I agree with you. The only benefit I see is its consumer friendly for them to release yearly to force depreciation and upgrades while we choose when to upgrade.
@@friedrichranada7133 i mean you still good with 12th gen intel or zen 3 for years, but i understand we always want the newest fancy stuff, me included, thats why a 2x years release would force us to stay on that specific plattform for longer wich is good
Most people keep their cpu for 5-6 years at a time if not more. Cpus have becomes just as bad as iPhones now. Overhyped and shoved down your throat by RUclipsrs trying to make you get the shiny new thing. My friend has a 9900k and can still run anything on the market just fine.
@@Zibbybob Hopefully they have a plan. But will they continue using TSMC? I thought it might be temporary. Is kind of embarrassing for a major fab owner like Intel to using other fabs to build their stuff.
It's frustrating to see how often Intel is changing their socket and therefore forcing new motherboard upgrades. I think it's actually lowering their sales for each particular chip generation because people who are any level of price conscious are waiting for a more long-lived socket.
10000000000000%. They should be like AMD's release of the 7000 and 7000v2 (sorry, "9000 series") parts on a single socket. Reminder that the 12th, 13th, and 14th (again... 13th... I don't even know if 13v2.. more like 13.1th) were all on the same LGA1700 socket. AMD abandoned Threadripper after 2 generations then a whole new motherboard was affectively abandoned after a single generation.
It’s not just intel, it’s also the motherboard manufacturers failing quality control as well lately.premium LGA 1700 EATX mobos don’t seem to be passing quality assurance lately (ASUS MSI)
Most people aren't upgrading their PCs every year; it's usually every five years at least. By that time, you'll likely want to upgrade to a better motherboard, as new technology will increase performance and support the latest components.
Actually the reason behind Intel’s decision for TSCM to instead of their own nodes was cost. Intel 20a was supposed to be the first step before 18a that was supposed to be an iteration of 20a but 20a was not going to be sold to other companies it was meant to be just for Intel products. The cost to transition to a new architecture 18awas too high when they already have a mature 18a process (which means they consider their 18a good enough to be sold to other companies without the feedback of 20a) so they went with TSMC instead
Because it is separate processor on the die and you would most likely use a dedicated GPU like Battlemage the integrated graphics would be more like backup or hooked to the TV to play video. Personally I think it is logical and cost saving.
@@XDAVideo "Tech Power Up" in their breakdown gives Xe2. Also PL2 "Extreme": 295 W . I did put it to 275W but it seems to have some overclocking room. In notes Intel Arc Xe2 Graphics Dynamic Frequency: 2000 MHz "Lion Cove" P-Cores "Skymont" E-Cores
untill business laptops ships with windows intel is not going anywhere near future, just like windows phone windows os will not work very well on ARM , microsoft will not survive without Intel, both are supporting each other.
Windows works very well on ARM. The problem is not the OS but applications compatibility. Next year Nvidia will make its first ARM SoC running Windows, ARM is here to stay and it's a good thing for competition
It's primarily because more people use Intel processors; therefore, why wouldn't Microsoft spend more time optimizing for them? Currently, 71% of laptop CPUs and 64% of all computers are Intel. Intel itself isn't directly responsible for this; rather, Microsoft knows that more people use Intel products. Just like when Windows 10 was released, Intel CPUs performed better on Windows until Microsoft started optimizing for AMD. The same thing is expected to happen with Windows 11. Microsoft just needs time to work on that optimization for amd
Will the socket be kept for more that 1 generation (like the 12th, 13th, 14th)? I want to upgrade, motherboard CPU cooler memory and hopefully this could accept the 16th or maybe 17th gen if Intel sticks to the socket lga1851.
@@XDAVideo Intel is losing consumers this way. They shouldn't leave people hanging. We aren't going to be buying on hope without confirmation. They should promise from now that they will be using that socket for Nova Lake as well. AMD is giving like 4-5 generations on a single socket, is not much to ask from Intel to at least commit to 2 generations. That could built some trust for someone that is thinking whether he should invest on an Intel system or an AMD one.
@@Consola260 Trust me I feel you. I have a family that will always come first. Younger me would have the latest but right now I would rather upgrade my son before me. That’s the main reason I want to hold out as long as I can even though my eye is twitching to upgrade but have to make sure funds are also right and won’t leave me in more debt.
As I I have gotten older I have realized that I don’t have to have the latest all the time. Yes the 4790k and 3770k may be older, but if they are doing what you want and you are good with it then ok. People are acting like they can’t play low FPS but consoles have been 30fps and sometimes less for years. I tell folks yes I’ll upgrade one day, but I want my next upgrade to BLOW my skirt up like I wear one. My monitor goes up to 144 anyway so all thus 200, 300fps stuff means nothing to me. At least 100 and I’m satisfied for now. Plus pc building is too expensive now with the economy, family and mortgage.
as a gamer programmer a music producer mix and mastering engineer i have both intel and amd ryzen and l like them equally they serve me well and ill upgrade to their latest offers to do my tasks
Man nobody gives af about Bartlett lake S. Will I buy one? Probably, cuz I’m a loser, but seriously, nobody gives af about a 12 core 13th gen, scaled down cpu. If u want more performance, buy arrow lake. If ur broke…get a job, & buy a new motherboard. Jesus Christ.
@@godnamedtay Why should I have to buy another platform when I bought one only a year and a half ago? I think you are childish and very impressionable by what is written in the press, ...where newest best crap is promoted. Arrow Lake is barely better than 14 gen. There is no reason to get excited by it!! Btw. how old are you? Most people on the internet nowadays are supposedly children and teenagers.
