*Official* NBR Desktop Overclocker's Lounge [laptop owners welcome, too]

Yes, and the thing about it is, nobody is even certain that the CPU Interconnect Error (event ID 19) is even a real problem. There is some indication that it is an error in Windows itself and other indications that it is the Realtek Audio USB device included in the UEFI that triggers it. And, oddly enough stability seems to be improved with the logging of that error event disabled. In fact, I have not experienced any USB dropouts in Windows since.

If I continue having the random USB dropouts at POST then I am going to try disabling the Realtek Audio in the BIOS and see if that stops the problem. If it does I will comment about it.

Disabling the event logging of the WHEA error also corrects a high DPC latency issue that the excessive error logging activity was causing.

People that believe error logging should never be blocked are just silly. Just because Windows believes it is an error doesn't actually make it one. For example, I enable SMBv1 on all systems I use on my home network because it works excellent for network access on drive shares on my home network. That being enabled on Windows 8.X and newer is logged as an event because it is considered a security risk. I also disable Windows Update on my systems and disable the updating of drivers through Windows Update. I know SMBv1 is enabled, Windows Update is disabled and driver updates are disabled. I did those things on purpose. I don't need Windows to create hundreds of event logs documenting it. It is stupid for me to not disable that event tracking because the information is totally worthless and logging it a bazillion times simply wastes system resources to accomplish nothing.

 

Maybe those WHEA errors that Windows was logging in Event Viewer were consuming even more resources than I thought.

https://hwbot.org/submission/4833018_ | https://www.3dmark.com/3dm11/14638922



https://hwbot.org/submission/4833016_ | https://www.3dmark.com/3dm11/14638931

 

https://hwbot.org/submission/4833054_ | https://www.3dmark.com/3dm/67083192

 

WOW. Amazing scores brother Fox. It appears you got your 5950X to stretch its legs a bit more and you managed to take out most of the 6900 XT's with your 3090. Again, amazing job bro.

 

Thank you brother @Rage Set it felt nice to be happy with the 5950X for a change, too. Sorry for the extra screenshots of the HOF, but I felt like a little kid for a few minutes, LOL.

Now the cynical conspiracy theorist inside of me wonders if the powers that be at AMD (hey that rhymes) are pulling an Nvidia by trying to keep some extra mojo tucked away for a rainy day when they can flip a switch in a driver or firmware update to make it run faster when the competition heats up. 5950X just loves that FCKL. Things just might get dangerous with an X570 Dark in the armory.

 

This is some good discovery and minimizing the performance impact and the USB instability (Also surprising to see PCIe add-in USB didn't fix the drops - Means that's the issue of PCIe, another known one again damn) of the X570 and Zen 3, but it's annoying to see the top end of the stack of the x86 processors from AMD display such behavior in the first place. Also on the DPC aspect, Anandtech's X570 DARK review got the highest DPC latency and lowest POST time of all the boards they tested. They ran a 3700X in it not even Ryzen 5000 series. A shame that they didn't even test any Memory OC or such. Also EVGA board is not the X570S series from their documentation or specs, it's the first X570 only. AMD and ASUS say it's just a firmware change that allows the Matisse PCH silicon to run at lower power improving it but I refuse to beleive how a simple firmware change enables that, if that's the case all X570 boards can remove the chipset fan entirely.

It's clearly looking like an I/O die issue on the CPU, it's made on 12nm GlobalFoundries. Some folks even have problems with Zen 2 Matisse but very few cases vs Zen 3 Vermeer, but not just limited to X570, it happens on B550 too which is an ASMedia PCH chipset unlike the X570(S) which is just a repurposed 12nm Global Foundries Matisse I/O die. It's probably due to the performance push that AMD did with very small ceiling left.

