IBM has just upended the chip world with 2nm production the US is back!!!

PEB

Sunflower in support of Ukraine
@Steel
Joined
Jan 20, 2004
Messages
30,382
Reaction score
19,268
TMSC who IBM is crushing it as I said before GlobalFoundries and Samsung got to 10nm thanks to IBM. IBM has been quiet but now is back in a huge way. Back in the 90's they where already doing advance design ahead of everyone then disappeared after their deals with TMSC and such. Now they are back again with 2nm production this is like 5 to 8 years ahead of everyone. They are also partner with Intel look out AMD it may explain the 20 billion dollar investment by Intel.

 
<Huh2>

time to market is ~4-5 years.
That's a bit optimistic because they are only at 7nm with X86 and 5nm with ARM their is timing to coincide with Intel 20 billion dollar plant investment and IBM development of 2nm chip production I was surprised there was a company making UV machines in Conn. There is a lot of reasons to see Intel choice of two factories.
 
Last edited:
That's a bit optimistic because they are only at 7nm with X86 and 5nm with ARM their is timing to coincide with Intel 20 billion dollar plant investment and IBM development of 2nm chip production I was surprised there was a company making UV machines in Conn. There is a lot of reasons to see Intel choice of two factories.

wat

tsmc has 5nm already well past mass production.
3nm (finfet) is set for 2023. (edit: technically, slated for 2nd half of 2022)
2nm (gaafet?) is tracking for 2024.

even samsung's been on track for 2nm gaafet for a while. (iirc, 3nm in 2022)

i have no idea where you're getting 5 years ("ahead of everyone") from, let alone 8.
 
Last edited:
Everyone is chasing Apple, now, and TSMC manufactures their M chips.

I'll believe IBM has reclaimed the lead when I see it.
 
2nm? Pfft, what a hunk of junk. Call me back when they have some planck length chips.
 
Everyone is chasing Apple, now, and TSMC manufactures their M chips.

I'll believe IBM has reclaimed the lead when I see it.
"
GlobalFoundries has begun baking chips at Fab 8 in upstate New York, and is doing so for none other than semiconductor process partner IBM.

Fab 8, when it is fully ramped, will be able to kick out 60,000 wafers per month from its 300,000 square feet of space and etch them with 32 nanometer HKMG processes. That ramp is now expected in the second half of this year, a bit faster than the original early 2013 date from a year and a half ago when the fab upgrade plan was announced by ATIC."

https://www.theregister.com/2012/01/10/ibm_globalfoundries_fab_8_ramp/

"
IBM says it has built a computer chip whose smallest features are about seven nanometers wide, meaning it could provide about four times the capacity of today's chips. Basically, it packs far more transistors into the same space, and that means data can travel more quickly between those transistors.

The announcement, made on Thursday, was hailed as a breakthrough, evidence that semiconductors will continue to improve in the years to come. But Patrick Moorhead, an analyst who closely follows the chip business,"

https://www.google.com/amp/s/www.wired.com/2015/07/ibm-seven-nanometer-chip/amp
 
"
GlobalFoundries has begun baking chips at Fab 8 in upstate New York, and is doing so for none other than semiconductor process partner IBM.

Fab 8, when it is fully ramped, will be able to kick out 60,000 wafers per month from its 300,000 square feet of space and etch them with 32 nanometer HKMG processes. That ramp is now expected in the second half of this year, a bit faster than the original early 2013 date from a year and a half ago when the fab upgrade plan was announced by ATIC."

https://www.theregister.com/2012/01/10/ibm_globalfoundries_fab_8_ramp/

"
IBM says it has built a computer chip whose smallest features are about seven nanometers wide, meaning it could provide about four times the capacity of today's chips. Basically, it packs far more transistors into the same space, and that means data can travel more quickly between those transistors.

The announcement, made on Thursday, was hailed as a breakthrough, evidence that semiconductors will continue to improve in the years to come. But Patrick Moorhead, an analyst who closely follows the chip business,"

https://www.google.com/amp/s/www.wired.com/2015/07/ibm-seven-nanometer-chip/amp
WTF? Why are you quoting articles from 2012 and 2015?