When we talk about Arrow Lake and INTEL architecture we can't not to talk about Battlemage. This is huge. It includes my suspension about Nvidia 5080/5090 also. INTEL ARC A770 has 32 Xe cores. Predicted top Battlemage has predicted more than 32 Xe2 cores. Xe vs Xe2 already is 50% jump in performance. I've seen rumors of 40 Xe2 cores if not more for the top Battlemage. Now we are closing to double the performance of the first gen INTEL ARC GPU's. A770 GPU clock speed is 2.1 GHz. The upcoming GPU BMG-G21 that is predicted to be the replacement for the A750 and has specs reaching to 2.85 GHz. So we should expect 3 GHz and above from the top Battlemage performer. Then there is the Memory speed/Bandwidth from 17.5 Gbps/560 GB/s from A770 to 19 Gbps/608 GB/s from the top Battlemage. Now add Intel Application Optimization (APO) that promise 10-50% gain with top INTEL CPU. INTEL XeSS XMX also is close to equal to DLSS. Don't forget that INTEL ARC via REBAR uses 2 silicon architecture taking the advantage of both silicon and doubling the performance. For example i7-13700k 253W (max) + A770 225W (max) = 478W (max). In general Wattage = performance. It is a good indication now that the clock speeds are predicted to match the competitions clock speeds. Plus with two silicon architecture there is no queuing or waiting meaning every frame is perfect. A770 in reality already gives you better experience on 4k than 4070ti. Current testing doesn't give you the performance of 2 silicon combined performance or recognize architecture where you don't have queuing or waiting, counting every error and broken frame. To beat you need at least double the FPS count from AMD/Nvidia single silicon architecture at high resolution demanding tasks where CPU plays only very small part for AMD/Nvidia GPU performance. Hence insane Wattage like 450W from 4090 or the predicted 600W from 5090 (that I suspect of being dual GPU). INTEL is cooking something big for sure. Don't forget that Arrow Lake also plays a huge part. Battlemage predicted Wattage is 250-275W. So it looks like it will be close to 4080/5080 cards with combined wattage of 500-550W with Arrow Lake. Arrow Lake max Wattage is predicted to be max 250-275W. It would make sense to match the Wattage on GPU and CPU to avoid bottlenecks (as you can also see speculation on 5090 doubling the specs of 5080 as two equal sides. also the INTEL CPU/GPU pairing tool supports this narrative). So when talking about Battlemage you should also add the Arrow Lake into the picture because it is going to be a big part of the performance if not half. Also new Motherboards for the Arrow Lake most likely will have enhanced REBAR for pairing via native 5.0 PCI interface to the CPU. Would make sense (also for 5090 to have two GPU's on the same card). 4090 24gb VRAM 384 bit bus 450W 4080 16gb VRAM 256 bit bus 320W 4070ti 12gb VRAM 192 bit bus 285W a750 8gb VRAM 256 bit bus 225W a770 16gb VRAM 256 bit bus 225W Price point for the top Battlemage I predict is $500 (for reference 4080 +$1300). But here is the catch. You need top upcoming Arrow Lake CPU to get all the performance from Battlemage and that is a new build $500 for the CPU and $500 for the Motherboard. That's $1500 as a rounded ruff estimate. Now it doesn't look that cheap anymore. But if this combo competes with 4080/5080 then it starts to look extremely good value if you are looking for a new build. Also when the Wattage of the 4080/5080 are 320W/predicted 400W and add the Wattage of the CPU it is looking very similar in power cost. Arrow lake Wattage predicted Wattage is 250-275W. Top AMD CPU 170W. Battlemage prediction is 250-275W. Top INTEL combo 275W + 275W = 550W vs Top AMD CPU 170W + 320W(400W)4080(5080) = 490W(570W). As a summery I think we are looking at doubling the ARC Alchemist/13th gen/14th gen CPU performance with Battlemage/Arrow Lake CPU that would compete with 4080/5080 performance. Potential also gives hope that it might even be more and beat these cards. It will come down to price and build new vs old for the top performance. So look for the Arrow Lake CPU launch and you can be sure Battlemage won't be far behind. Going back to the dual GPU of 5090 prediction or maybe just a feeling of it it would do similar job as the INTEL pairing to INTEL CPU. There is talk about Nvidia getting into CPU market and I think this might just be the reason to have two silicon instead of one in 5090. This is just guessing and making conclusions/speculations from the specs. When I look at the specs of 5090 it is pretty much double than the 5080 to the point. The first thought that comes to mind is 5090 has two (2) 5080 GPU's on the same card and because 600W is the limit/max that can be supported and that it has the same 12 pin power supply indicates two GPU's in the one card as it is 2 x 300W not single 600W. This in my knowledge can be susteined through the existing power architecture and the troublesome 12 pin connector from the 4090. Also because the two GPU's are not pushed as hard as 5080 (2 x 300W for 5090 vs 1 x 400W for 5080) there is less heat/GPU and similar size cooler can fit the 5090 as it is predicted it will be 2 slot card. 600W to single silicon would run so hot that it would need at least 3 slot cooler. Performance/Wattage/specs/temperature/cooling all indicate that there is two silicon instead of one GPU in the 5090. Old SLI comes to mind but in one card. Also I think Nvidia has seen the future of pairing two silicon from INTEL but they do not have CPU's at this point of time and the closest they would come is with dual GPU card. To me this is a scam because now you have 3 silicon in your system but only need 2 and your CPU is on idle and RAM use minimal as it has never really made that much difference in performance. INTEL Arrow Lake will change this because it is fully utilized and paired to take half the tasks of the paired INTEL ARC GPU. We end up with less hardware but same performance if not more with RAM counted in also. This does raise some questions like VRAM. Battlemage predicts to have 16 GB VRAM. With the new Arrow Lake can you add RAM as your graphics memory also if there is native link via PCI 5.0 between GPU/CPU is a question and how well. It almost looks like INTEL combo will have two hearts beating and syncronised together. Food for thought.