Note All EPYC series even including latest Milan based on Zen 3 doesn't use 12nm GlobalFoundries I/O die but rather 14nm GlobalFoundries I/O die, they do not have insane clocks on the CPU cores so probably went for that and eases the cost ? or some other reason dunno, but they have higher power consumption on the I/O die on a Server socket despite that they did not change since first EPYC. All this means the AM4 CPUs get different validation suite entirely, ofc a Server and Mainstream (not sure what's the I/O die on the HEDT side) will always be separate but this is me just assuming more things.

Also since you do not get Memory instabilities over 3800MHz DRAM / 1800 FCLK unlike others, it's again another inconsistent behavior observation on the Ryzen design.

Memory instability past 3600MHz with WHEA bonus on top of USB, PCIe dropping issues which are independent and dependent on each other at the same time and even happen for bone stock. AMD should man up and fix this really once and for all but it's way too hard to fix once it's past validation and in the market. The lack of documentation from AMD is another big bummer, Zen 3D might be the only way they can fix but it doesn't change the I/O die on the CPUs but however a new stepping for sure. Your Ryzen 5950X Zen 3 is B0 stepping, AMD said a new B2 is in works, since all folks already built a lot of machines and lack of GPUs means fewer new buyers on top of Intel ADL hype. It's hard to find this information...

 

I second this, excellent job. And these achievements were only made possible by continually searching out solutions and being proactive. Understandably, bug finding can get tiring and cause many to throw in the towel, but to he who puts in the work belong the spoils. You earned it!

Sent from my SM-G970U using Tapatalk

 

For your end goals, it makes complete sense to effectively face palm the whea issue when what you seek are overclocking results. Just about every red lining endeavor has risk/reward. Very nice scores indeed! I put nothing past any corporation when the sole focus is maximizing profits and shareholder value. That's the nature of the beast.

Reviews like that irritate me. A modern board running an older CPU amongst a bunch of other boards that have long since had their BIOSes updated. Instead of doing the groundwork and retesting with the newest versions of their BIOSes and running 5000 series CPUs and doing some real edge testing to differentiate the boards they go with the ole ho hum. Ma and Pa computer users aren't going to buy Dark boards (or similar high end models) for ho hum use. They need to differentiate their testing based on tiers and designed use.

It is usually when boards get into the hands of hardcore overclockers that you can start to separate the wheat from the chaff and you listen to voices you trust.

I like @Mr. Fox approach to basically turning off all safeties and error reporting to basically pursue pure results but the issues are still there when all is said and done. It is just that he doesn't care because that's outside the scope of what he is trying to achieve. You use every tool and advantage you can to get the best scores possible. I love it.

 

https://hwbot.org/submission/4833431_ | https://www.3dmark.com/3dm/67107321

https://hwbot.org/submission/4833443_ | https://www.3dmark.com/3dm/67107700

 

So my brother sold me his system for really cheap. And this is the exact reason that makes me seriously regret selling that Z490 Dark Kingpin I had!… I’m a doofus for doing that. (I wasn’t anticipating my brother selling me all of this stuff)

Yes, I love my new Evga Z590 Dark and my awesome 11900K as my main system.

However, I’m building the 10850K rig as the new direct die backup PC! And I’m unfamiliar with these Gigabyte bios. I really wish I had my old Z490 Dark Kp. This 10850K would easily do 5.2Ghz all core at 1.380-1390V range. And it could handle half cores at 5.3Ghz. (All on that Z490 Dark KP)

I delidded this 10850K earlier today. I just used standard liquid metal. I let the LM soak on the solder for a while, and it finally came off!

Also, the plastic thermal paste spatulas work well to remove most of the solder.

This was my first delid on a soldered CPU.

This 10850K is probably very capable of 5.3Ghz all core now that it’ll be direct die.

 

Windows 10 LTSC is still the best for gaming!

I already posted this in the Clevo X170SM-G owner's lounge, but I'm going to post it again here since I think it's more likely I can get a comprehensive answer from the people on this thread:

My X170SM-G is finally arriving on thursday and I'm thinking of buying a Rockitcool full copper IHS for the 10900K I bought. Will it fit in this system? I'm thinking it should but I'd like to confirm first.

