BGA Venting Thread ;)

You can normally get Two gen of the Flagship CPU Depends if you buy early or late in the life cycle for the MB. Don't push a Dellienware on him

Edit. What happen if you ask in AW social media

 

Perhaps I ought to have been clearer: 'chasing performance without keeping value for money in perspective'. Of course, there are people who use their 1080 Tis, Threadrippers and Xeons for real work and productivity. There's nothing wrong with playing games, too. All work and no play makes Jack a dull boy.

However, objectively speaking, value for money is entirely measurable: performance ÷ price, or FLOPS/$ (or IOPS/$ for storage/CPUs). Maximising this amount is what I would aim for, and it typically falls in the middle of product stacks. Subjective value is a different story—some people aren't satisfied with 1080p 60 FPS, and need 4K 120 FPS, or need faster encodes. Sure, carry on and buy the higher end parts.

But parading the LGA socket as it is today as though it will future-proof the notebook in question (I am not talking about moving up within the same product stack) for significantly longer than an equivalent notebook with a BGA CPU, doesn't make sense. The upgrade path for the P750ZM is dead, save the GPU (and it maxes out at a GTX 1060, IIRC—not so great). The P870KM1 cannot take a Coffee Lake CPU. Both are less than a year old.

If you want to show off the features of a modular notebook, an Intel LGA CPU is not it.

MXM? Definitely.

 

Why would they bother doing that?

 

Ryzen clevo when.

 

Zigackly. Where is it, though?

 

I hope yoo know P870KM1 has the second revision of socket 1151 FYI... Exactly same LGA socket 1151 is used both for desktop and some laptops See also https://www.digitaltrends.com/computing/what-are-intels-lga-processor-sockets/

 

-p750zm with MSI Gtx1070 http://tieba.baidu.com/p/4889598899

eurocom mxm 3.0b gtx1070/1080 are coming

 

I fear asking or wondering. But I avoid places that do this, this level of content moderation doesn't help anyone, if I write in broken grammar, so be it, if I have a bit too much space between my lines, so be it. I really hate the format they force on us as well there, I need one empty row between my replies sometimes

 

Woah there, that's some serious bull$it, how in the world can they even tell that, and to whom, we're all enthusiasts, we know what should be upgradeable...

 

OEM motherboards have had that issue since the year dot.

 

The new Apple iPhone X is going for $1k (and to the Google Pixel 2 XL is close to that). More BGA abuse? Seems like an inversion - now we're buying phones at laptop prices....

Sent from my Pixel 2 XL using Tapatalk

 

Yes. That is both ridiculous and pathetic. I question the intelligence of people that spend as much on a disposable smartphone. Because they do, we are seeing equally pathetic changes in PC tech, with the offerings being dumbed down and most consumers lacking sufficient knowledge to even recognize what is happening. Posted this in another thread.

 

Manufacturers see well specced glued together unserviceable stuff in a locked ecosystem making crapple a shed ton of money, and want in.

Meanwhile attempts at modular smartphones and laptops have failed.

Capitalism has no sympathy for minorities until they are big enough to sustain a niche market

 

Exactly. As long as the majority is comprised of extra-stupid people that love wasting money on trash we will all have whatever pathetic filth they are willing to tolerate (or whatever garbage they are not smart enough to reject). Oh well... it's all gonna burn anyhow, right?

 

I think my Samsung S5 was $250 some three years ago. It's unfortunately unable now, but I don't see why it couldn't be replaced by another $250 phone...

Sent from my Pixel 2 XL using Tapatalk

 

This is what people in short time can buy... ARM-based Windows 10 notebook Jokebooks. And the $2653.25 U.S. dollars garbage from eg Micro$haft is what they can buy today. Pretty sad to watch for us who appreciate high performance. And Microsoft follow the bandwagon and customizes its latest desktop smartphone OS for today and coming throttlebooks. As you can see, M$ have already started... Win 10 Creator Fall update.

 

lmao at phone making battery also integrated and non-replaceable, whole thing is a joke. which is why i will never spend more than $100 for a phone. s4 here, i recall people running out of power and have to charge daily, watching videos etc. the moment i got it i disabled whole bunch of things and now only listen to music on it lasts like 4 days without needing to charge with several hrs of usage each day.

 

@Papusan @Mr. Fox
https://videocardz.com/74170/aida64-gets-early-support-for-mobile-intel-i9-series

HAHHAHAHAHHAHAHA i9 BGA HK HAHAHAHHAHAHHAHAHAH LOLLLLLL

 

Yeah, A brand new Core i9 for next gen notebooks Jokebooks need proper cooling

 

That looks like what I'd call sad cooling...

