Disclosure: Some links on this page are monetized by the Skimlinks, Amazon, Rakuten Advertising, and eBay, affiliate programs, and Liliputing may earn a commission if you make a purchase after clicking on those links. All prices are subject to change, and this article only reflects the prices available at time of publication.
Intel’s Alder Lake-N line of processors are low-cost, low-power chips designed for entry-level laptops, tablets, and mini PCs. But Intel says the Intel Core i3-N305 processor, which is the highest-performance Alder Lake-N chip to date, offers the kind of performance you’d expect from a Core i3 processor… thus the name.
Now one of the first laptops featuring that chip is available for purchase. The Acer Aspire 3 with a Core i3-N305 processor is now available from Acer. The company is selling a 3.1 pound model with a 14 inch display for $480 and a 3.8 pound version with a 15.6 inch screen for $500.
Both models feature 1920 x 1080 pixel displays, 8GB of single-channel LPDDR5 RAM, and 256GB of PCIe Gen 3 NVMe solid state storage. Both support WiFi 5 and Bluetooth 5.0, feature 720p webcams and stereo speakers, and have USB Type-C and Type-A ports, a headphone jack, and HDMI port. Only the 15.6 inch version has an Ethernet jack and numeric keypad though.
But it’s the processor that really makes these laptops stand out from other budget notebooks.
The Core i3-N305 chip is a 15-watt processor featuring 8 Gracemont CPU cores with support for speeds up to 3.8 GHz. There’s no support for hyperthreading, but multitasking performance should still be pretty decent with eight cores.
For graphics, the chip features an integrated Intel UHD GPU with 32 execution units and support for speeds up to 1.25 GHz. While you won’t get the same level of gaming or graphics performance from this chip that you’d see in a Core i5 or Core i7 U- or P-series processor with Iris Xe graphics, there GPU should be more than good enough to handle 4K video playback on an external display.
Overall, these are still budget laptops. But it’s nice to see that even budget laptops are getting a little more horsepower these days.
Looks like both models of the Acer Aspire 3 with the core i3-N305 is now available directly from their website at $350.
“Adler Lake-N” when you have old fabs and need products made by them.
I can’t wait for the new TSMC 3nm i3 chip benchmarks. Rumor is that low-end 3nm i3 beats high-end 5nm Ryzen by at least 25% in single threaded integer.
NEVER buy laptop with half sized arrow keys and cramped/merged keys (enter).
I’m writing this comment from such laptop 😐
And check Chinese brands like chuwi, teclast, bmax, kuu, jumper if you want
Intel N + metal-ish + fanless + IPS
“There’s no support for hyperthreading”
Because Gracemont – like Celeron and Atom before it – is single threaded. But good grief, ARM chips like the Apple M1 and M2 that everyone raves about is single threaded with no SMT/hyperthreading support either and no one mentions it.
“While you won’t get the same level of gaming”
This is horrible for gaming because there are no big cores. The i3-1215U or similar is way better for gaming.
What you’re missing is that these are basically Atom cores, but given more bandwidth and thermal headroom (so more performance, less efficiency). As a refresher, they can be slightly slower than a Cortex-A53, and slightly faster than a Cortex-A515. Let’s just say Cortex-A55 on average.
The “small” cores on the Apple M1 are much more powerful. They’re faster than the Cortex-A57 and on-average they’re equal to the Cortex-A73.
So Intel is WAY behind the market… but since they are the establishment and basically “own” the market it’s a difficult thing to expand upon.
For more context, the 5th-gen under-clocked Intel Skylake (2015 Core-M) was roughly equivalent to the Cortex-A73. While the 9th-gen over-clocked Intel Skylake (2020 Core-i7) was roughly equivalent to the Cortex-A78. Intel’s “new architecture” with the 10th-gen and 11th-gen refresh are more or less equal to the Cortex-X3. All of these are much much slower than Apple’s “big” core.
Intel’s newest “medium core” in the 12th-gen and 13th-gen aren’t very fast nor very efficient. They’re pretty similar to the old Skylake cores. While Intel’s newest “very big” core in the 12th-gen and it’s 13th-gen refresh are the industry leader in performance. However they’re very inefficient.
