2.  There is no IA VROUT MAX control like on Asus and Gigabyte.

 

#3. Ring will cap at 4.5ghz when selecting all core no matter what. You have to go in and select per core and set your all core that way for it to boost to 5ghz

 

#4. Whenever you change your fixed Vcore, it will automatically switch back to LLC1

---

 

 

Right now, I have 5 profiles setup for testing/use

 

56x 45x Auto Ring 8200 Fixed 1.28

57x 45x Auto Ring 8200 Fixed 1.33

58x 45x Auto Ring 8200 Fixed 1.37

59x 45x Auto Ring 8200 (work in progress) But right now looking like 1.41 atm

Auto freq w/ Enforced Intel limits with adjusted AC_LL/DC_LL and LL/Vcomp set to min 8200

 

Until the Intel MC update, I'm sticking with my fixed Vcore profiles. After we find out their official cap, I'll start dialing in the Auto freq profile to fully test 6.2 boosting and various games for data collection. I know I boost only to 1.392v max 6.2 auto CB23 but it isn't worth taking a chance just to be safe till then also knowing I prefer fixed overall anyhow.

 

All are working on AIO for my use cases.

 

All do not thermal throttle and range anywhere from ~68-70c (5.6x fixed along with Enforced) up to ~87c (5.9 fixed) with CB23 loop testing.

 

I still need to test 8400. Trying to boot 8800 = hard lock up requiring a physical power reset which means it is the SA Bug kicking in. 8600 boots but as soon as I run TM5, SA Bug lock up. This shows the limitations (if you can call them that for my meager setup) but also shows really good potential for the TG sticks and this Asrock Lightning Z790i board. I'd love to test a known, good chip up to 8800+ just to see.

 

Oh, and since I was testing shader compiles in Fortnight, I decided to try and play (I've never played before) and my nephew (who is a die hard player along with COD) laughed so hard at me failing over and over again in Battle Royale as I was getting obliterated left and right while trying to convince him back in the day I was an upper level Quake player who routinely pwned...
🙂

 

I mean just destroyed.....

 

😁

 

DDR5 8800 wants that SA juice for most chips. This is with 1.360V SA below, I am messing around with 8800 as of now, just a quick test in TM5, never tried 8800 before. Good news is, even loose 8800c40 does perform decently though (Unless the IMC is under so much strain, it won’t perform very well) Testing with 1.500V vdd, and 1.500v vddq for the moment. I would see if you can manually enter your "DqVrefUP" and "DqVrefDOWN these slopes. Sometimes they do not work properly when training and going in to the 8800+ territory. I am not sure what your current motherboard sets them to that you are using. But it should be something like this below for frequencies on an APEX. 

8000 DqVrefUP 170 / DqVrefDOWN 88.
8400 DqVrefUP 178 / DqVrefDOWN 96.
8800 DqVrefUP 186 / DqVrefDOWN 104.
9400 DqVrefUP 198 / DqVrefDOWN 116.

[Talon]

8 hours ago, Mr. Fox said:

all of the negative publicity is going to be harmful to Intel and it could ultimately be more harmful to overclocking enthusiasts if they start locking things down and interfering with user tuning controls.

 

BINGO. My thoughts exactly. I would not be surprised at all if 12-14th gen are the last true overclocking platforms with all controls and knobs left unlocked. Intel higher ups are going to come down hard on all departments, and I really believe we might see overclocking heavily limited going forward, similar to AMD. 

 

Of note, https://www.tomshardware.com/pc-components/cpus/puget-says-its-intel-chips-failures-are-lower-than-ryzen-failures-retailer-releases-failure-rate-data-cites-conservative-power-settings

 

I find it highly interesting that this very well known and respected PC builder is actually finding AMD Ryzen 7000 fails quite a bit as well. 

[Mr. Fox]

2 hours ago, Clamibot said:

Does anyone have any tips for reviving a thermoelectric element or is it completely toast?

I have never used a TEC before, but they seem simple enough that fixing it should not be too difficult if you can confirm what went wrong. If I had to guess it would be the TEC plate itself probably gave up the ghost. Unless there is something proprietary about it, you might be able to replace it with a generic module. I would definitely look into fixing it. Please keep us posted and let us know if you figure it out. Sorry that this happened.

[Clamibot]

1 minute ago, Mr. Fox said:

I have never used a TEC before, but they seem simple enough that fixing it should not be too difficult if you can confirm what went wrong. If I had to guess it would be the TEC plate itself probably gave up the ghost. Unless there is something proprietary about it, you might be able to replace it with a generic module.

