Jump to content
NotebookTalk

RTX 4000 mobile series officially released.


VEGGIM

Recommended Posts

i said dumb because on nice settings on just about every game i can do 4k/60 and alot of times 4k/120 with my junk 3060

ZEUS-COMING SOON

            Omen 16 2021

            Zenbook 14 oled

            Vivobook 15x oled

 

Link to comment
Share on other sites

2 hours ago, 1610ftw said:

Indeed, will have to check the options. The Asus has the advantage of the best laptop GPU and a significantly brighter and faster panel, I think it is something like 500 nits 240Hz vs 300 nits 165 Hz. Personally I am not a fan of the looks of the Asus and i also do not like how they have pared down the connectivity but people get a lot of raw performance and capability for their money.

 

Is it already known how much down the GPU ladder the m18 will still have a vapor chamber? That would be good to know for people who do not really need a 4080.

 

I guess we will have some reviews of the Alienware QHD+ panel at some point. The specs may be worse but it should still be enjoyable for most games when used indoors.

 

Hopefully 4080 and 4090 both get the vapor chamber. Per usual Dell style the specs are ambiguous so we probably will not know for sure until someone gets one and takes it apart. I am happy to see the m18 has Windows 10 drivers on the support page so far. It also appears that the TB4 ports are likely connected to the iGPU based on the spec page.

  • Bump 1

Desktop - 12900KS, 32GB DDR5-6400 C32, 2TB WD SN850, Windows 10 Pro 22H2

Clevo PE60SNE - 14900HX, 32GB DDR5-5600 CL40, 4TB WD SN850X, RTX 4070 mobile, 16.0 inch FHD+ 165hz, System76 open source firmware, Windows 10 Pro 22H2

Link to comment
Share on other sites

7 minutes ago, MyPC8MyBrain said:

i was looking for that link the other day, how did you find it?
did you had to construct the query string yourself to load these results?

 

Yep, loaded it myself, ultra long link incoming:

 

https://www.3dmark.com/search#advanced?test=spy P&cpuId=&gpuId=1546&gpuCount=0&gpuType=ALL&deviceType=ALL&storageModel=ALL&memoryChannels=0&country=&scoreType=graphicsScore&hofMode=true&showInvalidResults=false&freeParams=&minGpuCoreClock=&maxGpuCoreClock=&minGpuMemClock=&maxGpuMemClock=&minCpuClock=&maxCpuClock=

 

😄

  • Thanks 1
Link to comment
Share on other sites

10 minutes ago, win32asmguy said:

 

I guess we will have some reviews of the Alienware QHD+ panel at some point. The specs may be worse but it should still be enjoyable for most games when used indoors.

 

Hopefully 4080 and 4090 both get the vapor chamber. Per usual Dell style the specs are ambiguous so we probably will not know for sure until someone gets one and takes it apart. I am happy to see the m18 has Windows 10 drivers on the support page so far. It also appears that the TB4 ports are likely connected to the iGPU based on the spec page.

 

I honestly detest the gamery Asus design even in the non-Strix version and I would never even consider it unless maybe it was socketed. I ti salso disappointing to see that Asus would have had enough space for at least one if not two more m2 slots but they chose not to have them..

 

Still we got to give them these advantages in CPU and screen and some people will surely also prefer that the G18 weighs about 11/2 to 2 lbs less. Hopefully that will not hinder m18 sales. For me 165Hz and 300 nits will be perfectly fine in almost all situations but I would also expect that Dell will do something about their screen weakness relative to their competitors - it just is an obvious point that they have to concede to the competition.

 

I am much more disappointed that despite its heft the Alienware cannot dissipate more heat nor does it have 4 memory slots or 4 full sized m2 slots. I guess we will know more pretty soon but it strikes me as off that Dell did not manage to do a bit more with such a big chassis. But then maybe everything is so over-engineered that the m18 will be very quiet - I would certainly settle for that!