So the Apple M1 is BOTH more efficient AND faster than ALL processors produced by Apple, with the exception of the newest ones (a minority). This Celeron is a poor showing of Intel’s technology, and shows more how they’re slow to adapt in this fast-paced market. There’s also AMD and I haven’t touched upon that, but I can if you’re interested. As a disclaimer, while I’ve compared the cores as best as possible, we must remember the context of software when thinking about x86 vs ARM vs RISC-V etc etc.
Wow, so much smartness, it’s hard to say where to begin!
Maybe just the highlights.
– Alder Lake N didn’t get more thermal headroom than it’s predecessor Jasper Lake, so your statement about less efficiency is moot. In fact there are plenty of real world tests on the desktop and mobile Alder Lakes with the p-cores disabled showing Gracemont is a large step forwards in terms of efficiency.
– All the Arm comparisons are highly misleading. You base these on what? Got a preproduction sample to try out? Back then I did a test to compare Gemini Lake (Goldmont Plus) vs. an A53. On CPU heavy tasks like encoding it runs circles around the Arm. So I’d highly doubt Alder Lake with more cores on a smaller node does worse.
– There is no such thing as a 5th gen Skylake. 5th gen Cores were Broadwell or Hasswell-E, and Skylake didn’t release until the 6th gen Cores.
– There is and never was such a thing as an Intel “medium core”. There’s only the large Golden Cove resp. Raptor Cove and the small Gracemont. All Intel cpus are composed by a mix of these two.
– There is also no such thing as an Intel “very big” core, see above.
– The design of Gracemont is completely different from Skylake and there are great articles diving deep into the architecture if you’re interested. The point where both are similar is performance under certain workload. And that is actually the impressive part about Gracemont.
With all these points I wouldn’t even start throwing Apple into your mix, let alone AMD or some RISC-V implementations. Maybe next time get your facts straight first.
Only thing I’d agree in your little rant is the point that all current desktop CPUs are driven way out of their sweet spot on the voltage-frequency-curve. There’s a great test on AnandTech about this (https://www.anandtech.com/show/18693/the-amd-ryzen-9-7900-ryzen-7-7700-and-ryzen-5-5-7600-review-ryzen-7000-at-65-w-zen-4-efficiency) showing you can easily reduce power consumption by half with only little effect on the performance. But this concerns only the desktop parts, not the low power ones.
Wrong link! This is the right one:
https://www.anandtech.com/show/17641/lighter-touch-cpu-power-scaling-13900k-7950x
edit: it is Brad Linders fault. He wrote this device has Goldmont cores. I just looked it up on the Intel website and it is actually the Gracemont cores. Huge difference.
main response:
Well, uhh, you’re wrong.
These “N-cores” are evolved from “Atom cores” which themselves come from “Celeron cores” which also are descended from “Pentium cores”. It is the same lineage. You can better look them up from their codenames; Silvermont, Goldmont, Tremont.
Comparing the likes of the Intel Atom x7-8750, which was the last and fastest Atom processor, you can see they increased the bandwidth and thermal headroom for the next generations. That was the PRIMARY source of the performance uplift. You can observe this by comparing the above processor to the Intel N3710, same class, but at a much higher 6W TDP. Or the next N4200 processor, or the next one the N5000 and N5030. These newer processors ARE faster but they use up MORE electricity, and overall they are LESS efficient. This isn’t the only case, as Intel has been forced to make this move throughout their entire product stack due to having a “node disadvantage” and the competitors putting out impressive designs.
As a note, I made a comment about how the Intel 12th-gen P-processors which was labelled as “28W” actually behaves more-so like a typical 45W laptop processor. I thought so, when I found it unfortunately applied to their desktop products. Yes, the desktop parts actually do have great scaling but the voltage curve (or Watts) is beyond the limits of most laptops. That is also why ages ago I made a comment to tell people should get a Zen3 for most laptops out there, particularly the 6800u (no dGPU needed). I stated overall it is the better product, whilst the Intel chipset is superior for anyone looking to buy a Gaming Laptop, which disregards battery life, thickness, weight, noise, heat and comes with dedicated GPU and beefier cooling.