Also, do I need to put anything in the order notes on their site?

 

Yes, it will fit just fine. zTecpc sold them that way and I tested them with the Rockitcool copper IHS. It doesn't make a huge difference, but every little bit helps. And, you can keep the stock IHS looking brand new if you are planning to use liquid metal. I recommend you follow the guidance from @Papusan and @jc_denton on the baking process so that the copper absorbs all of the liquid metal it can into its pores (both underside and top side) before you install it in your laptop. That will produce better results and make the liquid metal application durable and lasting. It also makes the application easier. It sticks to itself better than it sticks to new copper.

 

Thank you. This is good to know.

I wanted to be 100% sure it would work. I already spent a lot of money on this laptop and really don't want to spend any more.

I really hope I never have to get high end hardware again, but unfortunately that's all that would suit my needs this time around.

Just watch me get used to this and only wanting high end hardware in the future. Everything you had before becomes very unsatsfying when you experience something better, kinda like me jumping from a 60 Hz screen to a 120 Hz screen, and then to a 144 Hz screen.

Fortunately I should be able to easily afford high end hardware in the future. I'm going to be graduating with a Master's Degree in Software Engineering in December, and I hope to have a job lined up by then. I pretty much only spend money on food, computers, and video games. I don't have any other hobbies besides those two and programming.

 

AKA "The Electrosoft Template"

Toss in plenty of Sci-Fi/Fantasy/Anime and your journey to the dark side will be complete.

I'm much older than you but take comfort in the fact that template will remain static even as you get older.

 

Congratulations on the accomplishment. I hope that opens doors to an enjoyable and prosperous career for you.

I concur. That is how it played out for me as well. The Fantasy and Anime are not something I ever got into, but I will vouch for the fact that the sickness is permanent. If we are honest about it, I think we would all admit that we DO NOT want to be cured. We love our sickness. That's why the degradation we see within the industry is so disturbing and aggravating to some of us. We expect better.

 

I’m right behind you! I graduate next year. Looking forward to it.

 

I'm still in high school .

I'm aiming for Electrical Engineering or something to do with Thermodynamics. Hopefully I can do either one of those, but cheers to both of you!

 

Also, I've always had this question at the back of my mind because I've yet to find an answer (unless I'm just looking in the wrong places,) but why did AMD's Bulldozer and Piledriver based CPU's clock so high? I've always assumed it had to do with the ability to focus all resources and power to 1 module and that AMD traded IPC for clock speed, but the former doesn't sit right in my head.

Can anyone explain how they managed to clock so high?

 

I do not know if this is a technically accurate assessment as much as an assumption, but I suspect that they processed so few instructions per clock cycle (IPC) that it allowed much higher clock speeds. Maybe something like you can rev the engine on your car in neutral, but that high revving activity doesn't translate into power at the wheels because the engine isn't doing much. It might be charging the battery with the alternator and running the AC system, but it isn't really doing much.

 

This is a correct assessment.

Certain stages of the execution pipeline take more time than others. The more complex you make these stages, the lower the clock speed has to be, else you run into errors since the slowest stage hasn't finished yet. Unfortunately, you can't always split instructions into equal parts.

The limiting factor on clock speed these days is the maximum speed at which the slowest execution pipeline stage can run at. Since execution pipelines keep becoming deeper to introduce more instruction-level parallelism (which increases IPC) and keep becoming more complex, the clock speed must be lowered to prevent errors. However, smaller process nodes shorten the distance between transistors and shorten the wires inside the CPU, which negates this effect. This is why clock speeds have stagnated in recent years while the IPC has been going up. This is a more power efficient approach anyway, so I approve of this direction. Having said that, I do like being able to overclock the crap out of my CPU.