  • Bump 1
Link to comment
Share on other sites

technically speaking DLSS3 is a driver implementation NVidia is intentionally not adding to 30xx line atm, if i had to guess its an executive and marketing decisions designed to promote the 40xx line, there is nothing beside that preventing the 30xx line from getting the same DLSS3 treatment,

from what i read with the 40xx line NVidia cut more corners resulting in smaller die than 30xx gen had,
a smaller die translate to less cooling efficiency, i am curious to learn how efficient the 40xx line compared to the 30xx line,

 

12 minutes ago, 1610ftw said:

disappointing to see that Asus would have had enough space for at least one if not two more m2 slots but they chose not to have them

i believe this is an intel imposed cpu lane limitation with the exception to the HX line,

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

51 minutes ago, MyPC8MyBrain said:

@1610ftw
interesting... from your link if you arrange by cpu score my 12950HX score is the highest :classic_huh:

this has to be a bug or something as my score changes completely,

cpuscore.jpg

 

by overall score showing the right score for my unit (74th out of 34 total 😄),

cpuscore.jpg 

 

Two different scores. One is better when sorted by overall and the other when sorted by CPU - check the date.

  • Like 1
Link to comment
Share on other sites

30 minutes ago, 1610ftw said:

Two different scores

my score number actually changes too,
look at my overall score going from 13489 to 12177 when sorted by cpu, its not just my rank changes which is also incorrect btw my cpu score is not 16555 (on this bin i believe i can get close to that with LM TIM, on previous bin's i saw 15800+ with LM and even with just a good quality paste replacement, gpu score is low because at the time i didn't know how to get the gpu to use its full 150w yet 😊),

 

timespy2.thumb.jpg.34d3702ef3d9b66e5d64647252ee7b0d.jpg

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

28 minutes ago, MyPC8MyBrain said:

my score number actually changes too,
look at my overall score going from 13489 to 12177 when sorted by cpu, its not just my rank changes which is also incorrect btw my cpu score is not 16555 (on this bin i believe i can get close to that with LM TIM, on previous bin's i saw 15800+ with LM and even with just a good quality paste replacement, gpu score is low because at the time i didn't know how to get the gpu to use its full 150w yet 😊),

 

timespy2.thumb.jpg.34d3702ef3d9b66e5d64647252ee7b0d.jpg

 

Again, these are two different benchmarks from two different days, look at the dates on the right.

  • Thanks 1
Link to comment
Share on other sites

54 minutes ago, 1610ftw said:

these are two different benchmarks from two different days

so its pulling my highest ever cpu score for the comparison not highest score overall?

 

edit: i guess it does huh

found it here https://www.3dmark.com/spy/33634086 
(didnt save a screenshot, i was after higher dgpu score i couldn't get at the time), 
the closest bench i saved a screenshot from is this https://www.3dmark.com/spy/33539302 

Tweaked-3-DMARK-06.jpg 

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

not to harp on what i was trying to explain earlier with regards to what these bench represent,

from my confused example above we all learn together that despite my cpu scoring first place on that 4080 bench list it doesn't make the 12950HX a better/faster cpu than a 13900HX based on my bench results, this highlights clearly the point i was trying to make earlier with @ssj92,

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

25 minutes ago, MyPC8MyBrain said:

not to harp on what i was trying to explain earlier with regards to what these bench represent,

from my confused example above we all learn together that despite my cpu scoring first place on that 4080 bench list it doesn't make the 12950HX a better/faster cpu than a 13900HX based on my bench results, this highlights clearly the point i was trying to make earlier with @ssj92,

 

Time Spy has been all over the place for me for CPU, probably because of the combined TDP limit for CPU and GPU. You can also see that in the 4090 scores that show surprisingly low CPU numbers.

