Up to 4.8 GHz at 50 W, 2x GPU with Xe, brand new

In August, Intel hosted one of its rare Architecture Days in which the company went into some detail about its upcoming Tiger Lake processor. This included target markets, core counts, graphics counts, a look at some of the new acceleration features, and the promise of a product launch later in the year. The launch of that product is here, and Intel is providing Tiger Lake with speeds and feeds, providing details and the benchmark performance expected for the next generation of Intel handheld devices.

What we already know

Recapping what we learned at Intel Architecture Day 2020, Tiger Lake’s standard form is a quad-core processor with an updated graphics architecture based on the latest Intel manufacturing process.

These new 11^th The Generation Core processors use four of Intel’s latest cores built with the Willow Cove microarchitecture, a slightly modified version of the Sunny Cove microarchitecture found in Intel 10^th Generation Ice Lake processors. These new cores offer more performance than before, peaking at 4.8 GHz compared to the 4.0 GHz seen in the previous generation, a 20% improvement. However, the underlying clock-by-clock performance improvements are minimal, and Intel is instead focusing on that frequency gain.

The reason for the frequency gain comes from the manufacturing process, previously Intel’s 10 ++ process, renamed once to 10+ and then again during Architecture Day to its new name: 10nm Intel SuperFin technology. The goal of this manufacturing process was twofold: first, to increase efficiency and scalability to allow for higher frequencies, but also to improve performances. As a result, we will see Intel claiming higher frequencies, which lead to higher performance, at similar power levels to the previous generation. These cores support AVX-512 as well as Intel’s DL-Boost acceleration libraries.

On the graphics side, Tiger Lake uses Intel’s new X^me-LP graphic architecture, which the company also detailed on Architecture Day; we have a separate article that explains the differences. In short, X^me-LP increases the raw thread count and computational count per execution unit (UE), as well as the cache hierarchy and some acceleration functions. For Ice Lake, we saw 64 EU running at 1100 MHz, while for Tiger Lake we will see 96 EU (+ 50%) running at 1350 MHz (+ 22%). Throw in some of the other benefits and we should supposedly see a two-fold improvement in graphics performance compared to the previous generation on paper. For AI calculation, the charts also support DP4A instructions for INT8 inference workloads.

The graphics display pipeline has been improved, with support for AV1 decoding, as well as display channels for up to four 4K60 displays or a single 8K60 display. There is a direct-to-memory display port to improve latency – memory support at Tiger Lake includes LPDDR4X-4267 (up to 32GB), DDR4-3200 (up to 64GB), and when available we’ll see LPDDR5-5400 devices in 2021. For accelerators, Intel has also upgraded its Gaussian Neural Accelerator to version 2.0, to help offload simpler AI inference workloads, such as noise cancellation.

Other improvements to Tiger Lake include native Thunderbolt 4 support, with the controller built into the CPU allowing up to four TB4 ports per device. Wi-Fi 6 support is also enabled through a CNVi interface. Intel’s power / frequency algorithms are also updated, allowing for power and clocks to be scaled separately in the CPU, GPU, and memory fabric. Intel stated that it has doubled the internal bandwidth of the fabric, allowing all of these parties to communicate with each other with more data.

Today’s announcements

Speeds, performance, new style

The difference between previous Architecture Day and today’s Tiger Lake launch is that today is more about end-user performance – the speeds and benchmarks end-users will use to compare it to other products on the market. Not only this, but Intel partners can now start announcing future Tiger Lake-based products.

The details we should start with right away are the processors. Intel has two categories of Tiger Lake parts, ranging from 7 W to 28 W. First are the UP3 parts.

Higher 12-25W processors are technically known as ‘UP3’ processors, or what we used to call U-series hardware. These processors have a nominal TDP of 15W, but laptop companies can increase or decrease power depending on how they build their systems. At the higher end is Intel’s Core i7-1165G7, a quad-core processor with hyperthreading and the full 12MB of L3 cache. It has a base frequency of 3.0 GHz, a single-core turbo frequency of 4.8 GHz, and an all-core turbo frequency of 4.3 GHz. For graphics, it has the full 96 EU available, running at the frequency 1350 MHz maximum. Memory media is listed as LPDDR4X-4266 and DDR4-3200.

Intel divides its processors by number of cores, graphics, and frequencies. Anything listed in the ‘G7’ list means the highest class of graphics, although it’s worth noting that this means 96 EU for Core i7 parts, but only 80 EU for Core i5 (which also has less L3 cache) . Below are the G4 charts, with only 48 execution units enabled. Only the top Core i7-1165G7 also gets the highest graphics frequency. Core i3 processors have a reduction in memory support, and that lower Core i3-1115G4 only has two cores.

The 7-15W processors are the old ‘Y series’ processors, now known as UP4.

At the top is the Core i7-1160G7, with all four cores enabled with hyperthreading and 12MB of L3 cache. This processor has a 1.2 GHz base clock, 4.4 GHz single-core turbo, and 3.6 GHz all-core turbo. All 96 execution units run at 1.1 GHz and only supports LPDDR4X memory, up to 4266 MT / s. All processors support four PCIe 4.0 lanes.

There is only one Core i7 in this rage, and the Core i5 part has less EU and lower frequencies, before the Core i3 parts are divided into quad-core and dual core with only 48 EU per piece.

These processors fit into the UP4 power envelope primarily because the base frequency is so low, a trait of Intel’s low-power processors over the years.

The aim of these pieces is to go from fanless designs around 9 W with the UP4, and then up to 28 W perhaps with a discrete GPU at 28 W. As with the previous generation of laptops in Intel 10^th Generation, we expect to see some Tiger Lake designs that take the higher performance processors and allow for higher power modes, in that 25-28W range. This will allow these products to deliver better sustained performance beyond the turbo period, for example for rendering and video encoding.

Intel’s turbo boost technology has come under a lot of scrutiny lately, mostly from seeing impressively high power figures within that turbo period. Intel claims that the Turbo mechanisms are there to take advantage of improvements in system design until the system has other factors taking precedence, such as thermals and / or power. This time, Intel explains that Tiger Lake has better turbo algorithms, allowing the CPU / GPU / fabric turbo system as needed within the turbo window. The entire system is a feedback loop that takes inputs such as workload, system sensors, power limits, and a balancer to find the correct frequencies.

So while Intel’s ‘TDP’ value, also known as Power Limit 1, varies between 7W and 28W depending on the system it is installed in, the power level in turbo modes can be much higher. We have seen this secondary power level, PL2, increase over the years to +4 W over TDP to +20 W over TDP. In the previous generation of Ice Lake, we saw 15 W processors with a maximum of 44 W in turbo mode. Intel even provides a useful diagram:

What does this mean for Tiger Lake? According to Intel’s own numbers, the maximum power of the turbo is around 50W, either when it is plugged in or off. Intel would like to point out that it manages its turbo algorithms to ensure a constant experience for its customers, whether the system has access to wall power or is battery dependent.

Unlike the last generation, where we were able to test Intel’s own mobile development platforms before a launch, we are not sure which Tiger Lake product will be the first to test in our labs. It will be interesting to see what they send us and how that system works in comparison.

On the page, read about Intel’s new branding, Intel Evo, Tiger Lake performance, and future Intel announcements.