Intel What Are You DOING! Ice Lake Lineup Is A Huge Mess!
Well you all probably guessed that there was no way Intel was going to take all of those AMD product launches lying down, right? Regrettably, if you’re here to learn more about their new upcoming desktop processors, look away right now because I’m here to talk about mobile CPUs again. That means the Ice Lake Architecture, Gen11 graphics, and 10nm manufacturing process are sort of almost here. Supposedly you will be able to buy them starting November, and honestly if you thought Intel’s lineup was confusing before just wait till you see what they’ve cooked up.
Now to set the stage, let’s do a little test. Above are four of Intel’s newest processors, and pause if you have too because I’d like you to write them down in order of the highest performance to the lowest and we’ll discuss about them later in the video. With that done, I’m going to try to make some sense of Intel’s lineup.
Alright, so let’s get right into this with a quick explanation on how Intel and AMD usually launch their processors. This will give you a rough idea about why we haven’t seen Intel respond to the Ryzen 3000 series yet. Basically, AMD launches their desktop processors first and notebook processors last. What that means is their notebook processors are a generation behind their desktop parts. Intel does things a lot differently. Unlike AMD, Intel mobile processors are at least a generation ahead of what desktop users get. Remember, the Coffee Lake CPUs on the desktop actually use an architecture that was launched almost two years ago.
So I guess that brings me to Ice Lake, and if you thought Intel’s last generation lineup was confusing, this one is absolutely ridiculous. I can’t believe this is happening again. Intel… what are you really trying to do to us? I mean is what we’re dealing with. So this is Intel’s version of an alphabet soup, a mishmash of letters and numbers. This processor is called Core i7-1068G7 and I’ll try to explain what that all means. So the i7, i5, and i3 are supposed to tell you the general performance level, and that’s pretty well known, but after that things take a turn for the worst. So the 10 part of this number means that it’s Intel’s 10th generation architecture, but there are some exceptions and I’ll mention that a bit later. The next two numbers tell you the relative performance versus other parts in Intel’s lineup, for example a 68 will perform better than a 65 right? But to add a bit more complication to this, any CPU ending in 5 or 8 will belong to the U series family, while a name ending in 0 puts it into an ultra-low voltage Y series lineup. Finally, any CPU with a G in its name will have the new Gen11 graphics architecture. The G7 has 64 execution units, G4 has 48, and the G1 has 32 EUs. Remember, these new graphics cores are pretty important to Intel. Supposedly, a Gen11 GPU with 64 EUs can perform equally to an NVIDIA MX150, which is good news for Intel, but bad news for NVIDIA.
We actually looked at that GPU in Razor’s Blade Stealth and considering that notebook’s form factor it was pretty impressive. In fact, if you’re interested in a full performance analysis, you can check it out right over here. Now imagine that kind of performance from an even thinner and lighter system. Well, that’s what Intel is trying to accomplish.
So let’s dive into this confusing lineup starting with the U series models, which actually don’t have the letter U anywhere in their name. The lineup includes six processors from i7 to i3, with the i7 and i5 have having four cores and eight threads, while the i3 has two cores and four threads. Other than clock speeds the main difference between i7 and i5 CPUs is the amount of cache. As for the integrated graphics, well that’s all over the place and you can start seeing why this lineup is so damn confusing. While the i7 has the full 64 Execution Unit layout, the i5-1035 series can have either 64, 48, or 32 EUs. The i3 only has about 32 EUs. Oh, and did you notice that 1035 CPU? Yeah, there are three of them. All featuring different graphics configs and clock speeds. Another thing you’ll notice is how low the clock speeds are, even though the TDP values haven’t improved over the previous generation. If we take that 28W i7 out of the picture, the base frequencies barely stay about above 1GHz. The single and all-core Turbo speeds are a bit better, but in general they’re between 300MHz and 900MHz lower than the previous generation.
So what does that mean? It looks like Intel needs to rely on architectural improvements to carry Ice Lake instead of clock speeds. Personally, I think there will be a minor performance difference between the 9th and 10th Gen CPUs in regular workloads. But I might be wrong, you never know. And I’m going to be really honest here, I think Intel might be in a bit of trouble because it looks like 10nm can’t hit the same frequencies as the older 14nm process. You see with AMD and their launch of the Ryzen 3000 series CPUs, they were able to switch over to 7nm while offering better frequencies. Now sure that’s on the desktop site, but at least it’s worth mentioning.
All right, so back to the lineup with the Y series. Intel’s most efficient low-power notebook CPUs are popped into this lineup and there’s some exciting stuff here, but also some bad news. It’s great to see four core/eight thread ultra-efficient CPUs, but TDPs have climbed significantly from 5W to 7W on Amber Lake Y to 9W to 12W on Ice Lake Y. Sure, you’re getting a lot of GPU horsepower, but I’m interested in knowing if you see this as a benefit or sacrifice. Now about those base clocks, well it’s like we’re back in the Athlon Thunderbird days, and here’s a virtual cookie to anyone who gets that reference. Luckily, the Turbo and GPU clocks are pretty decent even when compared to the previous generation.
So remember that little quiz I gave you at the beginning of this article? Well here’s the list again and based on what you know, let’s put this in order from the fastest to the slowest. So the i7-1065G7 is the first, since it’s part of the Y series and has four cores and eight threads. Next is the i5-1035G7. Now to be honest, did anyone really guess that? Well, what about the other i7? Remember what I said about the low-power and low-performance Y series not having a 5 at the end of its name? Well here it is good people. Performance on the i7-1060GT will be less than the i5-1035G7. Finally, there’s the i5-1030G7, which is also a four core/eight thread processor in the Y series. And it’s also the lowest performing of the processors that I’ve shown here.
Now Intel also mentioned that these new processors will have support for Wi-Fi 6 and Thunderbolt 3, but ultimately it is up to the manufacturers to add those features into their designs. I mean, it’s nice to have them as marketing bullet points, but that’s about it. So this is the problem that I see right now. Intel’s names are so damn long that they actually stopped using the letters U and Y. Those actually helped folks understand the differences between the different series of processors. And remember most people who are investing their hard-earned money into these notebooks aren’t going to watch this video, and definitely aren’t as tech savvy as you and me. The amount of confusion that this is going to cause is going to be an absolute mess.