Intel Core Ultra 9 285K Physical Socket: Intel Socket 1851 Foundry: Intel Process Size: 7 nm Package: FC-LGA18V tJMax: 105°C Cache Cache L1: 112 KB (per core) Cache L2: 3 MB (per core) Cache L3: 36 MB (shared) E-Core L1: 96 KB (per core) E-Core L2: 4 MB (per module) Processor Market: Desktop Production Status: unreleased Release Date: Oct 2024 Part#: unknown Core Config # of Cores: 24 # of Threads: 24 Hybrid Cores: P-Cores: 8 E-Cores: 16 SMP # CPUs: 1 Integrated Graphics: Arc Xe2 Graphics 64EU Performance Frequency: 3.7 GHz P-Core Turbo: 5.4 GHz P-Core Boost 3.0: 5.6 GHz Turbo Clock: up to 5.7 GHz E-Core Frequency: 3.2 GHz up to 4.6 GHz Base Clock: 100 MHz Multiplier: 37.0x Multiplier Unlocked: Yes TDP: 125 W PL1: 125 W PL2: 250 W PL2 "Extreme": 295 W Architecture Codename: Arrow Lake-S Generation: Ultra 9 (Arrow Lake) Memory Support: DDR5 Depends on motherboard Rated Speed: 5600 MT/s Memory Bus: Dual-channel ECC Memory: No PCI-Express: Gen 5, 16 Lanes (CPU only) Chipset: Z890 Features MMX SSE SSE2 SSE3 SSSE3 SSE4.2 AVX AVX2 FMA3 SHA EIST Intel 64 XD bit VT-x VT-d AES-NI TSX TXT CLMUL F16C BMI1 BMI2 ABM ADX RdRand VNNI vPro DLBoost 3.0 TBT 2.0 TBT 3.0 TVB Notes Intel Arc Xe2 Graphics Dynamic Frequency: 2000 MHz "Lion Cove" P-Cores "Skymont" E-Cores The Intel Core Ultra 9 285K will be a desktop processor with 24 cores, that is expected to launch in October 2024. It is part of the Ultra 9 lineup, using the Arrow Lake architecture with Socket 1851. Core Ultra 9 285K has 36 MB of L3 cache and operates at 3.7 GHz by default, but can boost up to 5.7 GHz, depending on the workload. Intel is building the Core Ultra 9 285K on a 7 nm production process, the transistor count is unknown. You may freely adjust the unlocked multiplier on Core Ultra 9 285K, which simplifies overclocking greatly, as you can easily dial in any overclocking frequency. With a TDP of 125 W, the Core Ultra 9 285K consumes a lot of power, so good cooling is definitely needed. Intel's processor supports DDR5 memory with a dual-channel interface. The highest officially supported memory speed is 5600 MT/s, but with overclocking (and the right memory modules) you can go even higher. Actual memory technology support depends on the chosen motherboard, the processor itself supports multiple memory types, but most motherboards have only one kind of slot. For communication with other components in the machine, Core Ultra 9 285K uses a PCI-Express Gen 5 connection. This processor features the Arc Xe2 Graphics 64EU integrated graphics solution. Hardware virtualization is available on the Core Ultra 9 285K, which greatly improves virtual machine performance. Additionally, IOMMU virtualization (PCI passthrough) is supported, so that guest virtual machines may directly use host hardware. Programs using Advanced Vector Extensions (AVX) will run on this processor, boosting performance for calculation-heavy applications. Besides AVX, Intel is including the newer AVX2 standard, too, but not AVX-512. techpowerup
Decided to wait for the new intel arrow lake cpu before i build a new gaming pc. I hope Intel will be more power efficient and cool again. Otherwise i'm forced to buy amd, never really liked that brand when it comes to cpu. Just personal preference.
NOooooo the 15 Intel gen fu**ing sucks. Screw all this optimizaton for work station desktop pc. I WANT TO GAME! Looks like im forced to buy AMD ZEN 5. ( Never liked AMD btw.)
i have a 13900k and primarily game. i also have a asus ryujin II and will tell you, under load I have never seen my cpu hit 70c its usually sitting around 58-61c under load. maybe I got a great chip but I dont know what all the complaints are about. this is a fantastic cpu and have had it since release.
That is not the case for the vast majority of people. Maybe you have amazing cooling and also hit a jackpot on silicon lottery but for most even with a big good air cooler that thing hits close to 90c and even with water-cooling it goes over 80c.
More interesting is whether it will be worth considering that the last two generations were defective from the factory. Intel themselves said that only i3 and i5 which are not K version are not affected. I've always liked Intel, but after this disgrace with the last generations, I wonder if I'll ever buy anything made by them again. I'm glad I didn't get a 13 or 14 generation CPU.
The only good thing I see it doing is to lower the prices of the 14th generation and maybe update my Kaby Lake, which I have left over at the moment. It would be an update on a whim more than anything.
intel will be on par to amd for about a month, then amd will be coming out with their x3d variants like 9800/9950x3d and amd will be ahead, again, probably for a good chunk of 2025 due to the 3d v-cache. no major upgrades until maybe late 2025 from intel when they'll come out with panther lake and powervia. amd's gonna be crushing intel this next year.
So Oct 24th it will be released; I wonder how its power efficiency is when compared with AMD Ryzen cpus. Also the naming convention is weird, so Core 265 and up means more than 6 Performance cores; need 8 Performance cores for decent gaming I believe.
@@laszlozsurka8991 No game use only one core; you're forgetting that PCs also run background applications and system tasks. Some games even have additional apps that need to be running for the game to function properly. While the game itself may primarily utilize one core, it can occasionally use two. This isn't misinformation; if you had done just a little bit of research, you would understand this
@@n.d.n.e1805 "No game uses only one core; you're forgetting that PCs also run background applications and system tasks" When tf did I say that games use only one core? Don't put arguments which I never stated. Try running Hogwarts Legacy with only a 2 core CPU and see what happens. Most modern games do need at least 6 cores with 8 being the preferred. There's a reason why 8 P-core CPUs are faster than 6 P-core CPUs or lower. Also GTA 6 is not even out yet so how do you even know how that "GTA 6 will rely on one core"?