My guess is that AMD's Bulldozer and Piledriver CPUs had very shallow execution pipelines, which made it easy to keep the clock speed high. An ideal execution pipeline to keep the clock speed high consists of 4 steps: Instruction Fetch, Instruction Decode, Execute, and Write Back (Store). This is the simplest one I know of. Increasing the number of stages increases instruction-level parallelism, which in turn increases IPC.

 

I recommend always staying in school. As children we never realize that our parents actually do know best. Time goes by wayy too fast to veer off path for 10 years lol. Pick something you enjoy, or something you’re truly passionate about. (Don’t go to college and get a degree for a something that is just a phase your in) Ive always been interested in the optimization of games with PC hardware. Too many crappy performing games out there that can be drastically improved with the simplest of fixes.

That’s why I like overclocking so much. It’s free performance.

 

I say I’m still in high school sadly because I wish I was older and already doing what I’d like to be doing, which is working with technology on a higher level. I’m definitely going to college and aiming for degree in Electrical Engineering as this topic has interested me for quite a while, and is closely related to computers/technology, something I’ve yet to burn myself out of.

I also dislike not being able to talk technology with anyone. Well, I can, but it usually is a very simple conversation. All of the people I have ever talked tech with have a very basic understanding of it. If I wanted to have a surface level conversation about newer CPU’s and graphics cards, I know people in person that I can talk to, but for everything else, I’ve got all of you. It’s just a bit frustrating that I’ve yet to talk to someone who doesn’t just have surface level knowledge. (I apologize if it comes off like I’m bragging, I’m really not trying to.)

 

Here is a video on that ATi card mentioned in the other video linked before.

 

You know AMD could still do well in the unicorn market because there is none right now lol

Imagine having a dual gpu card today? Game/production on one and mine on the second. I know the mainstream consumer market would hate it but still take advantage anyways.

Could also do well for the internet cafes to offset the cost of an idle machine

Obviously I have my bias, but it has allowed me to "responsibly" diversify my income and make a larger dent into my student loans.

Plus I just want more unicorn cards to exist, really like them back in the day.

 

New"Game cashe" and no real performance improvements. This means socket AM4 is at the dead end for owners who hoped for a last ride on 6000/7000 series Cpu's. But a few extra FPS and maybe better binned chips (eat the last drop of the OC headroom) is probably better than nothing. Not the best news for the oc’ing nuts.

AMD ROADMAPZen 3 3D V-Cache and new notebook CPUs in early 2022

For the rough five-year anniversary of Ryzen, AMD has scattered more breadcrumbs over upcoming products.

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

Microsoft got it exactly the way they wanted. Push Nvidia infront with only DCH drivers and their Store apps down our throats... http://forum.notebookreview.com/threads/nvidia-geforce-driver-496-13-whql.836957/#post-11123177

 

So no 3d cache for am4? I don't have translator on my phone

 

Zen-3 = 3D Cashe. This is all you get of improvements. The 6000/7000 chips is a Stop-gap as the XT chips.

 

That's all that was really expected unless the typical AMD rumor mill was over hyping itself again?

I'm looking forward to the giant l3 cache and the longevity it will provide over time.

 

Maybe next time boys, maybe next time...

 

Couldn't agree more. I didn't go to college or at least get some certificates and I ended up doing intermediate tech support making $80k a year which isn't much to support a family on in Commifornia. More importantly I was stressed and unhappy and this obviously affected my personal and spiritual life immensely. I went back to school 2 years ago to get my Associates of Science Degree in Health Information Technology and I'm almost 40 now. I'll have to start at the bottom, at least as far as experience in the health care industry, and take a pretty substantial paycut initially, but after much prayer and discussion I felt it was more beneficial for my family that I do something I enjoyed more. It does sound so cliche but follow your heart, find something that enriches your life, and hopefully the lives of those around you, and you will find that the financial aspect will not be your primary motivation.

Schools done in about 6 months and then it's time for another big change, move out of CA. I won't digress on reasoning, most people here already know why. Looking at Texas, Idaho, Tennessee, Florida and a few other places who aren't trying to obliterate our constitutional rights on a daily basis.