I'm talking about showing me the money...not headline hype or promises of future products. I'm talking about a chip you can put in my hands today. I'm talking about a chip with market samples in benchmark reviews.
 
WTF? Why are you quoting articles from 2012 and 2015?

I'm talking about showing me the money...not headline hype or promises of future products. I'm talking about a chip you can put in my hands today. I'm talking about a chip with market samples in benchmark reviews.
I am showing IBM history of advancing chip architecture an development. They are hardly new at advancing chip design. In fact a sizable amount of technology they had a key part. There is no question they have a technical advantage.

"
IBM just popped up with an exciting breakthrough in transistor technology—the world's first 2nm chips. The news comes straight out of IBM's semiconductor research facility Albany, New York, and means great things for the likes of the company's partners, such as Samsung, and not least its newest partner: Intel.


In fact, it might just prove immensely valuable in Intel's race to catch up with rivals TSMC.

With 50 billion transistors in a chip the "size of a fingernail," IBM claims the research chip's "tiniest components are smaller than a strand of DNA." With it, the company has apparently surpassed its previous 5nm chips in terms of size, and it's even projected to have a 45% performance improvement over 7nm chips, as well as use 75% less energy (via Tom's Hardware).


IBM's $3 billion 'Seven Nanometers and Beyond' research program is moving us out of the age of the FinFET designs of yesteryear, which were fine back when 22nm was the norm, but tend to throw up issues when you get down to the ridiculous scales of today's chips. These new nanosheet designs are part of the gate-all-around design that means we'll no longer have to deal with FinFETs leaking current."

https://www.google.com/amp/s/www.pcgamer.com/amp/ibm-breakthrough-2nm-transistor-research-intel-cpu/

Oh an the US needs to lead the way to sub 1nm chip design possibly the most expensive race in the history of high tech.
 
Last edited:
I am showing IBM history of advancing chip architecture an development. They are hardly new at advancing chip design. In fact a sizable amount of technology they had a key part. There is no question they have a technical advantage.

"
IBM just popped up with an exciting breakthrough in transistor technology—the world's first 2nm chips. The news comes straight out of IBM's semiconductor research facility Albany, New York, and means great things for the likes of the company's partners, such as Samsung, and not least its newest partner: Intel.


In fact, it might just prove immensely valuable in Intel's race to catch up with rivals TSMC.

With 50 billion transistors in a chip the "size of a fingernail," IBM claims the research chip's "tiniest components are smaller than a strand of DNA." With it, the company has apparently surpassed its previous 5nm chips in terms of size, and it's even projected to have a 45% performance improvement over 7nm chips, as well as use 75% less energy (via Tom's Hardware).


IBM's $3 billion 'Seven Nanometers and Beyond' research program is moving us out of the age of the FinFET designs of yesteryear, which were fine back when 22nm was the norm, but tend to throw up issues when you get down to the ridiculous scales of today's chips. These new nanosheet designs are part of the gate-all-around design that means we'll no longer have to deal with FinFETs leaking current."

https://www.google.com/amp/s/www.pcgamer.com/amp/ibm-breakthrough-2nm-transistor-research-intel-cpu/

Oh an the US needs to lead the way to sub 1nm chip design possibly the most expensive race in the history of high tech.

...so you didn't listen to the video you linked, you don't know what the current state of tech is, and you reallllllllllllllllly want to believe the usa is the best at chip fabbing.
 
I am showing IBM history of advancing chip architecture an development. They are hardly new at advancing chip design. In fact a sizable amount of technology they had a key part. There is no question they have a technical advantage.

"
IBM just popped up with an exciting breakthrough in transistor technology—the world's first 2nm chips. The news comes straight out of IBM's semiconductor research facility Albany, New York, and means great things for the likes of the company's partners, such as Samsung, and not least its newest partner: Intel.


In fact, it might just prove immensely valuable in Intel's race to catch up with rivals TSMC.

