Видео

X86 is better

As far as gaming goes the issue was oc they just couldn't oc almost at all

The ddr5 memory argument is bad, we are talking about a different order of magnitude comparing l3 cache to ddr5/ddr4

I love the idea of on package memory. It's fantastic to get the perspective of someone who sees this idea as an opportunity for improved efficiency and cost, rather than just a lack of upgradability.

The Asus ROG Ally proved this video true. Their APU is way too focused on CPU, and thus the performance isn't much better than the steam deck, with CPU cores being idle and wasting a lot of extra power and worse battery life. They then make the device bigger and heavier and overall its more expensive and a worse handheld PC than the deck.

Apple always puts in the work to make their business model.

{JUNE .\' 25 % 2024} {🌎}HELL☯; IN~LIGHT👶🏿 OF👧🏿 THE♂ PAST🔄PANDEMIC♀ AN INTUITIVE🧔🏿 THOUGHT👩🏿 COMES 🙍🏿‍♂TO🙍🏿‍♀ MIND,..=&= THE 🧑🏿‍🌾RAID👨🏿‍🎤 ON👱🏿‍♂ THE👱🏿‍♀ LEGAL ⚖LIBRARY🧑🏿‍🔧 WAS👩🏿‍🏭 SHAMEFULL🕵🏿 IN ⛩[D;C]🏦 THERE SHOULD💂🏿‍♀ BE👨🏿‍⚕ A' CHIP🧕🏿 DISTRIBUTED🌍 TO;CHURCH^TEMPLES#PARK^DISTRICTS#PUBLIC^LIBRARIES' -&-SCHOOLS,🪐..TOO👮🏿 BETTER👩🏿‍🚒 AND🧑🏿‍✈ ENSHURE A~LEGITIMATE👩🏿‍🚀 WAY🤵🏿 OF👰🏿‍♀ LIVING👼🏿 FOR🧑🏿‍🎄 THE PRESENTING🧙🏿 GENERATIONS 🧝🏿IN🧑🏿‍🏫 =ETHNIC- RUAL\URBAN -CITY= AREAS 'U:.?,:S🌏 [=ESQUIRE # NOTE=]60637 BEST^BUY 3959.....

Let's face it! Intel dropped the ball!

I’m halfway through the video and I’m in now way an expert on this topic but to me it seems a way too complicated architecture.

I have an old computer with an AMD 1090 thuban black series processor, just checked it has 900 million transistors, 14 years later there are consumer chips with a 100 times more transistors is just insane.

Watched the full video, have no job related necessity for this knowledge but have been very interested in the performance achieved by apple with the M chips and how this relates to the use of LLMs.

I think we'll have to wait for the release of the Lunar Lake laptop and the benchmark scores, but if you simply multiply the graphics scores of the Meteor Lake-H's 3DMark benchmark Time Spy and Fire Strike by 1.5, you get TS: 5250 FS: 13800. In terms of desktop GPUs, it's close to the performance of the GTX1660. In the country where I live, there are several articles that say it's 50% better in performance than the Meteor Lake-U, but if you multiply the GPU performance of the Meteor Lake-U by 1.5, it will be the same as the Meteor Lake-H's GPU performance. On a different note, is the presence or absence of hyperthreading related to the high single-thread performance of Apple silicon?

Should I buy a new laptop at this time or not?

I work for a major computer vendor and you're spot on. Your conclusion 110% speaks my mind and maims exactly what I've been saying when Intel presented us the LNL 3 weeks ago. I said that if LNL matches almost the battery perf of ARM by Qualcomm, this is going to be another Windows RT. ARM for Windows doesn't really offer a difference. We have already more performance than needed, NPU's are available en masse thanks to NVIDIA, it's just MS that firewalls for now the marketing bullshit storytelling about Copilot and that blocks other than embedded NPU's from being recognized by copilot, but this will change probably next year and they'll have to open the gates. What's left ? Battery performance. Ok, but if this gets matched, what's the point of having the whole industry shifting away from x86 ? Zero... ARM will be the thing that made Intel rethink it's architecture and from there the power efficiency and that's a good thing.

This is the best explanation of these things I’ve seen so far! I also don’t fully understand everything but I feel like you made it really easy and enjoyable to follow in one video. Thank you!

Lunar Lake looks like the biggest improvement for Intel in over a decade. In terms of performance per watt and GPU performance, it looks like Lunar Lake will beat Zen 5 and Qualcomm's X Elite. The only downside is that Lunar Lake is focused exclusively on thin and light laptops and handhelds, its not their highest performance product for mobile or desktops that is Arrow Lake which looks great for performance but will lose some efficiency and iGPU gains Lunar Lake brings.