@@laszlozsurka8991When did I ever say that you said that games only use one core? I said that. And yes, most games primarily utilize one core, maybe two at most. This is evident because most AAA games typically rely on one core, with two being the upper limit in many cases. Just do research it only takes 2 seconds. Also are you dumb? Of course, games will run better on a CPU with more cores but the game itself only utilizes one core. That's why CPU with higher single-core performance usually get better performance when it come to games
you dont think it was great timing that two weeks before the so called cpu issues that the US gov told us they were going to be investing 100 Billion dollars into Intel over the short term, then like 2 weeks later there is a story that craters the stock to 50% of its value. Why by high when you can create a story that can help give you a controlling interest. Pretty funny stock story, cause big plays always go down exactly like this.
Pretty embarrassing if you ask me. A chip leader for many decades---no one was even remotely a competitor, and responsible for revolutionizing the PC industry but yet unable to manufacture their own chips in-house. So sad, Intel.
well at least their entire line up cpu will be cutting edge for first time since zen1 launch after arrow lake release server(granite) intel3, mobile(lunar) tsmc3, dekstop tsmc3(arrow)
How? They're still spending billions to catch up to TSMC in technology; they just aren't there yet. The last few CEOs just sat on their hands and did nothing. So the new CEO has to catch up to TSMC. Also, Intel still produces a lot of the technology that the 15th generation will use; they are just buying the nodes from TSMC. So intel still make CPU
No, it won't. the i9 285k TDP is 125 watts, and the maximum power consumption is 250 watts. The 9950X has a TDP of 170 watts, and at maximum, it can draw 250 watts.
Intel isn't buying nodes from TSMC for shits and giggles, it wants to COMPARE with real world HARD DATA how it's own 5nm/4nm/3nm/2nm nodes fair with TSMC in _volume_ production. The key is "volume" as any of the big three can do small "hand picked" batches easily. But Not only that, intel can run benchmarks to see if TSMC is truly better or worse than they're node-for-node from the SAME CPU!! but manufactured by TWO totally different companies! THEN, they'll figure out just HOW far behind or better they're in terms of volume efficiency and node superiority! Pretty sure they had to pay TSMC millions upfront to make those chips.😅😅
@@tim3172 Actually the reason behind Intel’s decision for TSCM to instead of their own nodes was cost. Intel 20a was supposed to be the first step before 18a that was supposed to be an iteration of 20a but 20a was not going to be sold to other companies it was meant to be just for Intel products. The cost to transition to a new architecture 18awas too high when they already have a mature 18a process (which means they consider their 18a good enough to be sold to other companies without the feedback of 20a) so they went with TSMC instead
i'm excided for Intel Arrow lake, also we dont need new cpus all years two year intervall would be good
As I agree with you, I don’t appreciate Intel producing overheating failing CPUs each year either
@@eevd350z nah, i belongs to all comapnies.
I agree with you. The only benefit I see is its consumer friendly for them to release yearly to force depreciation and upgrades while we choose when to upgrade.
@@friedrichranada7133 i mean you still good with 12th gen intel or zen 3 for years, but i understand we always want the newest fancy stuff, me included, thats why a 2x years release would force us to stay on that specific plattform for longer wich is good
Most people keep their cpu for 5-6 years at a time if not more. Cpus have becomes just as bad as iPhones now. Overhyped and shoved down your throat by RUclipsrs trying to make you get the shiny new thing. My friend has a 9900k and can still run anything on the market just fine.
I wish Intel success as everyone will benefit when there is competition.
Don’t u have driver issues to fix?
Lol come again?
@mars5268 if you get monopolistc pricing, no one will be LOL. 😁
@godnamedtay yeah, you have a good day too!
man Im waiting to build a new pc and i want this arrowlake so hard
Pause…
Same, however also want to hold for a month (minimum) after release to understand how it performs in the market.
And lets not forget about the mother board prices that will probably be wild.@@MishimaZaibatsuCorp
@@jeremyniels yes, rumours suggest so too
Crazy, don't worry, don't reply. I get it
the fact we're getting a intel desktop CPU on TSMC is going to be the biggest blessing since the launch of Zen 1 and the 1080 Ti.
Aaaand..it wasn't.
@@SIPEROTH Definitely does leave ground work for their next gen
@@Zibbybob Hopefully they have a plan.
But will they continue using TSMC? I thought it might be temporary.
Is kind of embarrassing for a major fab owner like Intel to using other fabs to build their stuff.
It's frustrating to see how often Intel is changing their socket and therefore forcing new motherboard upgrades. I think it's actually lowering their sales for each particular chip generation because people who are any level of price conscious are waiting for a more long-lived socket.
10000000000000%.
They should be like AMD's release of the 7000 and 7000v2 (sorry, "9000 series") parts on a single socket.
Reminder that the 12th, 13th, and 14th (again... 13th... I don't even know if 13v2.. more like 13.1th) were all on the same LGA1700 socket.
AMD abandoned Threadripper after 2 generations then a whole new motherboard was affectively abandoned after a single generation.
It’s not just intel, it’s also the motherboard manufacturers failing quality control as well lately.premium LGA 1700 EATX mobos don’t seem to be passing quality assurance lately (ASUS MSI)
@@eevd350z you don't know the difference between quality control and quality assurance so why listen to what you have to say?
Most people aren't upgrading their PCs every year; it's usually every five years at least. By that time, you'll likely want to upgrade to a better motherboard, as new technology will increase performance and support the latest components.
Unfortunately AMD is doing the same shit too... AM5 is likely gonna support Zen 4 and Zen 5... I doubt Zen 6 will be on AM5 but one can hope.
