Intel promised that it’s 11th-gen Core “Tiger Lake” processors would bring up to a 2x boost in graphics performance over previous-gen “Ice Lake” chips, as well as more modest CPU performance gains.
But up until now we’ve only had Intel’s word to go on. Now a handful of tech news websites have published the results of real-world tests and synthetic benchmarks, and it seems like Tiger Lake is indeed a big step up.
It could give AMD’s Ryzen 4000U series chips a run for their money, and may even make NVIDIA’s entry-level MX series discrete graphics cards obsolete.
But there are still a few important caveats to keep in mind.
The first is that websites including AnandTech, Engadget, Gizmodo, PC Magazine, The Verge, and Tom’s Hardware are all testing a reference design laptop provided by Intel. This is not an actual product you can go out and buy (although it was manufactured by MSI and sure looks a lot like some recent MSI laptops).
A second thing to keep in mind? The reference unit has Intel’s most powerful 11th-gen chip – the Intel Core i7-1185G7. It’s unclear if we’ll see the same kind of performance improvements from Core i3, Core i5, or even other Core i7 chips in the Tiger Lake family.
Finally, this chip has a configurable TDP, allowing it to run at 12 to 28 watts. Ice Lake chips only go up to 25 watts, and most laptops with the processor shipped with a 15W configuration. AMD’s Ryzen 4000 series notebooks also top out at 25 watts and, again, many laptops that ship with these chips are configured to run at 15 watts.
Anyway, I’d definitely recommend following some of the links above for more details. But in a nutshell:
- The Core i7-1168G7 offers stronger single-core performance than any other chip in its class in most tests.
- It also scores well in graphics benchmarks, even playing Overwatch at its highest graphics settings at close to 60 frames per second.
- A Lenovo Slim 7 laptop with a 25 watt Ryzen 7 4800U processor does come out ahead in some tests, including those that take advantage of multithreading (it’s an 8-core/16-thread chip, while Intel’s is a 4-core/8-thread processor), and video transcoding (using Handbrake, at least).
The first Intel Tiger Lake-powered laptops should arrive this fall.
I didn’t see any battery life evaluation done on Intel’s reference design. I get the feeling their reference design was tuned for performance.
On the one hand, it is wonderful to equip laptops with the power to meet high performance demands and also scale back power use in the same device… And on the other, we need exactly the data that you mention. In terms of power, a 2006-era Dell Latitude with an i5-3320m is not a lot less powerful than an i7-10**U outside the new-processor optimizations. But the 45W near-constant power use compared to the newer laptop’s ability to fit inside a 5W power envelope much of the time makes the real difference.
Since battery life is tied to more than just the CPU, it makes sense to not be talking on those terms for a CPU reference platform – battery capacity is equally important if not more-so, and component usage (including the screen) impacts it too.
A view of overall usage by watts for daily tasks would be interesting to see (Ars noted ~8W at idle, up to ~68W boosted at full load), but it won’t give an indication of what to expect from production laptops.
Anandtech’s write-up covered power consumption during benchmarks, but not day-to-day activities. Still, I believe their takeaway was ~15% improvement in performance for the same wattage (better with AVX512). Assuming the CPU is 50% of power consumption on the (lower-power) battery life tests, maybe looking at a 5-10%-ish battery life increase?
Comments are closed.