Sign in with
Sign up | Sign in
Your question

How is smaller architecture more helpful

Last response: in CPUs
Share
December 13, 2012 10:03:51 AM

How does having small architecture in CPUs be helpful? For example, people say the 22nm architecture Intel Ivy Bridge CPUs are better than the older 32nm architecture ones. What role does the architecture actually have in better or worse performance? Thanks.
a b à CPUs
December 13, 2012 10:50:13 AM

The smaller a chip gets in the Nm scale means you can pack more transistors on the cpu die formed on the silicone, so the cpu would or should get better in processing power, it also serves in some ways to reduce the heat output, and requires lower power to run say the cpu, reducing heat build up. Normal this is a good thing. untill you get to a point where the gap between the electronic pathways in the cpu and transistors become so small that electro magnetic leakage can start to cause problems, and power problems. Look up Moors law.
December 13, 2012 11:01:31 AM

For Intel's Tick-Tock model, the first move from one lithography size to the next generally has very little performance-per-clock difference in the instruction set. However, smaller CPUs allow for less power use, and will let the creator decide between shrinking die size or adding more cache/GPU compute nodes/etc to make up for the extra space they have. With overall-smaller die size reduction, they'll create more chips per wafer which should reduce the cost per chip. The reduced power draw will allow for more energy efficient CPUs and higher clock speeds within the same power envelope of the previous generation.

In short, smaller CPUs are generally cheaper and faster, while newer CPUs will gain incremental advancements in the instruction set and additional features, such as AES encryption acceleration. And in the world of highly integrated CPUs nearing system-on-a-chip, you'll have updated GPUs, memory controllers, PCI-express controllers, etc.

But again, with Intel's Tick-Tock method of first trying out a smaller process size on a known CPU design before upgrading the architecture to really knock our socks off, we've generally been seeing the first move to the new process that current software is within +/-5% of the speed than on the previous generation CPU when compared at the same clock-speed, while being more power efficient though.
a b à CPUs
December 13, 2012 11:34:55 AM

The Stealthinator said:
How does having small architecture in CPUs be helpful? For example, people say the 22nm architecture Intel Ivy Bridge CPUs are better than the older 32nm architecture ones. What role does the architecture actually have in better or worse performance? Thanks.

That's not just the architecture that the die size at play there too.
- Since the architecture has been mostly addressed, I'll just mention why 22nm is better than 32nm even without any changes.

Each die shrink is 1/SqRt(2) of the size before it.
- Over two axises this permits double the transistors per square millimeter.
- It reduces power consumption
- It reduces heat
- It increases clock speed so long as the surface area to contact ratio is large enough to dissipate the heat produced
- The same 500 million transistor CPU can now be mass produced at half the cost, so the profits are larger until your competitor catches up, and then you just reduce the margins while maintaining profitability.

The micro-architecture is more like comparing an Intel Atom series CPU to an Intel Core i7 series CPU...
- Wikipedia 'Intel Atom' and 'Intel Core i7' for a better idea.

The difference between the 1st and 2nd generation Core i# processors was quite large.
The difference between the 2nd and 3rd generation Core i# processors was only marginal, and sometimes the 2000 series outperforms the 3000 series at the same clock speed.
!