Actually the reason behind Intel’s decision for TSCM to instead of their own nodes was cost. Intel 20a was supposed to be the first step before 18a that was supposed to be an iteration of 20a but 20a was not going to be sold to other companies it was meant to be just for Intel products. The cost to transition to a new architecture 18awas too high when they already have a mature 18a process (which means they consider their 18a good enough to be sold to other companies without the feedback of 20a) so they went with TSMC instead
Lol uh duh
I think it has something to do with heat and power as well, especially after all the issues they had.
Did you say Intel Arrow Lake won't be using Xe2 graphics?
Leaks all point to graphics being Alchemist-based not Battlemage
@@XDAVideoMakes sense tbh, because RAM isn't soldered & cache isn't shared as good as on LNL
Because it is separate processor on the die and you would most likely use a dedicated GPU like Battlemage the integrated graphics would be more like backup or hooked to the TV to play video. Personally I think it is logical and cost saving.
@@XDAVideo "Tech Power Up" in their breakdown gives Xe2.
Also PL2 "Extreme": 295 W . I did put it to 275W but it seems to have some overclocking room.
In notes
Intel Arc Xe2 Graphics Dynamic Frequency: 2000 MHz
"Lion Cove" P-Cores
"Skymont" E-Cores
Lot of different information out there. But I did find Xe2 was also mentioned by "Tech Power Up" in their breakdown. Time will tell.
Looks like Intel does not care much about Hyperthreading anymore.
For their laptop cpu anyways
untill business laptops ships with windows intel is not going anywhere near future, just like windows phone windows os will not work very well on ARM , microsoft will not survive without Intel, both are supporting each other.
Windows works very well on ARM. The problem is not the OS but applications compatibility. Next year Nvidia will make its first ARM SoC running Windows, ARM is here to stay and it's a good thing for competition
@@fred-le-reveurWindows on ARM is a joke. No, Windows as a whole is a joke and it's showing even more now so.
It's primarily because more people use Intel processors; therefore, why wouldn't Microsoft spend more time optimizing for them? Currently, 71% of laptop CPUs and 64% of all computers are Intel. Intel itself isn't directly responsible for this; rather, Microsoft knows that more people use Intel products. Just like when Windows 10 was released, Intel CPUs performed better on Windows until Microsoft started optimizing for AMD. The same thing is expected to happen with Windows 11. Microsoft just needs time to work on that optimization for amd
@@fred-le-reveur unfortunately its very well behind the competition by a huge margin
Under the Leadership Of That Bum Jabraoni Sataya Nadella it's all gone down the shitter,Windows 7 Was the Last Great OS from Microsoft @@xslvrxslwt
Intel should have just left producing their precious 14nm +++++++++++ cpus in peace
they stopped that long time ago, now their using a 3nm transistor
Do arrow lake will come for laptops?
Lunar lake is the laptop version
@@n.d.n.e1805is lunar lake is a bit weak for creator or gaming oriented laptop?
Yes, there are H series chips coming to replace the Meteor Lake - H lineup. Lunar Lake replaces the Meteor Lake - U lineup.
Does the stability issues on 13th and 14th also affect laptop cpus? Or only desktop?
Only desktop
Will the socket be kept for more that 1 generation (like the 12th, 13th, 14th)? I want to upgrade, motherboard CPU cooler memory and hopefully this could accept the 16th or maybe 17th gen if Intel sticks to the socket lga1851.
Nova Lake is also rumoured to (eventually) use it, but not confirmed yet
@@XDAVideo Intel is losing consumers this way. They shouldn't leave people hanging. We aren't going to be buying on hope without confirmation.
They should promise from now that they will be using that socket for Nova Lake as well.
AMD is giving like 4-5 generations on a single socket, is not much to ask from Intel to at least commit to 2 generations.
That could built some trust for someone that is thinking whether he should invest on an Intel system or an AMD one.
Still using my 10850k. Yes I know it’s old now but it’s still working with no issues. I hope for it to last hopefully 2 more years.
10700K here, but I'm not waiting 2 years. Just waiting for Intel 15th gen and Nvidia RTX 5 series
@@Consola260 Trust me I feel you. I have a family that will always come first. Younger me would have the latest but right now I would rather upgrade my son before me. That’s the main reason I want to hold out as long as I can even though my eye is twitching to upgrade but have to make sure funds are also right and won’t leave me in more debt.
Here i am on my 4790K 🤣
@@87ranner rookie numbers. 3770k here.
As I I have gotten older I have realized that I don’t have to have the latest all the time. Yes the 4790k and 3770k may be older, but if they are doing what you want and you are good with it then ok. People are acting like they can’t play low FPS but consoles have been 30fps and sometimes less for years. I tell folks yes I’ll upgrade one day, but I want my next upgrade to BLOW my skirt up like I wear one. My monitor goes up to 144 anyway so all thus 200, 300fps stuff means nothing to me. At least 100 and I’m satisfied for now. Plus pc building is too expensive now with the economy, family and mortgage.
Can’t wait for this to come out and watch the 13th and 14th gen prices drop
Can’t wait to use DDR5 with my 15th gen with an RTX 5090… still trying to think of more high end things with 5 in it
Hmmm
Is the Lga 1851 already drop to buy
as a gamer programmer a music producer mix and mastering engineer i have both intel and amd ryzen and l like them equally they serve me well and ill upgrade to their latest offers to do my tasks
@@kenshijin-d3y the best now and still is 12900k without driver issues paired with your preffferd gpu
I am waiting to see Bartlett Lake cpu's... coming Q1 2025 and p-core only chips in Q3 .
Man nobody gives af about Bartlett lake S. Will I buy one? Probably, cuz I’m a loser, but seriously, nobody gives af about a 12 core 13th gen, scaled down cpu. If u want more performance, buy arrow lake. If ur broke…get a job, & buy a new motherboard. Jesus Christ.