  • Thumb Up 1
Link to comment
Share on other sites

1 hour ago, 1610ftw said:

You can also see that in the 4090 scores that show surprisingly low CPU numbers

hence my initial confusion when i saw i was first place for cpu on that list i thought it has to be a bug,
with that sorted i am indeed surprised i was expecting a much higher scores range from the newer 13xxx cpu line,

 

even my highest 12950HX score not close to highest in its class (though i believe this bin can get there),

12950hxhigest.jpg  

 

1 hour ago, 1610ftw said:

probably because of the combined TDP limit for CPU and GPU

my mobile system scored that with a 240w power supply which doesn't cover both TDP and TGP running at full throttle together, there are variants out there with 330w power supply that can push further,
at some point i removed my CPU PL4 limit and or on another occasion set it to 330w,
i also bought a new 330w Dell power supply, i had a power meter on the wall outlet,
i saw 330w being pulled but bench results remained the same with both TDP and TGP satisfied,

  • Thumb Up 1

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

Better but still not performing full power yet. Surprisingly the gpu score is exactly the same as before (thought I'd fluctuate at least a little but consistency is good I guess lol)

 

Unfortunately AW didn't release 4090 model so for now I have to wait to order. 

 

yc805zt39aha1.jpeg?width=4032&format=pjp

  • Thumb Up 2

Alienware m18             : Intel Core i9 13900HX @ 5.0Ghz | nVidia GeForce RTX 4090    | K1675 | 2x1TB SSDs 

Alienware Area-51M : Intel Core i9-9900K @ 5.3Ghz    | nVidia GeForce RTX 2080    | AX210 | Samsung 970 Evo+ 
Alienware M18x R2 :    Intel Core i7 3920XM @ 4.7Ghz | nVidia Quadro RTX 5000     | AX210 | Samsung 980 PRO   
Alienware 18 :              Intel Core i7 4930MX @ 4.5Ghz  | nVidia Quadro RTX 3000  | AX210 | Samsung 980 NVMe  

More Laps: M14x (555m) | M14xR2 (650m) | M15x (980m) | M17xR3 (880m) | M18xR1 (RTX 5000) 

BEAST Server:          Intel Xeon W7-3465X 28 P-Cores | nVidia Titan V | 128GB RDIMM | Intel Optane P5800X


CS Studios YouTube: https://www.youtube.com/c/CSStudiosYT 

Link to comment
Share on other sites

5 minutes ago, MyPC8MyBrain said:

at what TDP?

No idea, it’s not my computer. All they did was update drivers. They didn’t provide me with any other info. 
 

until someone from the forums gets one of the new laptops we’re at the mercy of whatever we find online

  • Thumb Up 1

Alienware m18             : Intel Core i9 13900HX @ 5.0Ghz | nVidia GeForce RTX 4090    | K1675 | 2x1TB SSDs 

Alienware Area-51M : Intel Core i9-9900K @ 5.3Ghz    | nVidia GeForce RTX 2080    | AX210 | Samsung 970 Evo+ 
Alienware M18x R2 :    Intel Core i7 3920XM @ 4.7Ghz | nVidia Quadro RTX 5000     | AX210 | Samsung 980 PRO   
Alienware 18 :              Intel Core i7 4930MX @ 4.5Ghz  | nVidia Quadro RTX 3000  | AX210 | Samsung 980 NVMe  

More Laps: M14x (555m) | M14xR2 (650m) | M15x (980m) | M17xR3 (880m) | M18xR1 (RTX 5000) 

BEAST Server:          Intel Xeon W7-3465X 28 P-Cores | nVidia Titan V | 128GB RDIMM | Intel Optane P5800X


CS Studios YouTube: https://www.youtube.com/c/CSStudiosYT 

Link to comment
Share on other sites

i was wondering if this is your unit, i just looked it up its 157W for that cpu sku,
same as the 12950HX only 8 more efficiency cores, higher clocks for performance cores in the same power envelope, from the graph it seems that cpu is struggling with heat dissipation, it plummets to below 3GHz, and in cpu specific test it completely takes a deep dive at the most critical stage (that indicative of heat buildup),

PassMark CPU Benchmark is showing the new mobile 13xxx with 50% increase performance from the 12950HX,

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

2 hours ago, MyPC8MyBrain said:

i was wondering if this is your unit, i just looked it up its 157W for that cpu sku,
same as the 12950HX only 8 more efficiency cores, higher clocks for performance cores in the same power envelope, from the graph it seems that cpu is struggling with heat dissipation, it plummets to below 3GHz, and in cpu specific test it completely takes a deep dive at the most critical stage (that indicative of heat buildup),

PassMark CPU Benchmark is showing the new mobile 13xxx with 50% increase performance from the 12950HX,

What’s interesting is, the guy said his cpu peaked 88c but ran in high 70s for most of the run. I really am wondering if something like AWCC is messing things up or some other software he has on the computer. I’ve seen on passmark these cpus get close to desktop chips so they have the potential. 
 

hopefully someone here gets one. I’m waiting for 4090 so I won’t have one anytime soon. 
 