 

This thread has been resurrected after 3 months of silence - because of what Apple are charging for their latest phone?!!!!! It's not BGA abuse, if anything it's consumer/marketing/brand abuse.




^^^ That's the HS from a 15W U series CPU isn't it though?

 

It is more like phones are a part of people’s life now more than pc. The people paying 1k for these phones consider 1k not too bad if they take into account how these devices are integral part of their lives. Besides, at the rate mobile socs are improving (Apple especially), I see a future of 100% bga lol.

 

Yes that's an ULV 15w part which is plenty for it.

 

We've seen single pipe HS for 45w mobile CPU before. Total size (amount pipes or single) for Cpu HS depends on what size pipes that will be used. Single pipe can remove this amount heat if cpu is the only hardware who shall be cooled.

 

I'd actually like an I9-8950HK. 6C/12T in a notebook the form factor of a Clevo P650... Nice.

Unless AMD steps up their game and standardises AM4 for their notebook environment, which they probably won't.

 

woaw, pay the premium tag of i9 even though it should cost as much as i7, dont get tricked by intel me boy.
desktop went from i7 7700k to i7 8700k just fine with 2 more cores, same price, you dont go from i7 7820hk to i9 mobile for 2 more cores while paying extra for that tag i9. do so and intel will continue to charge you more and brain wash you more

if intel gonna sell i9 in laptop, its gotta be at least those 12+ cores

 

I'm not talking about the brand or the name. Screw the name. I'm talking about the specs. I meant I would like a 6C/12T 45 W notebook CPU, which Intel has now done. Pricing is yet to be revealed—whether or not they will gouge remains to be seen. It is likely, but let's not jump the gun, eh?

Unless you want to pay to buy me a P750TM?

 

I fix it for you
Expect that Intel will love to put in the i9 tag in laptops. Long time since we have had more than 4 cores processors in laptops.

Intel can with with such naming charge more than normal aka premium for the i9 branding in notebooks. People should be happy Intel didn't change the name to Core i9-8950HX Aka X for Extreme But with 6 BGA cores!! Maybe we will see unlocked i7-8820HK. Aka 4 cores unlocked BGA with more normal Turbo boost. Because the 6 cores i9 BGA will have low clocks in full load. Even oc'd.

From Notebookcheck.com "Lastly, the Core i9-8950HK seems be the unlocked version of the mobile H-series similar to the Kaby Lake i7-7820HK or Skylake i7-6820HK. In essentially all of our notebooks equipped with these HK-class of processors, however, the base performance gains over the i7-7700HQ or i7-6700HQ are within just single-digit percentages each time. If the Core i9-8950HK is of any indication - and we believe that it is since it carries the same number of cores and threads as the Core i7 series - then we can probably expect minimal raw performance gains yet again over the i7-8750H and i7-8550H."

 

I don't think we'll see the 8700K vs 7700K +50% performance in the 6 core 8950HK over 7820HK simply because the 8700K can and does draw more power, but the same 45W limit applies to the mobile CPUs.
Edit: above definitely applies to H CPUs; HK may have unlocked power limit but unless cooling is redesigned (e.g. P870TM) there will still be an effective power limit imposed the thermal solution and any gains will only be due to revised 14nm++ process and efficiency gains of lower clocked cores

As for AM4, AMD did some kinda deal with Asus to make an all-AMD notebook, maybe that'll go somewhere further, depends on sales numbers. From an enthusiast POV it looks like a decent first effort for its price, if I was after a CPU-heavy workstation, the underwhelming soldered rx580 stands against it though and cooling isn't fantastic for what is a stock "65W" R7-1700

 

They used to sell extreme edition mobile parts (same silicon but unlocked) for £1000 a pop.

 

eww a 6c at 45w. that means intel is gonna be hardcore restricting that GHz on that CPU of yours, that turbo would only last a min or 2. now they could do the same as they did iwth 7820hk and have no throttle implemented in but that would not be 45w anymore, clock it like a desktop and it'll use more power than a desktop chip!

 

that will changed to BGA $1000 a pop, break a piece on mobo good bye to that $1000 lolol

 

That's quite risky for the machine manufacturer at that point however.

 

Yeah, you will have to run the ++$1000 silicon lottery each time you will need a MB swap. See... some of the Alienware’s. The record is 8 MB replacement and 2 whole system exchanges. For a single owner.