@@godnamedtay Why should I have to buy another platform when I bought one only a year and a half ago? I think you are childish and very impressionable by what is written in the press, ...where newest best crap is promoted. Arrow Lake is barely better than 14 gen. There is no reason to get excited by it!! Btw. how old are you? Most people on the internet nowadays are supposedly children and teenagers.
When we talk about Arrow Lake and INTEL architecture we can't not to talk about Battlemage. This is huge. It includes my suspension about Nvidia 5080/5090 also.
INTEL ARC A770 has 32 Xe cores. Predicted top Battlemage has predicted more than 32 Xe2 cores. Xe vs Xe2 already is 50% jump in performance. I've seen rumors of 40 Xe2 cores if not more for the top Battlemage. Now we are closing to double the performance of the first gen INTEL ARC GPU's.
A770 GPU clock speed is 2.1 GHz. The upcoming GPU BMG-G21 that is predicted to be the replacement for the A750 and has specs reaching to 2.85 GHz. So we should expect 3 GHz and above from the top Battlemage performer. Then there is the Memory speed/Bandwidth from 17.5 Gbps/560 GB/s from A770 to 19 Gbps/608 GB/s from the top Battlemage.
Now add Intel Application Optimization (APO) that promise 10-50% gain with top INTEL CPU. INTEL XeSS XMX also is close to equal to DLSS.
Don't forget that INTEL ARC via REBAR uses 2 silicon architecture taking the advantage of both silicon and doubling the performance. For example i7-13700k 253W (max) + A770 225W (max) = 478W (max). In general Wattage = performance. It is a good indication now that the clock speeds are predicted to match the competitions clock speeds.
Plus with two silicon architecture there is no queuing or waiting meaning every frame is perfect. A770 in reality already gives you better experience on 4k than 4070ti. Current testing doesn't give you the performance of 2 silicon combined performance or recognize architecture where you don't have queuing or waiting, counting every error and broken frame. To beat you need at least double the FPS count from AMD/Nvidia single silicon architecture at high resolution demanding tasks where CPU plays only very small part for AMD/Nvidia GPU performance. Hence insane Wattage like 450W from 4090 or the predicted 600W from 5090 (that I suspect of being dual GPU).
INTEL is cooking something big for sure. Don't forget that Arrow Lake also plays a huge part. Battlemage predicted Wattage is 250-275W. So it looks like it will be close to 4080/5080 cards with combined wattage of 500-550W with Arrow Lake. Arrow Lake max Wattage is predicted to be max 250-275W. It would make sense to match the Wattage on GPU and CPU to avoid bottlenecks (as you can also see speculation on 5090 doubling the specs of 5080 as two equal sides. also the INTEL CPU/GPU pairing tool supports this narrative). So when talking about Battlemage you should also add the Arrow Lake into the picture because it is going to be a big part of the performance if not half. Also new Motherboards for the Arrow Lake most likely will have enhanced REBAR for pairing via native 5.0 PCI interface to the CPU. Would make sense (also for 5090 to have two GPU's on the same card).
4090 24gb VRAM 384 bit bus 450W
4080 16gb VRAM 256 bit bus 320W
4070ti 12gb VRAM 192 bit bus 285W
a750 8gb VRAM 256 bit bus 225W
a770 16gb VRAM 256 bit bus 225W
Price point for the top Battlemage I predict is $500 (for reference 4080 +$1300). But here is the catch. You need top upcoming Arrow Lake CPU to get all the performance from Battlemage and that is a new build $500 for the CPU and $500 for the Motherboard. That's $1500 as a rounded ruff estimate. Now it doesn't look that cheap anymore. But if this combo competes with 4080/5080 then it starts to look extremely good value if you are looking for a new build. Also when the Wattage of the 4080/5080 are 320W/predicted 400W and add the Wattage of the CPU it is looking very similar in power cost. Arrow lake Wattage predicted Wattage is 250-275W. Top AMD CPU 170W. Battlemage prediction is 250-275W. Top INTEL combo 275W + 275W = 550W vs Top AMD CPU 170W + 320W(400W)4080(5080) = 490W(570W).
As a summery I think we are looking at doubling the ARC Alchemist/13th gen/14th gen CPU performance with Battlemage/Arrow Lake CPU that would compete with 4080/5080 performance. Potential also gives hope that it might even be more and beat these cards. It will come down to price and build new vs old for the top performance. So look for the Arrow Lake CPU launch and you can be sure Battlemage won't be far behind.
Going back to the dual GPU of 5090 prediction or maybe just a feeling of it it would do similar job as the INTEL pairing to INTEL CPU. There is talk about Nvidia getting into CPU market and I think this might just be the reason to have two silicon instead of one in 5090. This is just guessing and making conclusions/speculations from the specs. When I look at the specs of 5090 it is pretty much double than the 5080 to the point. The first thought that comes to mind is 5090 has two (2) 5080 GPU's on the same card and because 600W is the limit/max that can be supported and that it has the same 12 pin power supply indicates two GPU's in the one card as it is 2 x 300W not single 600W. This in my knowledge can be susteined through the existing power architecture and the troublesome 12 pin connector from the 4090. Also because the two GPU's are not pushed as hard as 5080 (2 x 300W for 5090 vs 1 x 400W for 5080) there is less heat/GPU and similar size cooler can fit the 5090 as it is predicted it will be 2 slot card. 600W to single silicon would run so hot that it would need at least 3 slot cooler. Performance/Wattage/specs/temperature/cooling all indicate that there is two silicon instead of one GPU in the 5090. Old SLI comes to mind but in one card. Also I think Nvidia has seen the future of pairing two silicon from INTEL but they do not have CPU's at this point of time and the closest they would come is with dual GPU card. To me this is a scam because now you have 3 silicon in your system but only need 2 and your CPU is on idle and RAM use minimal as it has never really made that much difference in performance. INTEL Arrow Lake will change this because it is fully utilized and paired to take half the tasks of the paired INTEL ARC GPU. We end up with less hardware but same performance if not more with RAM counted in also. This does raise some questions like VRAM. Battlemage predicts to have 16 GB VRAM. With the new Arrow Lake can you add RAM as your graphics memory also if there is native link via PCI 5.0 between GPU/CPU is a question and how well. It almost looks like INTEL combo will have two hearts beating and syncronised together.