 

btw @win32asmguy on the m18 bios you can turn on dgpu only mode apparently so theoretically those tb4 should output from dgpu in that mode but i’ll give it a test whenever I end up with mine. Also as far as I can tell, 4080/4090 will get the full vapor chamber cooler, not sure on lower tier gpus

Alienware m18             : Intel Core i9 13900HX @ 5.0Ghz | nVidia GeForce RTX 4090    | K1675 | 2x1TB SSDs 

Alienware Area-51M : Intel Core i9-9900K @ 5.3Ghz    | nVidia GeForce RTX 2080    | AX210 | Samsung 970 Evo+ 
Alienware M18x R2 :    Intel Core i7 3920XM @ 4.7Ghz | nVidia Quadro RTX 5000     | AX210 | Samsung 980 PRO   
Alienware 18 :              Intel Core i7 4930MX @ 4.5Ghz  | nVidia Quadro RTX 3000  | AX210 | Samsung 980 NVMe  

More Laps: M14x (555m) | M14xR2 (650m) | M15x (980m) | M17xR3 (880m) | M18xR1 (RTX 5000) 

BEAST Server:          Intel Xeon W7-3465X 28 P-Cores | nVidia Titan V | 128GB RDIMM | Intel Optane P5800X


CS Studios YouTube: https://www.youtube.com/c/CSStudiosYT 

Link to comment
Share on other sites

18 minutes ago, ssj92 said:

the guy said his cpu peaked 88c

maybe intel set a new lower threshold for heat? (usually its 100c)

how much wattage is his cpu actually drawing during the bench? (he can confirm with HWiNFO32)

do you know what's the wattage on power brick being used?

 

 

9 hours ago, 1610ftw said:

if you follow the link above, arrange by cpu score,
notice they all seem to show relatively low score for 13th gen cpu,

 

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

6 minutes ago, MyPC8MyBrain said:

maybe intel set a new lower threshold for heat? (usually its 100c)

how much wattage is his cpu actually drawing during the bench? (he can confirm with HWiNFO32)

do you know what's the wattage on power brick being used?

 

 

if you follow the link above, arrange by cpu score,0
notice they all seem to show relatively low score for 13 gen cpu,

 

330w which is correct. No idea on cpu power draw he hasn't mentioned it. 

 

If a m16 with 13900hx/4080 shows up at my best buy locally, I may snag one to play with until my m18 with 4090/13980hx is ready lol 

 

one issue with reviewers, they probably just ran tests without checking to see if there's any issues. 

 

these cpus should perform very close to 13900k (without all core turbo boost enhancement turned on) as long as temps are in check. could also be some bug affecting them, who knows, too early 

 

they should be able to beat a 7700x at least easily: 

 

https://www.3dmark.com/spy/35559314

Alienware m18             : Intel Core i9 13900HX @ 5.0Ghz | nVidia GeForce RTX 4090    | K1675 | 2x1TB SSDs 

Alienware Area-51M : Intel Core i9-9900K @ 5.3Ghz    | nVidia GeForce RTX 2080    | AX210 | Samsung 970 Evo+ 
Alienware M18x R2 :    Intel Core i7 3920XM @ 4.7Ghz | nVidia Quadro RTX 5000     | AX210 | Samsung 980 PRO   
Alienware 18 :              Intel Core i7 4930MX @ 4.5Ghz  | nVidia Quadro RTX 3000  | AX210 | Samsung 980 NVMe  

More Laps: M14x (555m) | M14xR2 (650m) | M15x (980m) | M17xR3 (880m) | M18xR1 (RTX 5000) 

BEAST Server:          Intel Xeon W7-3465X 28 P-Cores | nVidia Titan V | 128GB RDIMM | Intel Optane P5800X


CS Studios YouTube: https://www.youtube.com/c/CSStudiosYT 

Link to comment
Share on other sites

22 minutes ago, ssj92 said:

they should be able to beat a 7700x at least easily

somewhere here i saw score from CB23 around the 32k if i recall correctly,
best i could get with the 12950HX on CB23 was just below 26k (25997),

the impossible is not impossible, its just haven't been done yet.

