• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Intel Core Ultra 9 285k 'Arrow Lake' Discussion/News ("15th gen") on LGA-1851

I expect a slight improvement. perf. per watt yada yada.

tsmc 2nm (N2 blandly obvious naming) and Intel's supposedly adequately yielding 18A are both slated for 2025 production...tba releases. in the never ending battle for nerd supremacy in micro-computing devices. lunar lake seems good if youre into laptoppy things.
if they truly think theyll be shrinking at these rates year on year. mm yeh. im highly skeptical. zen 6 2nm could be 2025 or later . who knows. samsung have a 2nm process as well ive read.
angstroms. i have no idea what is possible but it seems there is room for shrinkage XD. barely 1 year between processes. seems as ridiculous as the market has become to me.
 
Last edited:
ooStevo said:
What's the actual plan for when the nm process comes to a halt? I'm genuinely interested. Just what are they going to do?
Click to expand...
that is anyone's guess.
the group efforts of r and d in materials science/engineering of silicon based semi-conductors. silicon was / is used because relative to other alternatives it is a cheaper/more expedient material to use for semi-conducting circuits. there are other mediums but it seems they are problematic, at least in terms of cost of production.

my knowledge is extremely limited. i just read online articles. :p
seems to revolve around lithography machines in printing the circuit traces, and solving various circuit factors on the substrates. the game is increasing density of transistors to have more of them as has been the case since the first silicon chip i guess..and more of them per unit area to keep power and thermals minimised.

"Robert Noyce invented the first monolithic integrated circuit chip at Fairchild Semiconductor in 1959. It was made from silicon, and was fabricated using Jean Hoerni's planar process and Mohamed Atalla's surface passivation process."

Invention of the integrated circuit - Wikipedia

en.wikipedia.org en.wikipedia.org

and the rest is history.

all the tricks of designing circuits themselves...always will be limited by power and thermals. so yeh ok you have supercomputers with millons of chips in arrays doing parallel processing. could probably get thousands or more fps playing some little kid's game on one of those? lol

the power of drawing sequential pictures rests with the graphics processors. massively parallel processing, direct die cooled Jensen chips. power hungry, expensive, and lucrative for resellers .lol.
and speaking of lucrative. not to mention advertising revenue. which youtubers can get from "views" something like 1-3 cents per view. so there is the craze of producing content 'for the views' right? say you got 1 million views . that is 10,000 simoleons/earth credits/inflation eroded dollars?. pretty handy?
 
Last edited:
ooStevo said:
What's the actual plan for when the nm process comes to a halt? I'm genuinely interested. Just what are they going to do?
Click to expand...
Optimisation.
Even if we had no new process nodes massive optimisations could be done to improve performance.

ASICs (Application Specific Integrated Circuits) are computer chips that combine several circuits on an SoC to allow custom programming for specific tasks.
As the cpus we use are so general purpose, and can do everything, they are not so great at doing specific tasks . Someone could for instance create a chip that is insanely good at a particular task such as coherency (rendering) however it would suck or not function at anything else.

Similarly with FPGA's ( field-programmable gate array) ASOC's (Adaptive system on chip) could be used to adaptively reprogram a chip to specific tasks before you launch to achieve maximum performance.

Asics are a one time affair whilst FPGA's/ASOC's can be reprogrammed depending on task.

Risc or similar reduced instruction set cpus, lots of our cpus have effectively wasted area set aside for things we don't use , however someone else does, a lot of that is old but kept for compatibility purposes.

So fear not, we won't reach a stage in our lifetimes when you can't get a better benchmark score.
 
Intel's 10nm, aka 'Intel 7' had a transistor density of around 100 MTr/mm2.

The 'Intel 3' fab. process has an unknown density. But it is probably worse than TSMC's N3 or N3E processes, because the transistor gate and interconnect pitches are higher:

3 nm process - Wikipedia

en.wikipedia.org en.wikipedia.org

The initial TSMC N3 fab. process has a transistor density of around 200 MTr/mm2.

It's a fair bet that the N3 process (presumably) used in Arrow Lake will be better than the 'Intel 3' process. But I doubt it's better than what they originally planned for their 2nm / 20a process.
 
Last edited:
graphene. - more expensive.

risc. thats amd.sort of they also use instructions which is more like cisc..partially. due to instructions that take longer than one cycle to execute.
and intel tried risc with itanium/EPIC. it didnt sell.

www.spiceworks.com

RISC vs. CISC: 20 Key Comparisons - Spiceworks

RISC and CISC architectures reveal distinct processor types. Discover more about RISC and CISC architectures and the key differences between them.
www.spiceworks.com www.spiceworks.com

so...? arm apple and amd all developed risc and intel has locked itself into a dead end. possibly. but thats up to them. various applications would have to be rewritten to utilise the risc architecture. i suppose. i have no idea really. just read bits and pieces.

www.extremetech.com

RISC vs. CISC Is the Wrong Lens for Comparing Modern x86, ARM CPUs

Try to investigate the differences between the x86 and ARM processor families (or x86 and the Apple M1), and you'll see the acronyms CISC and RISC. It's a common way to frame the discussion, but not a very useful one. Today, "RISC versus CISC" obscures more than it explains.
www.extremetech.com www.extremetech.com
quoted:
"The implementation-centric viewpoint looks to the ways modern CPUs have evolved since terms like RISC and CISC were invented and argues that we're working with an utterly outdated pair of categories."