Coming back on topic, while Atom cores got larger/thirstier, they did also get some upgrades. Intel has two development teams, the mainline which works on Core-i and the secondary which works on Atom. Over the years, they have slowly added some of the maths and calculations from the more advanced Core-i predictions into Atom. Initially there wasn’t much integration, but Tremont was the first with a decent headway there.
The next iteration (Gracemont) is a major divergence, which actually has less in-common with the classic Atom lineup, and is almost a copy of the classic Core-i architecture. I postulate it had something to do with transferring developers from one team to the other. Gracemont is what they’re calling as the “Efficiency Cores”, is the full unification, or basically the conversion of Atom into Core-i. You are right, Gracemont is NOT the same as Skylake from their design point BUT they have many similarities. Even Intel themselves decided to not compare Gracemont to Tremont due to the massive differences, instead they compared it to Skylake. There’s a deep dive on it here:
https://chipsandcheese.com/2022/01/02/intels-tremont-atom-changes-course/
Side note; I didn’t know but apparently they released the Intel Core i5-L16G7, which was the first x86 hybrid processor in 2020. I did put a comment somewhere that Intel should have done this back in 2015-2017, by pairing 4x Cherry Trail with 2x Skylake cores. That would’ve been great for tablets, convertibles, and ultrabooks of the day. They would have secured their lead in the mobile space longer had they borrowed from ARM’s playbook in that same period.
The ARM comparisons may be misleading, but they’re not wrong. I already put a disclaimer that it is difficult to compare the two system types. There are many instances where x86 has an advantage because it has certain calculations built-in, not to mention the decades old software maturity. However, ARM is relatively new and nimble, which allows them to be more efficient. All said, ARM still has performance to back, and is not merely bound to smartphones. It has shown it can scale up remarkably to tablets, laptops, desktops, and servers.
Sorry, I mixed up the date for Core-M. We actually got Intel Core-M as early as Q1-2013 (as Intel-Y) and these processor types did get both faster and more efficient until the 2016. But it doesn’t change my point, which is that ARM was ONLY competitive when it came to efficiency… but that all changed with the Cortex-A72 and Cortex-A73 when they could compete on performance as well. Sure the Cortex-A53 was competitive in performance against the Intel Atoms, but what we’re talking about here is the mainstream Intel processors. In fact, it was ARM themselves which released their public announcement on “Ares Core” as matching Intel Skylake in IPC. Later it was Anandtech which corroborated and confirmed those findings on the Cortex-A76.
I think you’re being disingenuous here. The Intel “medium cores” is what I put down as to communicate more effectively. Sure they are Small-er but it’s not correct to call them “small cores”. I would call the Cortex-A35 as “tiny cores” and the title of “small cores” with the likes of the Cortex-A53 and Intel Atom. They are not that efficient either, usually when it comes to certain x86 tasks, but these are not like BIG.little or DynamIQ that people think of. Overall it is more accurate to call these “medium cores” if you actually look at the die sizes, they aren’t too much different to the likes of the Cortex-A78 and Apple’s Firestorm.
What I would classify as “big cores” again goes back to the die size and capabilities. This would be more like the Coffee Lake (i9900k), Zen3, Apple M1-Firestorm and Cortex-X1.
The Intel “very big cores” actually makes sense. Now I know you’re being unfair here. Again, look at their die sizes, they’re huge, my point is made. Don’t be pedantic.
Sorry about that. I’ve updated the article to reflect that we’re talking about Gracemont rather than Goldmont.
Acer is back to netbooks… sad.
Not entirely. Here are some of the other models they introduced during CES:
https://liliputing.com/acer-swift-go-14-and-16-are-compact-laptops-with-high-res-oled-displays/
https://liliputing.com/acer-introduces-ryzen-7000-powered-swift-14-go-laptop-with-a-90-hz-2-8k-oled-display/
And I didn’t cover their new gaming laptops, which are a bit big and heavy for Liliputing.