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

It's a BGA chip with some features disabled compared to the 11980HK. ( AVX512 and PCI-E 3.0 x16 instead of PCI-E 4.0 ). In the past, those aliexpress shops have modified the 8700B ( stacked pcb ) to work in a Z370 board. There's even 8 core BGA Coffeelake variants modified for the Z390 boards. Although those modifications appear to lower the memory capability. It could just be incompatible by design? Hopefully they're only soldered to 1DPC boards for the shorter traces.

 

Videocardz says mobile:

https://videocardz.com/newz/intel-q...r-lake-desktop-cpus-including-core-i9-11900kb

So it's desktop now?

 

I personally think the 11900KB is just a well-binned Tiger Lake H45 CPU that is intended for use in SFF desktops. A good example would be my mothers' iMac which has an i7 5775R, which is a BGA CPU, but in a "desktop" class machine.

 

https://ark.intel.com/content/www/u...900kb-processor-24m-cache-up-to-4-90-ghz.html

 

This makes the most sense.

 

Tigerlake-H BGA trash is Gear 1 - 2933 MHz and Gear 2 - 3200MHz max, on top it's a mobile chip. I only remember Clevo laptops having XMP for 3000MHz DDR4 on a Skylake CPU design and that new Area51M disaster which has some 3000MHz RAM which has to be certified by Dell or some crap, they were Skylake which has a better IMC than Rocket Lake. So not sure how that will fare.

Also Intel really fked up hard in the naming scheme and SKU id numbers. I believe their Internal teams had a ton of prototype designs for every single uArch trying to cook something fast due to AMD Zen 3. So as per timeline of Intel CPUs and their uArch. Cypress Cove (last 14nm++ design, technically a Backport from 10nm - Rocket Lake) twin of Sunny Cove server design which is 10nm Icelake, and then we have that new Willow Cove design of Tigerlake BGA, next one is Golden Cove based design which is Sapphire Rapids server CPUs, the same one is Gen 5.0 and DDR5, new socket over Icelake. Also Alderlake uses same Golden Cove I guess.

Finally the Rocket Lake IMC Gear 1 maxing out at 4000Mhz is a diamond unicorn level chip I guess, then the 3866 which is hard, later 3733 which is also not easy I guess. What a nasty mess.

Then with the new DDR5 this whole Gear system is going to stick with us. So Intel really lost the IMC race with 10900K being last, such a shame as their CPUs had always top IMC. Now they are copying everything from AMD, be it the Gear system vs IF clock, Chiplet design (Intel did it way back during Core 2 era anyways but mocked AMD of glue lol), now out of all cards they are pushing the mobile crap cores onto the Desktops to maintain single CPU wafer production & have that SMT/HT performance, because with those small cores they are are going to market the hell out of them as "Ultra low Power" "Dynamic switching performance" "Super thin and light" because Apple ARM and AMD's BGA APUs. They are probably out of all tricks and have no architecture base design solid foundation, add the IDM 2.0 strat, where they are probably going to take help from TSMC or IBM to make their 7nm designs.

About that 11900KB is that a joke from Intel, I bet some idiot goofed it up ? On top it's even worse than Tigerlake H since, no PCIe 4.0 despite having same socket as TGL FCBGA1787. That is not a Rocket Lake class part at all for sure. I dunno why TPU and Videocardz reported it as a desktop or even RKL equal part it's inferior to all damn parts out there lol,

 

Tiger Lake H runs perfectly fine at 3200Mhz Gear 1 on my 11800H. BUT out of box is gear 2 by default. Of course MSI still includes the hidden/advanced bios with the key entry and you can simply change from gear 2 to gear 1 which is one of the first things I did after getting my GP66 laptop.

Out of box latency was around 93ns. Gear 1 brought that down to about 73ns. Gear 1 and tuned memory brought me down to around 63ns if I remember but will have to double check. Either way a huge improvement in latency and leads to some nice gains in CPU limited games like e-sport titles at super high fps.

The included memory is kinda garbage as expect so when I swap in a decent 3200Mhz kit, I expect to get down around 50ns.

https://imgur.com/a/yOXtvHl

 

Update 05/29/2021 7:22 am
An Intel statement on the models is now also available.

Intel has partnered with customers interested in expanding their product portfolio with enthusiast, small form-factor desktop designs. The Intel Core i9-11900KB processor is a BGA solution built with unique specifications and performance specifically for these designs.

https://www.computerbase.de/2021-05...ake-mit-65-watt-fuer-den-desktop-aufgetaucht/

 

Looks like @Tenoroon called that one spot on.