Food for thought.
Intel Core Ultra 9 285K
Physical
Socket: Intel Socket 1851
Foundry: Intel
Process Size: 7 nm
Package: FC-LGA18V
tJMax: 105°C
Cache
Cache L1: 112 KB (per core)
Cache L2: 3 MB (per core)
Cache L3: 36 MB (shared)
E-Core L1: 96 KB (per core)
E-Core L2: 4 MB (per module)
Processor
Market: Desktop
Production Status: unreleased
Release Date: Oct 2024
Part#: unknown
Core Config
# of Cores: 24
# of Threads: 24
Hybrid Cores: P-Cores: 8
E-Cores: 16
SMP # CPUs: 1
Integrated Graphics: Arc Xe2 Graphics 64EU
Performance
Frequency: 3.7 GHz
P-Core Turbo: 5.4 GHz
P-Core Boost 3.0: 5.6 GHz
Turbo Clock: up to 5.7 GHz
E-Core Frequency: 3.2 GHz
up to 4.6 GHz
Base Clock: 100 MHz
Multiplier: 37.0x
Multiplier Unlocked: Yes
TDP: 125 W
PL1: 125 W
PL2: 250 W
PL2 "Extreme": 295 W
Architecture
Codename: Arrow Lake-S
Generation: Ultra 9
(Arrow Lake)
Memory Support: DDR5
Depends on motherboard
Rated Speed: 5600 MT/s
Memory Bus: Dual-channel
ECC Memory: No
PCI-Express: Gen 5, 16 Lanes
(CPU only)
Chipset: Z890
Features
MMX
SSE
SSE2
SSE3
SSSE3
SSE4.2
AVX
AVX2
FMA3
SHA
EIST
Intel 64
XD bit
VT-x
VT-d
AES-NI
TSX
TXT
CLMUL
F16C
BMI1
BMI2
ABM
ADX
RdRand
VNNI
vPro
DLBoost 3.0
TBT 2.0
TBT 3.0
TVB
Notes
Intel Arc Xe2 Graphics Dynamic Frequency: 2000 MHz
"Lion Cove" P-Cores
"Skymont" E-Cores
The Intel Core Ultra 9 285K will be a desktop processor with 24 cores, that is expected to launch in October 2024. It is part of the Ultra 9 lineup, using
the Arrow Lake architecture with Socket 1851. Core Ultra 9 285K has 36 MB of L3 cache and operates at 3.7 GHz by default, but can boost up to 5.7 GHz,
depending on the workload. Intel is building the Core Ultra 9 285K on a 7 nm production process, the transistor count is unknown. You may freely adjust
the unlocked multiplier on Core Ultra 9 285K, which simplifies overclocking greatly, as you can easily dial in any overclocking frequency.
With a TDP of 125 W, the Core Ultra 9 285K consumes a lot of power, so good cooling is definitely needed. Intel's processor supports DDR5 memory with a
dual-channel interface. The highest officially supported memory speed is 5600 MT/s, but with overclocking (and the right memory modules) you can go even
higher. Actual memory technology support depends on the chosen motherboard, the processor itself supports multiple memory types, but most motherboards
have only one kind of slot. For communication with other components in the machine, Core Ultra 9 285K uses a PCI-Express Gen 5 connection. This processor
features the Arc Xe2 Graphics 64EU integrated graphics solution.
Hardware virtualization is available on the Core Ultra 9 285K, which greatly improves virtual machine performance. Additionally, IOMMU virtualization (PCI
passthrough) is supported, so that guest virtual machines may directly use host hardware. Programs using Advanced Vector Extensions (AVX) will run on this
processor, boosting performance for calculation-heavy applications. Besides AVX, Intel is including the newer AVX2 standard, too, but not AVX-512.
techpowerup
@paulboyce8537 shit did not cook bro...
Decided to wait for the new intel arrow lake cpu before i build a new gaming pc. I hope Intel will be more power efficient and cool again. Otherwise i'm forced to buy amd, never really liked that brand when it comes to cpu. Just personal preference.
NOooooo the 15 Intel gen fu**ing sucks. Screw all this optimizaton for work station desktop pc. I WANT TO GAME! Looks like im forced to buy AMD ZEN 5. ( Never liked AMD btw.)
i have a 13900k and primarily game. i also have a asus ryujin II and will tell you, under load I have never seen my cpu hit 70c its usually sitting around 58-61c under load. maybe I got a great chip but I dont know what all the complaints are about. this is a fantastic cpu and have had it since release.
That is not the case for the vast majority of people. Maybe you have amazing cooling and also hit a jackpot on silicon lottery but for most even with a big good air cooler that thing hits close to 90c and even with water-cooling it goes over 80c.
im excited im a intel fanboy even tho amd is better value for gaming
More interesting is whether it will be worth considering that the last two generations were defective from the factory. Intel themselves said that only i3 and i5 which are not K version are not affected. I've always liked Intel, but after this disgrace with the last generations, I wonder if I'll ever buy anything made by them again. I'm glad I didn't get a 13 or 14 generation CPU.
The only good thing I see it doing is to lower the prices of the 14th generation and maybe update my Kaby Lake, which I have left over at the moment. It would be an update on a whim more than anything.
Is Arrow Lake only for desktop or also mobile?