 

A few things.

I don't care about the 45 W TDP 'limit', because it is an arbitrary rating. Unlocked multiplier means I can under/over volt/clock as I see fit. Notebooks with sufficiently effective cooling like the Aorus X5, P650, P950 and even AW15 should manage the thermal output a 6C/12T part generates, even after overclocking.

Second, who said I was going to clock it like a desktop? Pushing 5 GHz on a notebook chip is already idiocy. I will try for a maximum all-core Turbo clock of 4.5 GHz, given the 7820HK does 3.9 GHz single-core Turbo Boost.

If not, I will reduce my expectations and clock down if necessary. Key thing is the flexibility that a K CPU gives me, without the headache and weight of a DTR. I refuse to buy a notebook heavier than 2.5 kg. There are people like me, who don't want gigawatt performance, and are perfectly happy with slower clocks trading off for weight—believe it or not.

Also, AMD managed 850 CB points with their Ryzen 7 8C/16T part at 3 GHz, drawing a mere 30 W.

It is possible.

 

Don’t compare apples with Oranges aka 6 vs. 8 cores chips in multi threaded benchmarks.
Ryzen R7 1700 in the new Asusbook run in fact stock 3.0GHz all cores as I remember. I’m sure it will need a lot more than 30w to perform normal in Cinbench or multi threaded workflow. If the goal is 850cb with 6 cores, Oh’well. Not my cup of tea. But that would suck!!

 

but your 4.5ghz all core turbo will use more power than desktop chip at 4.5ghz all core turbo is what we're saying. literally anything on bga sucked when clocking past 3.7-3.8ghz in terms of efficiency.

 

I was talking about efficiency—I think you understand what that means? My Core i7-4710MQ was clocked at 3.5 GHz on all four cores, here are some Cinebench results:


It achieves 138 CB on the single-core test, BTW. Compared to 678 CB at 52 W package power (4C/8T), I think 850 CB at 30 W is extremely efficient. If I want raw performance, I'll plug in the notebook or get a desktop. I don't know if you remember, but a notebook is about efficient hardware...

For your information, the i7-7820HK already hits 820-830 CB at ~ 4 GHz—see this NBC Aorus X5 v7 review. Also see the Aorus X3 v7 review—i7-7820HK performing well in a 13" notebook. Key paragraph, which I know you will end-up cherry-picking from the entire review to point out a 'flaw':

All this in a cramped, hot, 13" chassis.

??? The 7700K can draw 140-150 W under a full XTU/AIDA64 load. As shown by iunlock, the i7-6820HK draws around 85-88 W under Intel XTU/AIDA64 load. How is that desktop levels of thermal output? The very fact that we now have notebook CPUs running at > 4 GHz is a big deal already.

 

The TDP limit is not "arbitrary" in the vast majority of HQ cpus - it's the long turbo power limit, or a hard system firmware throttle. Whatever multipliers you throw at it don't stick under those conditions.

HKs I dont have a lot of experience with but I doubt they're all fully unlocked re power limits, and the thinner n lighter the lower that cap would be (along with a thermal solution and adapter)

 

Thanks for the clarification—I forgot about that. Regardles, it would be inane to expect 'fully unlocked' power limits on notebook CPUs—we don't have custom pumped refrigerated watercooling to allow for 100-150 W power dissipation. The best-performing notebooks havea versatile, effective cooling system that accounts for everything—including the VRMs and VRAM chips.

Some of those desktop cooling systems weigh more than entire notebooks themselves, and I still think that the progress we've made today as far as cooling is concerned, is great. Lots of room for improvement, but compare an Aorus to an older notebook from 2010 or so...

 

HK CPUs can draw over 110W if the cancer firmware doesn't limit it. I've seen over 115W. What matters is if the firmware allows it to exceed TDP limits (the CANCER firmware in the MSI 16L13 prevents even the DESKTOP CPU's from exceeding their rated TDP), and if the mainboard's power delivery system can handle the power draw and heat.

 

Has someone actually done a scientific test of this?
When I mean scientific, i mean, running BOTH the desktop socketed CPU and the laptop CPU at the EXACT same voltage, precisely. This also means turning off any crappy VID Boosting going on in the background. One reason why the HK CPU draws so much power on the MSI laptops is due to the "VID Boosting" occuring when AC/DC Loadline is set to Auto (a poor man's way to use loadline calibration, but instead of working on the actual vcore voltage, it works on the target VID, which is FAR less graceful).