With 50 billion transistors in a chip the "size of a fingernail," IBM claims the research chip's "tiniest components are smaller than a strand of DNA." With it, the company has apparently surpassed its previous 5nm chips in terms of size, and it's even projected to have a 45% performance improvement over 7nm chips, as well as use 75% less energy (via Tom's Hardware).


IBM's $3 billion 'Seven Nanometers and Beyond' research program is moving us out of the age of the FinFET designs of yesteryear, which were fine back when 22nm was the norm, but tend to throw up issues when you get down to the ridiculous scales of today's chips. These new nanosheet designs are part of the gate-all-around design that means we'll no longer have to deal with FinFETs leaking current."

https://www.google.com/amp/s/www.pcgamer.com/amp/ibm-breakthrough-2nm-transistor-research-intel-cpu/

Oh an the US needs to lead the way to sub 1nm chip design possibly the most expensive race in the history of high tech.

Sub 1nm? Wouldn’t that be literally impossible due to the laws of physics? Here’s a video from 2013 of a university professor explaining how quantum tunnelling means that there is a limit to how small these things can go and he estimates we’ll hit that limit by 2025.

 
TS got handed his ass back oh boy

So who is the current leader TSMC?
 
Can you tell me how this matters in real life ? How this affects me ?
All I know is more processors (?) on chip good.
Generally speaking (this is just an attempt at a really basic recap of stuff I've read), a smaller CPU die is more efficient and takes less power. That means less heat, which means you can push the CPU further.

That's my basic understanding, at least.
 
TS got handed his ass back oh boy

So who is the current leader TSMC?

yeah. and it's not even close. korea, just today, announced some gigantic initiative ($450B or something?) to expand and take the lead, but i dunno.

samsung is clearly #2 behind tsmc, but they seem to have yield issues with their most advanced nodes. their (samsung, not ibm [although, that could make things more interesting] ) bet on gaafet 2nm before tsmc (ie: risky jumps vs steady steps) could be the biggest factor.
 
Sub 1nm? Wouldn’t that be literally impossible due to the laws of physics? Here’s a video from 2013 of a university professor explaining how quantum tunnelling means that there is a limit to how small these things can go and he estimates we’ll hit that limit by 2025.


I am not completely saying it is possible but different uses and atom placement can make 14 atom transistor or .5nm transistor can they use that in a actual chip getting into quantum tunneling that runs into problems no doubt.

"
We are always being told that Moore’s Law is breaking, but increasingly today it’s breaking not because we’re at the technical limits, for example researchers recently demonstrated their latest 1nm transistor and virus sized computer architectures, but because we’re at the economic limit because every time the chip manufacturers step down a size it costs them anywhere between $15Bn to $30Bn in new investment.

"

"
The team created a two-terminal transistor with a diameter of about 0.5 nanometers and a core that consisted of just 14 atoms, and the device can reliably switch from insulator to conductor when charge is added or removed, one electron at a time, something known as a “Current blockade.”

"

"
Now, as a follow up step the researchers plan to design better and improved molecular cluster systems with better electrical performance, such as higher on/off current ratio and different accessible states, and increase the number of atoms in the cluster core, while maintaining the atomic precision and uniformity of the compound.

Elsewhere other studies into creating atomic scale and molecular scale transistors have used Quantum Dots, nanoscale semiconductor crystals of nanometre dimensions with distinctive conductive properties that are determined by its size, to produce similar effects, but the dots are much larger and aren’t uniform in size making it a problem to reproduce them in a repeatable manner – something that’s crucial if the technology is ever going to be mass produced.

All this said though the ultimate size reduction for any transistor would ultimately be transistors made up of just a single atom, which would mean they’re less than 0.1nm in size, but at the moment using current technology they’d require ultra-cold temperatures of -196C in order to cancel out some of the electronic “leakage” effects which, at the moment, cripple the technology.

"

https://www.fanaticalfuturist.com/2...s-to-build-the-worlds-first-0-5nm-transistor/
 
Back
Top