moores law aint dead - /g/ (#105983962) [Archived: 824 hours ago]

Anonymous
7/22/2025, 4:10:04 AM No.105983962
Imec2Droadmap630-800x600-c-default
Imec2Droadmap630-800x600-c-default
md5: 66c42b8ed2f0b1ae37ed314a0f66d3e1🔍
https://www.youtube.com/watch?v=0wRvbIaTUQw&embeds_referring_euri=https%3A%2F%2Fwww.reddit.com%2F

were back baby
Replies: >>105983967 >>105983998 >>105984021
Anonymous
7/22/2025, 4:10:54 AM No.105983967
MV5BOWM1OGNkODQtZjI3Ny00ZWJhLTg1YmItNmVlYmJiMDY1YzNkXkEyXkFqcGc@._V1_
>>105983962 (OP)
>reddit
GET THE FUCK OUT!!!!!!!!!!!! REE
Anonymous
7/22/2025, 4:16:33 AM No.105983998
>>105983962 (OP)
>how to get beyond 0.2 nanometers
Change how you measure your shit and then say your stuff is smaller than the other company
Replies: >>105984026
Anonymous
7/22/2025, 4:20:17 AM No.105984021
>>105983962 (OP)
Moore law is about transistor density not about scaling down transistors.
Planar is dead, 2.5D and stacking will be the future -if people can fix the throughput issue-
Replies: >>105984033
Anonymous
7/22/2025, 4:20:52 AM No.105984026
>>105983998
naming kept the same as scaling but with finfets is AS IF it was 2 nm
Anonymous
7/22/2025, 4:21:53 AM No.105984033
>>105984021
2.5D will def be the future. the question is WHEN? 2030? 2035? its exciting to know that the computers of 2030s will be like 3x faster than the best of the best fitfet or even nanosheet
Replies: >>105984076
Anonymous
7/22/2025, 4:22:01 AM No.105984036
square wafer when?
Anonymous
7/22/2025, 4:29:36 AM No.105984076
>>105984033
WDYM? 2.5D has been a thing since the early 2010s, finfet, GAA is 2.5D, they using deeper transistor to reduce their footprint.
Replies: >>105984080
Anonymous
7/22/2025, 4:30:41 AM No.105984080
>>105984076
sorry i meant true 2d
Replies: >>105984112
Anonymous
7/22/2025, 4:37:46 AM No.105984112
>>105984080
When a company solves the problem of cooling and interconnecting at micrometer scale, NAND memories are deeply stacked ('hard' 2.5D) but just because they have a very simple connecting structure like lift slab construction of normal buildings
Replies: >>105984359
Anonymous
7/22/2025, 5:28:36 AM No.105984359
>>105984112
youre right its the reason weve been stuck at 5.5 Ghz for years now. we really need someone to discover some new material that doesnt have the thermal issues that current tech has. Its the only way we will go beyond 6ghz.
Replies: >>105984375
Anonymous
7/22/2025, 5:30:30 AM No.105984375
>>105984359
No consumer-grade CPUs reliably exceed 7 GHz out of the box in 2025. The highest is the Intel Core i9-14900KS, reaching 6.2 GHz with Turbo Boost Max 3.0, but overclocking to 7 GHz+ is possible with extreme cooling (e.g., liquid nitrogen) and high voltage (1.83V), as seen in enthusiast experiments. Such overclocks risk instability and hardware damage.

AMD Ryzen 10000 series (Zen 6, expected ~2026) is rumored to target 7 GHz, but this is unconfirmed speculation. For now, 6.0–6.2 GHz is the consumer peak.