Nicely told until you start adding up TOPS from NPU, GPU and CPU into a 120 TOPS bundle that's never going to be usable. Making use of multiple GPUs for ML workloads is already much like a bespoke design that's only ever attempted on single huge scale-out workloads, but trying to make this work with such heterogeneous tiles with client side workloads that quickly change over the life time of such a device is an effort that has no chance of ever paying off. And while really smart automated tools could do some of that work eventually, they might only arrive long after this hardware has long gone into landfills. You're nicley hunble about the limits of your insights, but that statement seems like some marketing bogus we've all been fed, that just shouldn't have slipped through what I'd otherwise judge a reasonably well adjusted biological intelligence. Lunar lake sounds almost interesting but the real attraction would be to take advantage of the underlying flexibility to buy a smaller and cheaper variant without the NPU. Because that is a lot of transistors for which I can see very little use. They could do tons of stuff with dense models, but those are mostly on video or audio streams, which--unlike on a phone--are not very likely to pass through laptops. And the bigger bandwidth hungry LLMs still have to prove any consumer value on consumer devices, before they surmount the challenge of fitting into to a sea of transistors that were set in stone/sand at design time and have a big chance of being as useless as iGPUs for all those HD generations. IMHO most of Zen's original genius was to simply use all that space for the extra 4 cores over Intel's i7: at least those saw occasional use, even if the vast majority of workloads still suffered from scalar design. Intel used those iGPUs mostly as a weapon to destroy the chipset companies (e.g. Nvidia and ATI/AMD), which worked out really well. At the volumes they produced these chips that's quite a lot of physical chip real-estate, foundry capacity and sheer waste that was never used for any good. The iGPUs have finally caught up and deliver value across quite a big range of real workloads, but they are just about to break even after around ten generations. I see a repeat here with a similarly sized slice of the SoC allocated to the NPU remaining either uselss or actually having to be fused off because regulators catch up earlier than valuable consumer use cases can be found.

Watching it after a year!

Another new Intel laptop CPU that won't support Thunderbolt 5. Yeah! Let's hope to see some Lenovo Thinkbook laptops with oculink ports at least.

The card is tiny, the GPU is massive. Also: No one bought them.

Very interesting, thank you for the great analysis.

Sorry for the long wait, the video got longer and longer the more I worked on it... Let me know if you enjoy these (very) deep-dives, or if it's too long/detailed for you. PS: the dynamic caching doesn't have anything to do with the system memory, but it's about the on-chip GPU memory. The whole GPU seems to be complete game changer, something a lot of ppl seem to have missed. This might very well be the most advanced GPU architecture right now and it will take a while until we see it's full potential.

Thank you for the informative History reveal about this new Snapdragon Chip.

LunarLake is made by TSMC Not made by Intel. 😅That's very strange.

Every video is a masterpiece. Keep doing what you're doing!

Very good

I defo switch for Win on Arm!!

Thorough, clean, comprehensive deep dives (without three times music for entertainment) are what I am looking for. Not this lalala and noise that some others do. I don’t know another channel that has such high density and concentration of facts. Thanks a lot for your work!

I wonder when Apple iplements Back side power delivery. Their m3max chip already idles at plus minus 0.1W and scales very well.

The problem with arm is it is missing a lot of instructions so it is lets say lite lol but fine if not worried about good backwards compatibility! As x86 starts to drop older instructions, you will see arm not being so special

they should've waited for the 4nm.. like nvidia did

So basically buying the laptop having RTX 40* series is a bad choice? Ive been usong an old laptop for like 13yrs and finally got a chance to get a new laptop but seems like Nvidia is scamming peiple nicely in GeForce RTX 40* series. I really wanted to get that Acer Nitro V which has i5-13gen and RTX 4050 for casual gaming and daily stuff.

I love that I came across this channel. I can't believe how good the quality of content you have while still being a sub 50,000 sub channel.

The phoenix rises from the death of the 8CX!. Nuvia's Oryon uses the rest of the snapdragon funtional blocks like gpu, dsp, modem, wifi but leading edge cpu cores and new NPU cores. This is going to change the direction of laptops forever. Apple must be really scared because they will a casualty of this success.

Shift the entire product line up down in terms of tech, but shift it up in terms of price, and then increase prices on top of that. In Battlemage we pray - but perhaps AMD's clean-sheet for next year's gpu's will bring competition if Intel doesn't - or maybe Intel's 3rd and AMD's clean-sheet hit at the same time, and we are ushered into Nvidia-free heaven.

For my laptop that basically only need to run Word, Excel, Firefox, and not much more yes Windows on ARM would be fine. For me to upgrade there needs to be A LOT more battery life that's it

ANY xx50 will be ENTRY level, not LOW end. A xx60 are going to be low END!

Its nice to have competition but if it doesnt run Linux then I am not interested.

Snapdragon should enter the desktop competition because huawei is now entering the mobile chips. Once china announce the banning of the use of foreigner chips, it will be the end of snapdragon era. Because mostly chinese phone like xiaomi oppo etc use snapdragon chips

So which is best for gaming lunar or meteor lake.

Tbh stupid consumer is the problem here. Just dont buy bad product then they will have to improve. But nah, your heads are full of shit that you cant even think before spending your hard earned money

until we got benchmark vs a M4 chip, its all talks

Its very sad there's no perfect chip, i like some moves intel made with this desing like interposer for connectivity, the complete turn off of some tiles if they're not required to mantain a low idle power draw, the extra efficiency on E cores to be the main cores used to low the power draw under workloads and monolithic design on the computing chiplet, if just E cores where more like Zen c and less like big.LITTLE and the ram doesn't being integrated on the SoC...

for *Corporate America's Automation of Middle Management,* e-cores are more than enough.

Great info! Thanks

Came from the future to tell NO.

Great analysis! Really a real breakthrough.

Haswell just saying..

I always thought they had transistors on multiple layers, not just the bottom.

Do the Z1 extreme SoC next, please!