"Here's an example. Today, both x86 and high-end ARM CPUs use out-of-order execution to improve CPU performance. Using silicon to re-order instructions on the fly for better execution efficiency is entirely at odds with the original design philosophy of RISC. Patterson and Ditzel advocated for a less complicated CPU capable of running at higher clock speeds. Other common features of modern ARM CPUs, like SIMD execution units and branch prediction, also didn't exist in 1981. The original goal of RISC was for all instructions to execute in a single cycle, and most ARM instructions conform to this rule, but the ARMv8 and ARMv9 ISAs contain instructions that take more than one clock cycle to execute. So do modern x86 CPUs."

im guessing that the architectural differences are significant but obviously there are cross compatabilties for various softwares that are either better suited to one architecture over the other. i dont delve too deeply into the detail. which is why im practically useless at understanding these things at a programming level . i find it really boring. :)
 
Last edited:
I remember that 16-bit consoles and 32-bit consoles used RISC instructions,and they were very efficient because of it. I wonder if x86 instruction set CPUs would be outperformed by RISC architecture. I'm probably just talking ********,but it's an interesting question all the same.
 
ooStevo said:
I remember that 16-bit consoles and 32-bit consoles used RISC instructions,and they were very efficient because of it. I wonder if x86 instruction set CPUs would be outperformed by RISC architecture. I'm probably just talking ********,but it's an interesting question all the same.
Click to expand...

It’s hard to say. The two big questions would be how core scaling effects efficiency and of course the operating system leveraging the performance of hardware.

RISC V 64 support was added to Linux last year IIRC.
 
www.electropages.com

Intel Terminates RISC-V Development Project: What Happened?

Find out more about the surprising decision made by Intel to drop its RISC-V development project and explore the potential ramifications on the RISC-V processor market and the future of CPU architecture.
www.electropages.com www.electropages.com
article from 2023
another thing amd has.

which is a market differentiator in the server market.

quote:
"At the same time, future server designs will likely adopt RISC devices due to their higher core densities and lower energy consumption, meaning that Intel would be supporting the tools that could eventually upset its server market."

it's already upset.
lower cost and higher efficiency.

..which goes back to 'intel as a foundry'.
and using tsmc for packaged cpu's. teehee.
==
and that aspect of 'more of the same' versus new and different. that was mentioned XD . if you want something different to intel then obviously you dont buy intel.
i prefer to have both. but also im just comfortable with the familiar, which is intel and windoze basically.
 
Last edited:
*Rumour Mill*

videocardz.com

Intel Core Ultra 9 285K, 7 265K(F) and 5 245K(F) listed by first retailer - VideoCardz.com

Intel Arrow Lake-S coming, UK retailer lists Core Ultra 200K series ahead of announcement A clear sign that Intel is just about to announce the new desktop CPU series. UK-based retailer LambdaTek has listed the upcoming Core Ultra 200K processors, totaling five models. As expected, Intel is...
videocardz.com videocardz.com

Intel Arrow Lake-S Processors:
  • Core Ultra 9 285K 24 cores 5.7GHz LGA1851 (BX80768285K): £471.83, £566.20 inc VAT
  • Core Ultra 7 265K 20 cores 5.5GHz LGA1851 (BX80768265K): £327.35, £392.82 inc VAT
  • Core Ultra 7 265KF 20 cores 5.5GHzLGA1851 (BX80768265KF): £314.60, £377.52 inc VAT
  • Core Ultra 5 245K 14 cores 5.2GHz LGA1851 (BX80768245K): £250.86, £301.03 inc VAT
  • Core Ultra 5 245KF 14 cores 5.2GHzLGA1851 (BX80768245KF): £238.11, £285.73 inc VAT
 
My dad is still looking to build a new PC, so I'll see if anything cheaper becomes available.

Or, maybe I'll recommend the 245K.
 
Last edited:
I suppose there’s still a chance that the 20a fab. process could be used for a 285KS / Limited edition chip, maybe clocked up to 6ghz? That would allow for a late release, sometime next year.

I’d guess that Intel originally made plans to fabricate Arrow Lake on both internal and external nodes.

But, these chips are rarely worth it. But it would still have the prestige of being the first mobile or desktop 2nm class CPU.
 
Last edited:
RSR said:
*Rumour Mill*

videocardz.com

Intel Core Ultra 9 285K, 7 265K(F) and 5 245K(F) listed by first retailer - VideoCardz.com

Intel Arrow Lake-S coming, UK retailer lists Core Ultra 200K series ahead of announcement A clear sign that Intel is just about to announce the new desktop CPU series. UK-based retailer LambdaTek has listed the upcoming Core Ultra 200K processors, totaling five models. As expected, Intel is...
videocardz.com videocardz.com

Intel Arrow Lake-S Processors:
  • Core Ultra 9 285K 24 cores 5.7GHz LGA1851 (BX80768285K): £471.83, £566.20 inc VAT
  • Core Ultra 7 265K 20 cores 5.5GHz LGA1851 (BX80768265K): £327.35, £392.82 inc VAT
  • Core Ultra 7 265KF 20 cores 5.5GHzLGA1851 (BX80768265KF): £314.60, £377.52 inc VAT
  • Core Ultra 5 245K 14 cores 5.2GHz LGA1851 (BX80768245K): £250.86, £301.03 inc VAT
  • Core Ultra 5 245KF 14 cores 5.2GHzLGA1851 (BX80768245KF): £238.11, £285.73 inc VAT
Click to expand...

Good value if true but Intel always advertise the bulk tray try price, i have a feeling they will be 10 to 20% more.
 
You must log in or register to reply here.
Back
Top Bottom