Sent from my SM-G970U using Tapatalk

 

Oh right l4 cache, been so focused on everything else (probably for the bette).

Though if the size of l3 is giant then the need for l4 is reduced especially with ddr4 as it is. Though that also begs the question of what the relative speeds of l4 made today versus how they were in 2014

 

Ugh. This CPU is so stinking lame, LOL. Based on what I see online, it might even be a better than average silicon sample, which is pretty damned pathetic. It can't do 4.5GHz, and the memory clock seems like 3400 is the most its chintzy IMC can handle. I hope to not have to use it very long.

Anyhoo... 5950X is on the way to James at Rockit Cool for a delid. I hope it makes a measurable difference in overclock headroom.

 

LOL... the 9750H in my Tongfang turdbook overclocks higher and has a better Cinebench R20 score. All I can say is the Ryzen fanboys that had this CPU must have been drunk on red Kool-Aid if they stayed with Ryzen and upgraded to the next generation, because it would have been very disappointing to have a locked CPU in a turdbook tearing your desktop with an unlocked CPU a smelly new posterior orifice.

https://hwbot.org/submission/4834392_mr._fox_cinebench___r20_ryzen_5_2600x_3207_cb


https://hwbot.org/submission/4305083_mr._fox_cinebench___r20_core_i7_9750h_3368_cb

 

Even I was not really interested in the 2600x when it dropped. The 2700x for the right price or bust, luckily that happened.

Speaking of competition what are a few decent boards for x299? Probably nothing comes of it but I have been entertaining the idea of building a sister ITX system or mATX to go with my AM4 ITX system

 

It will do fine until I get the 5950X back. I'm sure it would be OK for work or gaming. I didn't expect it to be impressive, but it's got to be way better than the dual-core Athlon 3000G that I was going to use as a temporary CPU. I think the maximum temperature on the 2600x running Cinebench R20 was like 42°C under the monoblock, so about 50°C less than the 5950X.

I am only familiar with E-ATX X299 boards. But, perhaps Brother @RMSMajestic can chime in. He purchased my 7980XE and X299 Dark, but I think he may have put the CPU in a small form factor system. I'd try to stick with EVGA if possible. That would be my recommendation.

https://www.amazon.com/EVGA-Micro-Intel-Motherboard-121-SX-E296-KR/dp/B07J9KWHK7

 

Oh yeah, as a stop gap CPU I wouldnt complain. At the time it dropped I also didnt like the idea of 6c/12t especially when the 8700k had been in the wild already.

That EVGA board looks nice, going to have to research tomorrow and see if its bifurcation friendly. Right now my 5820x precision is the dumbest system I have ever used. It looks like the Precision towers can only do 1 CPU per 2 GPU's. So even if I have 5 PCIE slots I can only use 2 of them. Luckily it has the 9800x in it so it can still be fun but making sure it can serve the function I have in mind first.

 

What I find to be probably the most discouraging thing about being a PC enthusiast in 2021 is the almost non-existent interest or participation in overclocking and benching that exists among Linux users. At a time that Micro$lop is scraping the bottom of the digital cesspool with how truly pathetic and disgusting Windows has become and moving to Linux has never carried greater potential to be a more enticing alternative, the unfortunate reality is that it would be a truly mundane and unstimulating experience. Being already familiar with the general audience, I expected an unimpressive level of engagement, but this is truly pathetic and very deflating.

​

Keeping the big picture in perspective, this probably should be the last thing any of us should be concerned about. With our nation teetering on the verge of becoming a economically and morally bankrupt Totalitarian slum there are certainly bigger fish that need to be fried right now.