Lunar lake is mobile
ARL will be both desktop and mobile,
@@tech4free334 so its worth waiting? Will they also be efficient or pure power
intel will be on par to amd for about a month, then amd will be coming out with their x3d variants like 9800/9950x3d and amd will be ahead, again, probably for a good chunk of 2025 due to the 3d v-cache. no major upgrades until maybe late 2025 from intel when they'll come out with panther lake and powervia. amd's gonna be crushing intel this next year.
thanks for video
All Task Manager showing „NPU not used at all“, not even by Geekbench AI Benchmark😂😂
This is very interesting so their skipping calling the chips 15200/15900
So Oct 24th it will be released; I wonder how its power efficiency is when compared with AMD Ryzen cpus. Also the naming convention is weird, so Core 265 and up means more than 6 Performance cores; need 8 Performance cores for decent gaming I believe.
"The 265 has 8 P-cores and 12 E-cores. Most games typically use a single core; for example, GTA 6 will likely rely on one core.
@@n.d.n.e1805 What kind of BS is this? Games do need at least 6-8 cores to run well. I think what you meant to say is that games rely on a single CCD.
@@laszlozsurka8991 No game use only one core; you're forgetting that PCs also run background applications and system tasks. Some games even have additional apps that need to be running for the game to function properly. While the game itself may primarily utilize one core, it can occasionally use two. This isn't misinformation; if you had done just a little bit of research, you would understand this
@@n.d.n.e1805 "No game uses only one core; you're forgetting that PCs also run background applications and system tasks"
When tf did I say that games use only one core? Don't put arguments which I never stated.
Try running Hogwarts Legacy with only a 2 core CPU and see what happens. Most modern games do need at least 6 cores with 8 being the preferred. There's a reason why 8 P-core CPUs are faster than 6 P-core CPUs or lower.
Also GTA 6 is not even out yet so how do you even know how that "GTA 6 will rely on one core"?
@@laszlozsurka8991When did I ever say that you said that games only use one core? I said that. And yes, most games primarily utilize one core, maybe two at most. This is evident because most AAA games typically rely on one core, with two being the upper limit in many cases. Just do research it only takes 2 seconds. Also are you dumb? Of course, games will run better on a CPU with more cores but the game itself only utilizes one core. That's why CPU with higher single-core performance usually get better performance when it come to games
Arrow lake will support 10000mhz ram which is amazing for gaming
How old is this video? It shows charts that think it's still 2023???
Maximum TDP 250W. We know what that means. Separate PSU needed for it.
Based on Lunar Lake... I wonder what performance Apple's M4 could deliver with the 250 W TDP😈
it could probably run sims and that's about it
@@TheeAcid far from it... stay positive, u might witness one in your lifetime😉
I for one will be buying core 9 day one
when will 15th gen computers come on the market.
October 24th if I'm not mistaken. Although some leaks claim an October 10th release.
@@laszlozsurka8991 thanks, December it's perfect (winter sale)
Any 9700k boys still with me? 🤝
you dont think it was great timing that two weeks before the so called cpu issues that the US gov told us they were going to be investing 100 Billion dollars into Intel over the short term, then like 2 weeks later there is a story that craters the stock to 50% of its value. Why by high when you can create a story that can help give you a controlling interest. Pretty funny stock story, cause big plays always go down exactly like this.
I'll never forgive Intel
Pretty embarrassing if you ask me. A chip leader for many decades---no one was even remotely a competitor, and responsible for revolutionizing the PC industry but yet unable to manufacture their own chips in-house. So sad, Intel.
well at least their entire line up cpu will be cutting edge for first time since zen1 launch after arrow lake release
server(granite) intel3, mobile(lunar) tsmc3, dekstop tsmc3(arrow)
they will push further 2025 up for node advancement especially intel because they will have new toy from asml for first fab of high na euv machine
How? They're still spending billions to catch up to TSMC in technology; they just aren't there yet. The last few CEOs just sat on their hands and did nothing. So the new CEO has to catch up to TSMC. Also, Intel still produces a lot of the technology that the 15th generation will use; they are just buying the nodes from TSMC. So intel still make CPU
the xeon 6900p is on intel 3 and is the highest core count x86 on the planet AMD no longer leads and you are cluesless as to that fact
5718 Destiney Glens
i7-6700 user here. Will probably pick 15th gen intel for upgrade :))
Doesn't matter... Intel is getting rid of hyper-threadding which is a very dumb move.
Exactly. That will keep many people on last gen.
Unless they bring back L4 cache , i realy don't care
Crazy, the milk is putrid
Another chip, another socket... 🤦🏻♂️
Should I buy the arrow lake or 7950x 😂😂😂
Probably Arrow Lake, since the i7 leaks show that it outperforms the 9950X.
9800X3D, 9950X (underrated processor) or Arrow Lake :)
i9 will cost you 450-500w
No, it won't. the i9 285k TDP is 125 watts, and the maximum power consumption is 250 watts. The 9950X has a TDP of 170 watts, and at maximum, it can draw 250 watts.
Intel isn't buying nodes from TSMC for shits and giggles, it wants to COMPARE with real world HARD DATA how it's own 5nm/4nm/3nm/2nm nodes fair with TSMC in _volume_ production. The key is "volume" as any of the big three can do small "hand picked" batches easily. But Not only that, intel can run benchmarks to see if TSMC is truly better or worse than they're node-for-node from the SAME CPU!! but manufactured by TWO totally different companies! THEN, they'll figure out just HOW far behind or better they're in terms of volume efficiency and node superiority! Pretty sure they had to pay TSMC millions upfront to make those chips.😅😅
it's = it is
they're = they are
@@tim3172
Actually the reason behind Intel’s decision for TSCM to instead of their own nodes was cost. Intel 20a was supposed to be the first step before 18a that was supposed to be an iteration of 20a but 20a was not going to be sold to other companies it was meant to be just for Intel products. The cost to transition to a new architecture 18awas too high when they already have a mature 18a process (which means they consider their 18a good enough to be sold to other companies without the feedback of 20a) so they went with TSMC instead
no ht no buy
LAVA LAKE
61383 Willa Walk
what this video aspect ration is about
it is a shame!