• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Intel Intros Atom N570 Dual-Core Processor

Cheeseball said:
Not much of a difference from the old Atom 330, aside from:

1. Both cores being on-die instead of combined.
2. Integrated DDR3-667 controller.
3. Integrated GPU (Intel HD, not HD 3000/2000)
4. Integrated PCI Express controller.

An AMD Bobcat Fusion processor will eat this alive though. Probably cheaper too.
Click to expand...

Yes not much difference except the much lower CPU+chipset power usage.:rolleyes:
 
Stupid....intel... my mobile has better graphics capability
 
My dads pc has an Atom 330, 2gb ddr2, nVidia ION,..and it runs Windows 7 smoothly. the cpu has a score of 3.3
 
in short time will come a bobcat with 4 cores :D
 
There is a lot of Atom bashing going on here. Atom isn't bad, esp. the lower-power envelope models. And remember that Atom has been out a long time, and AMD's is a new release, so, following Moore's law, should be twice the performance or half the power consumption. But it isn't.

And everyone saying E350 smashes the Atom. Absolutely not! Have you read the reviews? Do you own these devices? The E350 CPU is not a lot faster. IIRC between 0%-30% depending on application, an average around 15%. And for multithreaded apps, 0% or -ve. It is the same order of magnitude, same ballpark performance, and nothing to get excited about. NOW, the GPU half of the E350 is better than Intel's offering, without a doubt. But so is ION and ION2 that pairs with Atom. And you need to ask yourself what the "Atom machines" are being used for. They certainly are not workstations or games platforms! For Office, productivity, small servers, or NAS-type devices, they are quite fine. In all of these examples who gives a hoot about 3D performance?

Remember: usage scenario. Atom is fine for what it was designed for. Low power terminals, x86 NAS, and netbooks a-la 2009.

I agree for small laptops today the A350 is far better... because netbooks have become laptop replacements and we want more video processing power for flash, basic 3d, etc.
 
Last edited:
Atoms are pretty decent, but their main problem is the fact that its graphics dept is absolutely shit. Which is why people thought Bobcat will cream Atom. If we are allowed to bolt on a 5450 to Atom, then suddenly Bobcat looks rather average.
 
Are we sure this has 4-threads?

Untitled530.jpg
 
There is a lot of info missing from what i see, so it's possible that the nuber of threads is wrong there, because the slower one has the same characteristics and 4 threads apart from clock speed.
 
Yeah. There is no way that would release a chip with the same naming scheme that clearly positions a chip to be an updated version of another, and cripple it by making any drastic changes like that.

This is little more than a clock rate update.
 
TheLaughingMan said:
Yeah. There is no way that would release a chip with the same naming scheme that clearly positions a chip to be an updated version of another, and cripple it by making any drastic changes like that.
Click to expand...

So your saying that Intel made a mistake? I don't understand why Intel would make this chip to begin with. They already have a chip that is the same speed (D510) and a faster chip (D525).
 
HossHuge said:
So your saying that Intel made a mistake? I don't understand why Intel would make this chip to begin with. They already have a chip that is the same speed (D510) and a faster chip (D525).
Click to expand...

those are higher wattage chips
 
my acer one has a n550 and i didn't even know it had 4 threadsand it looks like an overclocked 550? must be a mistake on the threads right
 
Ok, so it says about DDR3-667 that if you use a (more) standard DDR3-1066 it will work with a lowered latency.
So, if I change the SO-DIMM by a DDR3-1066 with CAS 7, will it lower the latency proportionally down to something over 4? Will it be automatical?
Or I'll end up, if I don't or can't change timings with a dimm woking at 667 MHz and latency 7 which, of course, will be slower than a native 667 dimm with CAS 5?
Thanks!
 
It all depends on the SO-DIMMS SPD JEDEC settings. You need to look at the specific SO-DIMM specification to see what timings it will use for CAS when at 667.

I would advise you to NOT worry about it though. CAS timings on memory make very little real world performance difference. On a PURE MEMORY benchmark, you might see 5% difference. But run an application or a game or something "real" and you will see less than 1% if not 0.1%.

So just buy according to your budget, dont pay more for lower CAS timings on an Atom system.
 
Thanks for the info.
I didn't know that it the cas at different frequencies was specified for a given SO-DIMM.
In fact my worry is not about buying a module with smaller CAS, even if it's more expensive.
I was worried about buying a more expensive 1066 Mhz CAS 7 module when in fact I can get better performance with a (hopefully) cheaper 677 MHz CAS 5 one.
Because... this diverges a bit from the main subject of this thread, but I wonder if a module operating at 677 MHz with CAS 5 is equivalent to one operating at 1066 with CAS 7 and to one operating at 1333 with CAS 9, for example, as my short knowledge about this makes me think that, if an access takes a given number of cycles, you win nothing with a higher frequency if you also need proportionally more cycles for one access. If so, then the speed of a memory module could be roughly calculated as its frequency divided by it's latency in cycles.
Am I too wrong?
 
You must log in or register to reply here.
Back
Top