 

I have got some nostalgic vibes. After instalation od 3dmark11 I found top scores form my previous laptops. https://www.3dmark.com/compare/3dm1...3dm11/13499642/3dm11/13924549/3dm11/14643771#. From P170HM with modded 680M and coil whinning power supply, where sleep was bugged, through P157SM-A 8970M and custom vbioses with power mods. Later "BGA junk" aka Medion X7849 (Based on Clevo P670RS-G) where I first experimented with liquid metal. And finally big heavy guns. First contact with 1080 SLI and 6700K was wonderfull, and scary (due to fan noise while installing windows). Later some basic mods like changing thermal paste and CPU delliding with conductonaut. After a while I gained enough courage to install Dsanke bios, 9900KF and 64GB CJR/JJR based memory wich again gived me signifcant bump of performance. And that's it. I sold it with conviction that I build desktop... well not this time xD. First GPU prices were horendous, second I found bargain for new P870TM1-G with two year warranty. This time I spend only 2 months with orginal setup and 9900K P0, 64 GB of B-Die ram changed that. Again after bit of experimenting and tweaking changes were made form little beast to literally portable monstrum. Currently I'm running http://forum.notebookreview.com/thr...o-debloated-with-several-enhancements.835513/ "gutted" 2004 windows 10 Pro without a lot of bloatware and NVcleaninstall'ed 510.06 drivers.

 

It seems Nvidia finally understand that 10GB vram ain't future-proof for a 4K card, LOOL

Oh'well, better later than never. But I'm sure they saved a few $$$ cap the vram limit at 10GB for their newest gaming flagship. And it also help Nvidia a lot push people/gamers over on their next gen cards before they had to (when the miners is gone and if the GPU/chips shortage have eased up). Oversupply of chips won't help Nvidia and its partners to be equal greedy as with Ampere cards. With time forwards we will see Nvidia and their partners try hard appreciate all Gamers again. Miners will be gone.

NVIDIA preparing GeForce RTX 3080 with GA102-220 GPU and 12GB memory for January 2022 launch videocardz.com

"NVIDIA is clearly focusing on memory upgrades with Ampere’s mid-life refresh"

 

Two steps forward and six steps back. Increasing VRAM is a no-brainer that they should have done at launch. What is not overtly stated is what worries me. Nvidia and MS have a partnership. That partnership has eliminated the standard drivers; is crippling AMD CPU performance; requires new hardware to address perceived security concerns. And, the cherry on top, is the increase in prices.

 

The one and only thing that will increase linux support is AAA game titles being available for linux. From a business perspective companies will not widely adopt it because the support staff isn't there in the IT world. BUT gaming would bring light and large movement to the platform. Dell tried the linux computer sales a while back didnt last long. The corporate world and everyday business will not make the move. Retraining employees, major software support and migration etc.,

 

They could've game the cards the VRAM allocation they designed them for but redacted before launch, not really something to be praised imho.

It's really weird that Nvidia keeps pulling these kinds of tricks. The only thing I've learned is that even if I seriously considered Nvidia for my next gpu is don't buy on release or anywhere near release, because you are only getting 80% wait for the next 2 refresh and you may actually have the full product but oh wait they've redacted functions for particular scenarios.

Why should I have to navigate that maze when I know exactly what I get from the competition?

 

Yes, this is true. The business world is absolutely dominated by Windows, and for good reasons. The world of IT professionals has very limited knowledge of Linux and even if that were not true, the workforce wouldn't handle it well. You can't really expect employees to do their jobs effectively using something they are not familiar with and don't want to become familiarized with. A fairly significant number of people whose jobs require using a computer struggle with anything new and more than a few barely have enough computer smarts to use Windows. They use what they are trained to use and don't want to learn anything new that isn't necessary or essential for them to perform their job duties. Introducing change for the sake of change doesn't set well with them, and it's also not very smart from a business perspective to introduce change (or waste time and money) when doing so isn't essential to success. If you gave people like that a Linux machine they'd be turned inside out, if for no reason other than it would be new and different because they have a mental block when it comes to technology.

I bet (and hope) that the business world will be very slow to embrace Windows 11. Some are still using Windows 7 and a fairly large percentage of those that moved to Windows 10 did so begrudgingly, and probably won't be eager to squander any resources on making another switch when it would be totally pointless to do so.