 

I've been toying with the idea of upgrading the peltier in this cooler with a more powerful one just to see how hard I could push my 10900K. I've been wanting to top your 5.6 GHz all core achievement on it. I suppose this situation gives me an excuse to do that experiment.

 

In any case, I think I'll go ahead and get the Li Galahad II Trinity Performance 360 as @electrosoft recommended so I can at least have a high performing cooler to fall back on should my experiments not be fruitful. I trust your guys' recommendations a lot more than any other group. I just needed confirmation that someone here thought the cooler was excellent, so thank you guys for that!

[Mr. Fox]

3 minutes ago, Clamibot said:

I trust your guys' recommendations a lot more than any other group. I just needed confirmation that someone here thought the cooler was excellent, so thank you guys for that!

Yup, I agree with that. And, if Brother @electrosoft says it is good, you can generally take that to the bank.

[tps3443]

17 minutes ago, Clamibot said:

 

I've been toying with the idea of upgrading the peltier in this cooler with a more powerful one just to see how hard I could push my 10900K. I've been wanting to top your 5.6 GHz all core achievement on it. I suppose this situation gives me an excuse to do that experiment.

 

In any case, I think I'll go ahead and get the Li Galahad II Trinity Performance 360 as @electrosoft recommended so I can at least have a high performing cooler to fall back on should my experiments not be fruitful. I trust your guys' recommendations a lot more than any other group. I just needed confirmation that someone here thought the cooler was excellent, so thank you guys for that!

5.6Ghz with a 10900K? That has to be an amazing chip. 

[Mr. Fox]

1 hour ago, tps3443 said:

5.6Ghz with a 10900K? That has to be an amazing chip. 

If it is the one I sold him it is a phenomenal sample.

[electrosoft]

56 minutes ago, Talon said:

 

BINGO. My thoughts exactly. I would not be surprised at all if 12-14th gen are the last true overclocking platforms with all controls and knobs left unlocked. Intel higher ups are going to come down hard on all departments, and I really believe we might see overclocking heavily limited going forward, similar to AMD. 

 

Of note, https://www.tomshardware.com/pc-components/cpus/puget-says-its-intel-chips-failures-are-lower-than-ryzen-failures-retailer-releases-failure-rate-data-cites-conservative-power-settings

 

I find it highly interesting that this very well known and respected PC builder is actually finding AMD Ryzen 7000 fails quite a bit as well. 

 

Agreed, as soon as this fiasco popped off I immediately had visions of lockdowns like laptop BIOSes coming next for desktops and I definitely do not want that.

 

If Intel (and AMD to a degree) want to maybe dial down the e-peen contest a tad and not be so aggressive in their boosting algorithms, I'm absolutely fine with that as that would just mean more headroom for us to dial in at our own discretion. To me, this works the best if Intel and AMD want to be a bit more conservative out of the box but leave MB/BIOS options to the discretion of MB makers.

 

If Intel and AMD decide overclocking is no longer a worthy endeavor and the boost you get is the boost you get and recommend/enforce severe non-K like chips, that would be a sad, sad day.

 

In the end, Intel's turbo/boost algorithms and vid requests on 13th and 14th gen were kerfuffled resulting in ridiculous voltage that fit under en envelope of ~2  cores being able to boost to 6.2 and pulling insane amounts of voltage trashing the cores.

 

Don't punish enthusiasts for your mistakes.

 

3 hours ago, Clamibot said:

 

I've been toying with the idea of upgrading the peltier in this cooler with a more powerful one just to see how hard I could push my 10900K. I've been wanting to top your 5.6 GHz all core achievement on it. I suppose this situation gives me an excuse to do that experiment.

 

In any case, I think I'll go ahead and get the Li Galahad II Trinity Performance 360 as @electrosoft recommended so I can at least have a high performing cooler to fall back on should my experiments not be fruitful. I trust your guys' recommendations a lot more than any other group. I just needed confirmation that someone here thought the cooler was excellent, so thank you guys for that!

 

It is good and especially at that price point.