Look at this screenshot.
Notice CAREFULLY the power draw, and heat and the VID shown as well as the current set bios voltage:
1135mv, AC DC Loadline=0 (Auto).


Now notice this.
1182mv, AC DC loadline=1 (0.01 mOhms).


Or someone care to tell me exactly how the VID shown is the virtually the EXACT SAME, and how a chip at 1135mv can use more heat and draw more power than a chip at 1182mv? (obviously lack of a vcore sensor makes it hard to see what's really going on but the culprit is the AC DC loadline auto setting. Auto can use a mohms up to the Intel reference value of 2.10.

I'm sure someone who knows exactly what mOhms is reads this thread. Someone. Besides me cuz I don't know. I wonder if anyone even cares. The last time I posted this not one person even cared about it.

@Mr. Fox @Khenglish @Prema
One of you must know.

Raja @ Asus said that IA AC/DC raises the VID based on current power draw (sounds like a half-assed loadline calibration (LLC) that isn't exactly LLC to me....)

 

mOhms = milliohms. It is a value of resistance. For a constant voltage, as resistance increases, power usage ought to decrease: Power, P = (Voltage)^2/Resistance.

1 Ohm = 1 volt/1 ampere.

 

something you're not realizing is the voltage and wattage goes up expoentially once it goes past the silicon's optimal frequency. if your 6820hk cpu was to use around 90w peak at around 4ghz, then once it hits say 4.2 say around 100w, it'll add 15w then 20w then 30w for any additional 100mhz, that'll at most put you to 4.5ghz before you using more power than a desktop chip at say 4.8ghz which draws 150w.

this is undoubtedly the truth. if you choose to have your laptop chip remain at just 4ghz, desktop chip can simply do that with less than 1.1v and run on low power and runs cool.

 

The silicon is the same so with the same voltage and frequency as the desktop you are looking at the same power consumption.

 

this reminds me of @Danishblunt saying bios/firmware dont matter lolol.

 

This reminds me of Ole!!! saying that CPU's don't exist.

 

There are people like ole!!! and mr.fox who want a power hungry heavy chunk of metal that reaches 97c rather than something else for the sake of having unusable power. There is literally no reason to argue with people like them since they are completely unable to see from another perspective than their own. In their little world desktop CPU is the only way, everything else sucks. It has to be heavy, hot, and power hungry, otherwise it's not a notebook.

Ironicly your bga chip destroys ole!!!s old garbage and mr.fox sold his notebook because apparently I made him realize that desktop is the only way to go when he wants massive performance without having to get those 97c, he wouldn't admit it since he dislikes me but the fact that he went with his notebook for so long until I showed him how his notebook gets crushed by desktops. Only 1.5 weeks later he sold his junkbook and bought himself a proper desktop.

 

Contrary to what that user claims, I believe firmware matters. But regardless—

Would you let a 8520HK run at a power draw of 115W with a puny two/three pipe squirrel fan cooler like that seen on notebooks?

I don't get it. Someone buys a notebook rated for at most 180 W of power dissipation. He/she wants to run the CPU at 200% its rated power. Complains that the firmware doesn't let it, and then calls the firmware 'cancer', 'castrated', or any other number of C-words.

I've used this analogy hundreds of times here, but here we go again: would you complain that a B787 broke up in flight, and was thus built lousily, because the pilot tried to pull a Pugachev's Cobra on the airliner? Or would you complain that jet airliners aren't safe because they've been shot down by SAMs? Who's going to pay for the anti-missile defence systems, the RnD, the added weight, etc. etc. ?

Everything has its use case. I don't deny that some firmware (notably Clevo's stock option) has terrible implementation. However, many others aren't.

Have you or @Papusan seen Aorus' notebooks? They rip through games and professional workloads with their processing capability. The CPU and GPU are fully unlocked and overclockable. Yes, they do throttle—when an unrealistic, unreasonable load like P95 + Furmark is run simultaneously.

I would hope they throttled, or I'd be very worried about the lifespan of my components.

Sometimes the language used becomes downright insulting and ad hominem, which some users (they know who they are) are wont to do.

 

It does matter, he simply made that up for some reason. I never said that Bios or firmware doesn't matter, quite the opposite, Bios and Firmware can make a CPU more stable and unlock some as well. Also the Overclock people want to achieve on notebooks won't give people any significant gains which would be worth the massive heat output. If you overclock a CPU to the point where you get the most "heat to performance"- ratio, then that's the perfect line. Getting 90s temp in gaming is a red flag imo.