The guy the wrote this article is very much on point. Totally pointless and an utter waste of time to entertain the idea. And, if you value performance and don't want to lose some for no legitimate reason, a very bad idea.

Windows 11: Microsoft's pointless update | ComputerWorld

 

As far as client facing IT I would agree but Linux basically runs the backend to my understanding. Linux SysAdmins make a very attractive paycheck. I have been debating dropping Windows as my "toolbox of choice" in favor of Linux

 

Like that article you liked to said, Microsoft tends to alternate between good and bad versions of their Windows operating system whenever they come out with a new release.

Windows 11 looks like a MacOS ripoff to me. It doesn't offer any better functionalities than Windows 10, so yes it definitely is a pointless upgrade. It's essentially Windows 10, but with a MacOS inspired GUI.

 

My largest client is a bank with 132 branches. The last few meetings we had with my IT staff like with past windows versions we will bot be moving to win 11 for at least 3 years. Fiserve is our banking software infrastructure backbone applications and all their apps still require iE11 and do jot work with edge or google chrome. Manny of my Dr’s office’s with medical software will not support windows 11 for at least 2 years.

 

Yes, in that one area Linux is common because it is affordable, (free if no support is needed,) involves a lot less nonsense, and it is more reliable. I suspect one of the reason those System Admins make so much money is because they are harder to find.

It's ugly. If I wanted ugly with limited options I would build a Hackintosh, or I'd go full Kool-Aid and waste money on an Apple computer that costs more and sucks worse. But, regardless whether you find it attractive or disgusting with respect to the UI, Windows 11 is a Windows 10 downgrade because it performs worse, not better. The average consumer isn't going to know it performs worse. Some will buy a new PC with Windows 11 pre-installed. It may offer better performance than an older machine it replaces and they will have nothing to compare to as a point of reference and not know that it runs worse than it would have with Windows 10. But, they will have a problem with features they are used to using being hidden, deprecated or moved to a more obscure and inconvenient location. Micro$lop has demonstrated time and again that they don't care what users like or want based upon their illogical and arbitrary decision-making process that results in their products becoming less functional, more inefficient, less convenient to use and harder to love.

I believe this is going to be the norm across all lines of business. There is no incentive whatsoever for any company to flush their money down the toilet on a buggy solution that will provide no benefit and will likely degrade the job satisfaction level of a large percentage of their employees. It is unfortunate that they continue to show us how incredibly stupid they are. They keep repeating their mistakes with respect to making major UI changes with no awareness of the fact that UI changes are not wanted by the vast majority of their user base. For the most part, the only people that embrace the UI changes are the fanboys that would never find fault with anything because they are fanboys that go with the flow because they believe whatever Micro$lop does is the right thing to do.

 

That is one aspect to be certain but they also have to have certain thresholds for Windows and MacOS given that they need to be able to function together pending the circumstance

 

https://www.tomshardware.com/news/leaker-allegedly-exposes-alder-lake-inner-workings

"Memory will function differently with Alder Lake, depending if it's DDR4 or DDR5. DDR4 can run on Gear 1, but the ceiling is at DDR4-3600. Increasing the uncore voltage could bump it up to DDR4-3866, though. By default, DDR5 will only run on Gear 2, due to the high data rate. In order to see any improvement over DDR4, DDR5 needs to be clocked over DDR5-8000. Gear 4 is present for future DDR5 memory modules that exceed DDR5-10000."

Hmmm, I don't know how I feel about this.

 

I know it's not a very popular position to take in today's world, but for me it's all about the results at the end of the day. If the crazy numbers actually produce crazy good results then sure, let's do it. If the numbers are just a number and the end result sucks or is mediocre, then they can put their fancy new tech where the sun don't shine. Results matter. Period. Offering something new and having good intentions is worthless without results. The other stuff looks good on paper for marketing and gets silly people excited simply because it's new, before they can even say if it's better or not.