 

5 hours ago, tps3443 said:

DDR5 8800 wants that SA juice for most chips. This is with 1.360V SA below, I am messing around with 8800 as of now, just a quick test in TM5, never tried 8800 before. Good news is, even loose 8800c40 does perform decently though (Unless the IMC is under so much strain, it won’t perform very well) Testing with 1.500V vdd, and 1.500v vddq for the moment. I would see if you can manually enter your "DqVrefUP" and "DqVrefDOWN these slopes. Sometimes they do not work properly when training and going in to the 8800+ territory. I am not sure what your current motherboard sets them to that you are using. But it should be something like this below for frequencies on an APEX. 

8000 DqVrefUP 170 / DqVrefDOWN 88.
8400 DqVrefUP 178 / DqVrefDOWN 96.
8800 DqVrefUP 186 / DqVrefDOWN 104.
9400 DqVrefUP 198 / DqVrefDOWN 116.

 

Yeah I'm going to be held back because of SA limitations (1.18 max solid, 1.19 benching quasi solid, 1.20 insta lock up) and SFF considerations and heat. 1.36 SA is a pipe dream for me. 🙂 Luckily, the chip is firing on all cylinders everywhere else that I need it. 8000 was the goal. 8200 is icing on the cake for this build. 🙂

 

 

 

[Clamibot]

6 hours ago, tps3443 said:

5.6Ghz with a 10900K? That has to be an amazing chip. 

 

Indeed, it's the super 10900K I acquired from @Mr. Fox. It's a beast of a chip. This is the same CPU I ran at 5.4 GHz all core in my X170SM-G when first taking it for a spin and am now running in my desktop. My super fast, low latency memory in my desktop was able to give a significant boost to my framerates in games despite using the same CPU, hence why that chip got moved to my desktop, to take full advantage of its capabilities.

 

I remember some dude on either overclock.net or overclockers.com that had a mythical grade 10900K that could do 5.7 GHz all core on air cooling. It sold for about 3 grand. An absolute top tier chip.

[Rage Set]

13 hours ago, Talon said:

 

BINGO. My thoughts exactly. I would not be surprised at all if 12-14th gen are the last true overclocking platforms with all controls and knobs left unlocked. Intel higher ups are going to come down hard on all departments, and I really believe we might see overclocking heavily limited going forward, similar to AMD. 

 

Of note, https://www.tomshardware.com/pc-components/cpus/puget-says-its-intel-chips-failures-are-lower-than-ryzen-failures-retailer-releases-failure-rate-data-cites-conservative-power-settings

 

I find it highly interesting that this very well known and respected PC builder is actually finding AMD Ryzen 7000 fails quite a bit as well. 

 

The key takeaway from Puget's data is how conservative they set their BIOS settings for both Intel and AMD. I.E. Puget sells workstation PCs where longterm stability is the selling point, performance is likely second or third. 

 

AMD's 7000 series is a furnace with default settings just like Intel's initial "let the mobo makers choose whatever they want" settings. 

 

I stated in the past few years that Intel was heading towards catastrophe. This was coming from me at a time when I was a proud self-proclaimed Intel fanboy. Admittedly, it was due to my now defunct company and having some direct contact with Intel. I saw the writing on the wall when the reps we were dealing with were changing every 4 to 6 months. Sure TO is high in consumer CS roles, but B2B reps generally stick around for years until promoted.

 

Intel will make it through this but they will be vastly different than they are now. I won't be surprised when they go fabless just like AMD. 

[Talon]

1 hour ago, Rage Set said:

The key takeaway from Puget's data is how conservative they set their BIOS settings for both Intel and AMD. I.E. Puget sells workstation PCs where longterm stability is the selling point, performance is likely second or third. 

 

Puget claims that with their reasonable and ideal settings they lose at most 1-2%. That isn't exactly something worth worrying over. But of course reasonable and competitive power doesn't make headlines like "Intel uses 400000w of power!!!!" *insert dog in house on fire picture*. 

 

Our approach has always led us to be conservative with our power settings, especially when we have shown that the real-world performance impact to be a small 1-2% range.

 

I personally don't think Intel will go fabless, their stated goal and the money they are burning is putting in the entire opposite direction. They are spending 10s of billions on their new fabs and new fab tech. I think they see the writing on the wall for the x86 CPU. They are pivoting to becoming an fab only company honestly. I would not be surprised at all if in the future, Intel could be producing ARM CPUs and GPUs for Nvidia. 

 

[Mr. Fox]

11 hours ago, electrosoft said:

If Intel and AMD decide overclocking is no longer a worthy endeavor and the boost you get is the boost you get and recommend/enforce severe non-K like chips, that would be a sad, sad day.