Link to comment
Share on other sites

15 hours ago, ryan said:

20k woah...now thats some serious future proofing.

 

I can game at 4k with my 3060.

Try this one below🙂 And Nvidia/Jensen said their 3080@10GB (desktop card) was their new 4K flagship 2 years ago. Damn futureproofing for 4K (RT enabled - 15.0 FPS), LOOL

 

FPS = frames per second😁

 

 

Very enjoyable. And only god Nvidia knows how the FPS will fade away in newer games 2 years down the road. 

performance-3840-2160.png

 

And where should we put 4090 Mobile@85-110W versions?

image.thumb.png.ee406be47ea4db809e9a64d40ff2c5ab.png

Hogwarts Legacy Benchmark Test & Performance Analysis Review - VRAM Usage Record

REVIEW GAME TESTING
Hogwarts Legacy lets you relive the steps of Harry Potter and is a fantastic adaptation of the books. In our performance review, we're taking a closer look at image quality, differences between the graphical setting presets, VRAM usage, and performance on a selection of modern graphics cards.

 

Atmospheric VRAM eater with upsampling fed up

ComputerBase takes a look at performance, image quality, ray tracing and upsampling in Hogwarts Legacy for the PC including benchmarks.

 

10 GB is no longer enough even without ray tracing in Ultra HD with DLSS at a sufficient frame rate. If the graphics card only has 8 GB, there are already clear signs of wear and tear in WQHD including DLSS, even without ray tracing, and even in Full HD one or the other texture is missing when playing. Ultimately, 8 GB and maximum textures are not possible.

  • Haha 2

"The Killer"  ASUS ROG Z790 Apex Encore | 14900KS | 4090 HOF + 20 other graphics cards | 32GB DDR5 | Be Quiet! Dark Power Pro 12 - 1500 Watt | Second PSU - Cooler Master V750 SFX Gold 750W (For total of 2250W Power) | Corsair Obsidian 1000D | Custom Cooling | Asus ROG Strix XG27AQ 27" Monitors |

 

                                               Papusan @ HWBOTTeam PremaMod @ HWBOT | Papusan @ YouTube Channel

                             

 

Link to comment
Share on other sites

8 minutes ago, VEGGIM said:

https://twitter.com/JarrodsTech/status/1623822414814777345?s=20

Image

Well this looks interesting. looks like some have ctgp configurable.

While this is nice, it is useless without current limit support. hopefully we can increase current limit 

Alienware m18             : Intel Core i9 13900HX @ 5.0Ghz | nVidia GeForce RTX 4090    | K1675 | 2x1TB SSDs 

Alienware Area-51M : Intel Core i9-9900K @ 5.3Ghz    | nVidia GeForce RTX 2080    | AX210 | Samsung 970 Evo+ 
Alienware M18x R2 :    Intel Core i7 3920XM @ 4.7Ghz | nVidia Quadro RTX 5000     | AX210 | Samsung 980 PRO   
Alienware 18 :              Intel Core i7 4930MX @ 4.5Ghz  | nVidia Quadro RTX 3000  | AX210 | Samsung 980 NVMe  

More Laps: M14x (555m) | M14xR2 (650m) | M15x (980m) | M17xR3 (880m) | M18xR1 (RTX 5000) 

BEAST Server:          Intel Xeon W7-3465X 28 P-Cores | nVidia Titan V | 128GB RDIMM | Intel Optane P5800X


CS Studios YouTube: https://www.youtube.com/c/CSStudiosYT 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. Terms of Use