 

While Intel release new BGA junk in desktop shape, AMD prepere for a new Threadripper to try come in front of Intel's coming HEDT chips.

AMD Ryzen Threadripper “Zen3” now rumored to be available in September vidocardz.de

 

Coming soon... Available in limited quantities only from scalpers for at least double MSRP.

Does it overclock worth a damn? TBD... probably not.

 

At least it perform better than the BGA Junk Intel released yesterday as desktop chips

Chip shortage forcing AMD to prioritize flagship CPUs over cheap Junk

Back to Intel and their new move......

Intel's "secret" Core i9-11900KB uses a different architecture on the 10nm node

WTF?! At some point, Intel secretly released the Core i9-11900KB, i7-11700B, i5-11500B, and the i3-11100B. The B stands for badass because that’s what these processors are: they use the 10nm SuperFin node, the superior Willow Cove core architecture, and are clocked higher than their regular counterparts...

But let’s circle back, these are Tiger Lake CPUs. These aren’t Frankenstein prototypes. They’re overclocked mobile processors.

 

The person that wrote that needs to stop drinking the Kool-Aid and pull their head out. I corrected his mistake. Castrated TDP soldered to mobo is going to suck no matter how you slice it.

 

Ok so, first of all I am BLOWN AWAY by how low my temperatures are on a 18/36 cpu at 4,800mhz all cores with just (3) 360’s and a D5 lol.

Although, I am not quite out of the woods yet. After swapping to LM I have double checked, and triple check the fitment of my waterblock on to my direct die mount.

Now,

(16) cores are literally within very tight tolerances of each other. superrr low temperatures. Like I just ran R15 a few times and this is the max lol.

I ran R15 twice to see how they do at 4.8Ghz on all cores.

My temperatures are amazing. But these same two cores have been a thorn for sometime now.. (It is always these same two cores. On the last block, and now on this block) (Temps never this high though)

I feel like contamination could be on my silicon in one single specific area that hurts the temperatures of these cores.

I can see some tiny black staining on the silicon, and that hazy look too. Alcohol simply doesn’t take it off.

Look at these temps and tell me what you guys think? One single core hits 96C while the rest sit in the 50’s and low 60’s. Makes no sense.

Should I lightly lap my die? Well, not even lap it. But more or so just to remove anything that may be on there that alcohol isn’t getting off.

I made sure I coated very well with LM. No bare spots on the die, copper cold plate was coated well too. All soaked in very nice, with no dry areas that wipe the LM away.

I figure you guys are probably gonna look at that one super hot core and be like WTH? lol.

I haven’t ever seen anything so good and so bad at the same time like this before. very weird.

 

I would not lap the die because you could end up with worse contact considering the die is just barely taller than the top of the die frame. The only times I have seen that problem (and I did encounter it exactly as you have now with my 7960X and 7908XE) is when the liquid metal is not sticking correctly to both surfaces and you end up with a "hole" in the TIM. I would thoroughly polish the die and water block with something like Flitz polish or a baking soda paste and then do a final clean with acetone. Then I would use the scuff pad that comes with liquid metal or a fine sandpaper like 1600 or 2000 grit just enough to knock the shine off of the contact surfaces (don't sand them down, only scuff the surface) to help the liquid metal have a better surface to bond to before a final alcohol cleaning and applying the liquid metal to both mating surfaces.

 

The 3733/4266 numbers are from the Icelake-U ( Microsoft Surface Book 3 with 32GB 3733mhz ) / Tigerlake-U's specs.
https://www.igorslab.de/wp-content/uploads/2021/03/11700k_gears_specs.png
I think those might be the defaults? Rocket Lake had similar listings of 3200/2933 for the Jedec specifications. https://www.igorslab.de/wp-content/uploads/2021/03/11700k_gears_table-980x550.png
It can't go higher than 31x something and 28x seems to be the average for 11900K with the upper limits being ~29x+.

The 10900K is a lot better than the 6700K for ram in a desktop but only slightly better in a laptop?

Hopefully SPR-X at least releases around the same time instead of 1-2 years after the consumer ( ADL-S ) lineup unlike Skylake-S ( 2015 ) and Skylake-X ( 2017 ).
It does come with a dual ring design compared to RKL/ICL so it should be slightly better on the memory controller.

 

I finally got it right guys!!!


I reduced that old hot core from 96C to 60C. So that’s a 36C reduction for that particular core. And I was able to reduce the other core from 72C down to 66C.

So my hottest core is 66C. I am good with that!

This is at 4.8Ghz with adaptive voltage. System is super stable.

I removed the waterblock, and rubbed the silicon to death with my LM Q-Tip lol haha. Of course the LM gets a little chunky looking after rubbing so much, so I wiped away the chunks with a new Q-tip, and added a little extra LM.

I am ready for 5Ghz! (In the summer time)

 

5Ghz was easy guys!

I just typed in X50 multiplier, 1.200 adaptive voltage with a +135 offset.

5Ghz booted right up at 1.358V and it seems to pass right through R15 as many times as you wish!

Hottest core was 79C@5Ghz.

^ Something I have never been able to do. I feel as if I can actually reduce the voltage. This cpu runs so cool.

7980XE with all cores at 5Ghz lol. wow.

Now, I am not saying this is 100% stable or anything (too early to tell) lol. But my temps at 5Ghz are equivalent to my old setup at only only 4.7Ghz. And I had 4.8Ghz stable previously. So its actually quite possible.

Using the best CPU waterblock available, with direct die has really shown a tremendous temperature reduction.

 

I knew you could do it, bro. it's not hard, but with that many cores and threads everything has to be just right. When you get it all just right it's totally awesome. Nice job.

I think there are 2 reasons for it, neither of which have anything to do with the CPU. SO-DIMMs are generally inferior and laptop manufacturers do a really crappy job of building motherboards. They cut corners on components and don't know how to design them correctly. In all cases, their firmware is totally botched as well. Laptops are basically trash.

 

You are right though, there is a particular way about installing direct die, and installing the block too. But now I know, and I have full confidence in removing the entire system and even re-mounting it if I had to.

My current Windows install is 6+ months old and under performing. It is spitting out some pretty lowly benchmark numbers. Once I re-install a fresh Enterprise windows tonight or tomorrow. Then I will start getting some benchmark numbers up here.

breaking that 5K in R15 should be very doable now

 

If you want the best possible Cinebench R11.5, R15, R20 and wPrime scores you should go dual-boot with Windows 7. Scores are going to be lower with Windows 10.

 

Well, look at this. It's basically a near-identical knockoff of the Lian Li O11 Dynamic XL case for less than half the cost.



https://www.newegg.com/black-metall...-tower/p/N82E16811700008?Item=N82E16811700008

 

Locked down stable settings for 5.1 and 5.2 no AVX offsets on this 11900k
Bench settings / gaming settings vs. long run stress settings as always differed greatly (shocker).
The 5.1 to 5.2 jump is pretty vicious but I've read worse.

The sheer amount of power and juice this backport can suck down and heat generation is just insane.

Final settings:

Memory @ 3866 G1 15-15-15-32-332

5.1 @ 1.325 w/ 1.338 max pull. Max temp = 80.
5.2 @ 1.445 w/ 1.470 max pull Max temp = 96.
5.3 @ 1.515 w/ 1.530 max pull will pass but thermal throttles into uselessness (great for gaming though)

Going to test 5.3 and 5.4 @ auto w/ HT disabled next

Passed:

Realbench 2.56 stress test
Prime95 29.x small fft 1hr AVX disabled
CB23 30 min stress test
CB20 30 min stress test
CB15 manual loops


5.1:



5.2: (Note: CB23 shows minimum off because I accidentally closed the window and reopened it and reran it again once)

 

Looking good Brother @electrosoft

 

NVIDIA Announces GeForce RTX 3080 Ti ($1200) and RTX 3070 Ti ($600)
http://forum.notebookreview.com/thr...-whopping-margin.833406/page-43#post-11097823

 

Well, for the price I couldn't skip getting one just to see if it is any good. If it is almost identical and built as well, there are a couple of things about it that I like better than the Lian Li version. And, I guess Phanteks may have actually purchased MetallicGear. I have seen a couple of places where Phanteks is identified as the principal company. Whether that is accurate or not, I do not know.

I will share my thoughts once I gather them. Maybe this will be a good choice for your next build, Brother @Rage Set. Can't hate spending half as much (or less than half) for basically the same thing. It will be here on Thursday.

I also ordered some thick brushed aluminum plates to fabricate mounting brackets to attach the dual Photon D5 270 reservoirs to each side of the MO-RA 360. I will have only water blocks and flexible tubing and a few slow-spinning fans inside of the case.

In addition to the desert dust, I am going to have pet dander to deal with now, so it will be even more difficult to keep things clean with an open bench.

Say hello to Ace and Grace. We're feeding them milk with a syringe.

@Spartan@HIDevolution

Spoiler: At least it's not hot... yet...

...and, yeah... it's a dry heat, LOL.

 

Ok, fresh windows install. Running the usual XOC 1.07 bios.

System is running pretty strong. I am right near that 1 Point/1Mhz in R15.

I have dialed back to 4.9Ghz. Which is seriously stable. I could run this test all night long, and I would never pass 75C on my hottest core which uses the most voltage of all of my cores. 5Ghz is still possible I think, but you can go run a game for 4 hours, or run just about any benchmark, and out of no where the system will just fall over and lock up while opening paint or something stupid lol. So I think it needs some more CPU input voltage, or some additional voltages need to be tweaked to find perfect stability.

So 4.9Ghz is a good enough compromise for me as a daily stable frequency. Also, these temps are with really high VCCIO voltages which maintain a stable MESH frequency of 3,200Mhz. IF I reduce the MESH to 3,000Mhz and use the default VCCIO voltages. The CPU MAX temps are even lower. By roughly 3-4C.

 

I await your thoughts on the cases. They appear to be good replicas of the O11D but not the XL version. Based off the reviews I saw, we are going to have a bit of a clearance issue when adding an EaTX mobo and fans on the bottom.

Cute kittens, although, I always pictured you as a dog person. Keep that damn heat right where you are, lol.

 

I like dogs and cats best when they belong to someone else and, in general, I am not a pet owner type. The kitties are more for my wife and our grandkids. My personal preference is no pets in my house. They are cute, though and I will get some enjoyment out of them as long as they don't become a burden. I tend to favor cats over dogs because they are more independent and not as needy. If I were looking for companionship in a pet, then a dog would be better.

We still need to get them spayed, neutered, chipped, vaxed and declawed. We are desensitizing them to noise and water while they are tiny. Daily baths and running the vacuum cleaner, blender and food processor a lot are giving them nerves of steel.

I hope the case is not overly cramped. I hate that. The placement and horizontal orientation of the 24-pin and EPS cables on the Z490 Dark may be helpful. We shall see. I am also concerned about how well my big 1600W Supernova PSU will fit. It is difficult to tell looking at photos and videos of other people's builds.

 

Hmm, my analysis was not as complete as intended. I see it is much closer to the O11 (non-XL.) However, it is interesting to see where there are differences. It seems odd that the NEO Qube advertises triple 360 radiator support, whereas the O11 (non-XL) is more limited. Unless the specs are misreported (which is not uncommon), my massive PSU should be less of a concern than it would be with the O11D XL. It does claim to support E-ATX and looking at stand-off placement I do not expect an issue there. Nonetheless, it is just a little bit smaller than the XL... which is an undesirable trait. Where the smaller dimensions may introduce challenges will be interesting to discover. I would still view both the O11 and O11DXL as being too expensive for serious consideration (for me personally). I am hoping I will appreciate the product as much or more than I do the Corsair 5000D Airflow. While I do like the 5000D, it instantly reminded me of why I prefer an open bench over an enclosed case.

 

Are you able to change all the timings on that laptop too?

 

If the NEO Qube ends up being too cramped for my Z490 Dark and 1600W PSU, I will just swap the guts from the work computer and put the Z490 Dark setup in the 5000D. With no radiators inside of the chassis, the 5000D would provide more than ample space for the setup intended for the NEO Qube.

 

Why not grab a Corsair 10000D? It can handle (4) 480's up to a massive thickness. It may look kind of ridiculous on a desk.

 

Best Buy is taking it to the next level and going Microcenter model with the 3080ti:

 

Primarily because the price is totally absurd and I don't think there is a case anywhere worth what they are charging for it. The Lian Li O11 Dynamic XL is half as much, but I consider it to be overpriced. I'm just not into the idea of spending a ton of money on a box no matter how nice it is. It would need to have some extremely compelling to-die-for features before I would entertain the idea. I do like how massive the 1000D is, but I have questions about the intelligence of some of its design characteristics.

That's good. I am glad to see they are doing the in-store only and not allowing orders for stock that is not in-store. The only bad part of it is the "select stores" thing. If you do not have one nearby that is a select store, then it's a totally meaningless gesture and you're just out of luck if you want one. If they have limited quantities available, they should distribute what they have evenly among all stores. If they don't sell out within 8 or 10 days, then they can sell them from a store that did sell out and let the customer pay for shipping.

 

I found a retailer that was selling that Corsair 10000D for around $384 USD, if you’d entertain the idea, I can certainly find it for you.

Yeah it has some down sides to it. I am not a fan of the corsair software obviously. I just like the size of it lol!

Anyways, check this out. I finally got 5Ghz stable. My bios wasn’t sending more than 1.397 volts to my 7980XE. And this is why 5Ghz was falling over on me at random times.

I enabled “extreme voltage mode” in my XOC x299 dark bios. Now I am able to send proper voltage to all cores.

I made a video of the endeavor, these are some screenshots from the video. I am
blown away by the single threaded prowess of this nearly 4 year old processor haha.

 

Thank you very much. I appreciate the offer a lot. Even that is way more than I am interested in spending on a case. That's just too much money for something that serves such a simple purpose in my mind. I do consider a case to be a durable part that can span many generations though and if it is truly amazing I do make exceptions on prices being stupid because there is value in that durability. It's the kind of thing you don't replace unless it doesn't fit your needs or you just get tired of looking at it. Otherwise, it lasts forever.

But, it is due to be delivered tomorrow and unless the NEO Qube turns out to be a bust and doesn't suit my needs, I am all set now. I'd have to pay to return it unless there is something wrong with it. If it is a good product and I like it, not only will it be a durable part that spans many generations of hardware, it only set me back $99 + tax. So, can't hate that, LOL. I don't like things that are "cheap" (referring to quality) but I love good things that are inexpensive or priced proportionately low to the value you derive from them.

Dang, I am sorry you did not know about that, bro. I could have saved you a lot of headaches had I known you did not have that enabled already. The same applies on the Z490 Dark.

The same is also true for ASUS mobos. For ASUS you need to disable voltage monitoring in the BIOS page with the monitoring options, otherwise, it caps your voltage and limits overclocking. If you try setting the voltage higher it will throw errors about CPU overvoltage protection and halt the system during POST. You also need to disable the CPU fan monitoring if you are not using the CPU fan header on an ASUS mobo or it throws errors about the CPU fan not working during POST.

 

This is the first time that I was actually required to increase the CPU input voltage to compensate for my CPU overclock of 5Ghz. All X299 OC guides always say to increase it (No matter the CPU OC). But honestly it is not required at all. Well at 5Ghz I had to finally adjust it, I was getting phantom throttling to 4.9Ghz on a core or two every now and then. And my cpu temps are in the low 70’s.

I barely increased the CPU VIN voltage, and I am
now maintaining that 5Ghz on all 18 cores without phantom throttling anymore. This CPU is actually dialing in really nicely at 5GHZ. I have reduced the Vcore even more.

 

Very nice. Good work.

I ran the CPU input voltage for the 7960X and 7980XE between 1.900V and 2.100V, depending on how high I had the cores clocked. What do you have it set at now?

 

It is currently at only 1.835-1.850V during a heavy load.

The default input voltage is around 1.700V-1.735V. (Depending on load)

The Voltage input is supposed to be +0.400V higher than the average V core voltage of the CPU for optimal performance. (Based on what intel suggest)

Previously I have always usually ran around 4.7 to 4.8Ghz all core OC. With core voltages of around 1.225-1.280 so the default Voltage input of auto has always worked perfectly of only 1.7V. That’s well over +400MV higher than the core voltage.

I know that the lower the input voltage is, the less power and less heat the platform will consume. But if it’s too low, it’ll phantom throttle and spit out weird and random inconsistent R15 numbers.

Running 1.9V input is fine. But that’s around 80 watts of extra power just from default input voltage. The cpu input voltage can also instant kill a CPU. 2.1V seems high to me. And I think is usually for the LCC skylake X chips. The HCC Skylake X chips were supposed to stay under 1.9V input.

The input voltage is also a big misconception too. I have never had a issue running it on Auto until now. And I only had to adjust it a tiny amount.

 

Might be trading one of my 5700 XT's or 1070's for a RTX 3080 plus cash, also found a line on the Asus Strix 3090 for 2,000. Hopefully have them both by Friday. Then I am officially broke for a while lol

 

AMD confirmed that 3D V-Cache SRAM powered Ryzen processors are coming to AM4 platform. This is insanely huge move from AMD. Intel on the other hand abandoned their eDRAM 4980HQ and 5775C processors (well they still shove the eDRAM chips specifically into those BGA trashware Apple Macs) which are still solid today and compete so well in 2020.

When I decided to get Z590 due to the stupid instability issues with USB which are still showing up for people on even 1.2.0.3 AGESA BIOSes for even X470 boards not just X570 with Ryzen 5000 processors, AMD comes in and is making this exceptional move which nets them 15% gaming performance boost. I think even in RPCS3 Emulation this might help (not more than AVX512 RKL, but that is as hot as Sun at that point)? Also they do not even have Latency penalty it seems since it's vertically integrated and TSMC optimized for the Zen 3 uArch. On top B2 hardware revision for the Ryzen 5000 CPUs with X570S (nothing changed the chipset is still Matisse I/O die on same node, they are citing new AGESA code to optimize the TDP) which is the only hope that USB problems get addressed, I strongly believe it's a HW issue which AMD Is not admitting the fact that iss. A simple reddit search will show what I'm telling, now again I have to wait ? "what if AMD fixed these issues now"..

So ADL is DOA ? Sketchy x86 mobile crap shoved in Desktop parts with DDR5 guinea pig cost on top of Windows disasters (Scheduler problem + new Windows 10 UI change) with less cores than 10900K and Ryzen 5000, and V-Cache 15% gaming boost over Zen 3 parts, yep. DOA Lake confirmed.

 

Don't let the high VCCIN voltage scare you from try out whats possible http://forum.notebookreview.com/threads/the-throttlestop-guide.531329/page-866#post-10793297

 

Is that 1.397v at full load?

 

The writing was on the wall since they had already started doing online ordering based on regions recently where once your "region" sells out, that's it even if they show as available online. This is the next logical step.

At least EVGA tells you when their website is being inundated with a deluge of auto notifies.....

 

Wow, that's crazy. Shark frenzy, LOL.

 

I was wondering why NVIDIA didn't have any meaningful information on their web site or in press release suggesting why anyone would want a 3080 Ti versus an ordinary 3080.

 

Oh for sure. Yeah I have tested high voltage input before. It brings the system power consumption up really high. I just don't need it any higher.

I run 1.200V adaptive with a +145V.

The voltage ranges depending on specific core quality. The (18) cores on my 7980XE use between 1.355V-1.422V under a sustained full load.

5Ghz is dialed in nicely for daily use.

Also setting a fixed static voltage is difficult because it sets that voltage for all cores. When your gonna have 1 or 2 maybe even 3 cores that need more than that. So you would have to guess high to
maintain stability, and you’d be wasting heat and power on some cores due to that much voltage not being required.


Now I am running in to power shutdowns at 5Ghz. I am trying to run R23 at a straight 5Ghz with no AVX offsets, and it almost finishes the test when the entire PC just shuts off.

 

I wonder how people really thought this was gonna perform.

Nvidia politics is all.

RTX3080=330watts higher boost.
RTX3080Ti=350watts lower boost.

I think the TDP going from RTX2080 to RTX2080 Super was higher than 20 lol. And those were practically the same cards.

Now let’s see the difference when they give them a 100 watt TDP power difference like the 3070 to 3080 has. But then it would crash the 3090 market, nvidia politics controls all.

I’m surprised gamernexus didn’t mention this.

I mean, I would buy a 3080Ti in a heart beat for $1,200! I could only dream of someone robbing me and then throwing a 3080Ti Fe at me lol.

People are paying $2,400 for a 3080 FE. And we get a card for $1,200 that is better all around besides mining, and it’s a bad deal. Gamernexus gets buckets of RTX cards for free to play with. So I suppose I understand where he is coming from.

PS: I was at the best buy early. People camped all night long. The store I went to had 60+ cards in stock. I was not able to get one.

I have (2) 2080Ti’s. And I offered one of my 2080Ti’s and $1,200 cash to several lucky people who managed to get a 3080Ti FE (No takers) (If this doesn’t show how bad our market is) then I dunno what will lol.

 

You can change the timings yes.

0724 3967 entry

Got my notify earlier and it should be here Monday. Either selling off my KPE locally or giving this to my bro on a 3080 FTW3 and letting him sell his. I had $500 EVGA bucks so this thing is a steal at $1070 out the door after tax and shipping. Should easily wipe a $1499 3090 FE that isn't shunted.