Indeed. It would mark the end of my interest in computers. I would own what I needed for work and spend as little as possible on that. Ultimately, I would no longer contribute anything on YouTube, or in technology-based social media, (which already is the only form of social media I actively participate in,) and probably abandon what little interest I left have in gaming. (The lack of interest has more to do with the repulsive selection of games than the act of gaming.) Literally everything I do for recreational enjoyment is either directly or loosely associated with overclocking in one form or another. It would become a totally pointless endeavor for me.

11 hours ago, electrosoft said:

Yeah I'm going to be held back because of SA limitations (1.18 max solid, 1.19 benching quasi solid, 1.20 insta lock up) and SFF considerations and heat. 1.36 SA is a pipe dream for me. 🙂 Luckily, the chip is firing on all cylinders everywhere else that I need it. 8000 was the goal. 8200 is icing on the cake for this build. 🙂

While it is certainly fun to set memory clock and AIDA64 read benchmark records and gather a few points on HWBOT, once that notch is added to your belt there is virtually no benefit that I can identify moving from 8200 to 8400 to 8600. Above 8000 it is hard to measure a difference in anything other than a memory benchmark. Other than maybe Geekbench, the vast majority of benchmarks and games and productivity workloads show virtually no quantifiable or meaningful change above 8000.

Running Y-cruncher and Prime95 and memory stress tests for hours (or days for some people) is actually more idiotic and harmful than the current stuff everyone is chirping about with Intel degradation. That is not much different than believe being a wife-beater makes your marriage stronger. Super stupid.

[electrosoft]

14900KS 5.9 all core testing, 8200 CL36 3060ti (only card that fits atm). In the SFF, it will be either 5.7 or 5.8 all core but i'm leaning towards 5.7 as 200mhz isn't going to make or break this build but does add a disproportionate amount of power draw and increased heat for minimal gains.

 

Crazy how on all my Intel BGA laptops this will spike to 100c on both 13th gen and Ultra 7/9 (I have them all) and even on my NH55 w/ the 12900k dialed in as much as possible it will still hit the 80's at much lower frequencies but desktop on AIO? 5.9 no problem once tuned.....this 14900ks is going to rock in the SFF. Modern laptop heatsinks are bleh as always.

 

Doing 1080p testing for the SFF and either the 18" or 23" display I'll be taking with it. This is at low settings to show you how WoW and cities just go to town on CPUs.  Don't let the fps fool ya. Once I switch it to ultra settings and turn RT on, fps plummet and the lows get real nasty (36ish) on the 3060ti.

 

I'm tempted to give a 4070 Super a whirl. I feel like that would be the sweet spot but I want to test the 7900xtx and 4090 first to kind of gauge.

 

Total system draw according to kill o watt meter is ~336w while playing.

 

This is why we call Valdrakken the CPU slayer with all the player data. Out in the open world, 500+ fps GPU Utilization almost always 91-99%. As soon as you come into "hubs" in WoW where calculations shift decidedly to the CPU, fps tank and GPU load plummets to sub 60%.

 

Valdrakken:

 

 

"Not" Valdrakken:

[tps3443]

10 hours ago, Clamibot said:

 

Indeed, it's the super 10900K I acquired from @Mr. Fox. It's a beast of a chip. This is the same CPU I ran at 5.4 GHz all core in my X170SM-G when first taking it for a spin and am now running in my desktop. My super fast, low latency memory in my desktop was able to give a significant boost to my framerates in games despite using the same CPU, hence why that chip got moved to my desktop, to take full advantage of its capabilities.

 

I remember some dude on either overclock.net or overclockers.com that had a mythical grade 10900K that could do 5.7 GHz all core on air cooling. It sold for about 3 grand. An absolute top tier chip.

Thats awesome! The LTT Golden Sample 10900K’s could all do the 5.7Ghz I think too. Those were all SP115-SP130 range. Just insane chips. but pretty hard to find as they only made #200 I think.

[electrosoft]

1 hour ago, tps3443 said:

Thats awesome! The LTT Golden Sample 10900K’s could all do the 5.7Ghz I think too. Those were all SP115-SP130 range. Just insane chips. but pretty hard to find as they only made #200 I think.

 

I had a golden sample LTX 10900k and it was good but not as good as the Silicon Lottery SP101 10900k I picked up used. That one was a monster. It performed better still lidded than the delidded other ones. It was crazy good.

 

I did a comparison when binning it along with a few other trash sample and 1x "binned" from ZtecPC 10900k's here: