Comparing 15 W TGL to 15 W ICL to 15 W Renoir

Despite the hullaballoo with the 28 W numbers on Tiger Lake, we suspect that most OEMs will still be positioning the hardware inside chassis built for the 15 W ultraportable market. This is where most of Intel’s OEMs have had success over the last decade, as the lower cooling requirements allow for a more user-friendly design. At 28 W, there is more of a cross-over into laptops that have discrete graphics options, and the main company that has succeeded in offering 28 W laptops without discrete graphics has been Apple - most Intel partners, if they want discrete graphics, end up looking at the 45 W processors with more cores.

So in that respect, our main battle should occur between the products built for 15 W. To that end we have been able to put the three together that will command this holiday season’s offerings: Ice Lake, Tiger Lake, and AMD’s Renoir.

  • For our Ice Lake system, we have the Microsoft Surface Laptop 3. This has the top-of-the-line quad-core Core i7-1065G7, along with 16 GB of LPDDR4X-3733. Base 1.3 GHz, Turbo 3.9 GHz.  Because this is an OEM design, Microsoft have determined the PL1 and PL2 values, and so they might be different from a ‘base’ design, however this is data from a real system.
  • The Tiger Lake system is our Reference Design from Intel, running the quad-core Core i7-1185G7 at 15 W TDP mode. It has 16 GB of LPDDR4X-4266. Base 1.8 GHz, Turbo 4.8 GHz.
  • Our AMD Renoir system is one of the most premium examples of AMD’s Ryzen Mobile in a 15W form factor, the Lenovo Yoga Slim 7 with the eight-core Ryzen 7 4800U processor. Even when set to the highest performance mode, the system still operates with a 15 W sustained power draw. It comes equipped with 16 GB of LPDDR4X-4266. Base 1.8 GHz, Turbo 4.2 GHz.

Compute Workload

For our 15 W comparisons, we can look again at the same benchmarks as the previous page. First up is y-Cruncher, an AVX2/AVX512 compute workload that tasks the CPU and the memory by calculating 2.5 billion digits of Pi, and requires ~11 GB of DRAM.

As we saw on the previous page, our Tiger Lake system in green at 15 W turbos up to ~53 watts before very quickly coming down to 15 W for the rest of the test.

The Microsoft Surface Laptop 3, by virtue of an OEM system, has different behavior - it turbos for longer, settles into a short turbo limit of 25 W, and then after about two minutes comes down to 20 W. The system then appears to opportunistically up the power draw until the end of the test, likely due to detecting extra thermal headroom.

The AMD Renoir processor does not turbo as high, peaking at only 38.9 W. Over the course of the next 100 seconds or slow, we see a small ramp down to just under 30 watts, before a more consistent decline over 30 seconds to 15 W, before staying at 15 W for the full test. The Renoir here has eight cores rather that four, but is running AVX2 rather than AVX-512 code.

The results are as follows:

  • Ice Lake: 233 seconds, for 6072 joules, averaging 26.1 W
  • Tiger Lake: 241 seconds for 4082 joules, averaging 17.0 W
  • Renoir: 234 seconds for 5386 joules, averaging 23.0 W

All three systems perform the test in roughly the same amount of time, however the Tiger Lake system is very much ahead for efficiency. Tiger Lake effectively shaves off a third of the power from the previous generation Ice Lake system. We weren’t expecting this much of a jump from Ice Lake to Tiger Lake, but it would appear that Intel has done some work on the AVX-512 unit, and is putting that new high-performance transistor to use.

Professional ISV Workload

Moving onto the Agisoft test - as mentioned on the previous page, this is a 2D image to 3D modeling workflow where the algorithm comes in four stages, some of which prefer full multi-thread throughput, while others are more frequency and memory sensitive.

First, the Renoir finishes in almost half the time, mostly due to the fact that it has double the number of cores - there is no AVX-512 codepath in this test, and so all the processors rely on a mix of SSE, AVX, and perhaps some AVX2. That aside, the turbo behavior of Renoir is very interesting - we get almost 10 minutes of higher-than-base performance before the algorithm sets into a routine, hovering around 22 W. Because this test doesn’t attack the vector units as hard as the previous test, it may be a case that the Renoir system can manage the power distribution a bit better between the eight cores, allowing for the higher turbo.

Between the Ice Lake and the Tiger Lake, from the graph it would appear to be a double win for Tiger Lake, finishing in a shorter time but also consuming less power. The results are:

  • 15 W Renoir: 2842 seconds for 62660 joules
  • 15 W Ice Lake: 4733 seconds for 82344 joules
  • 15 W Tiger Lake: 4311 seconds for 64854 joules

In this case, it’s a win for Renoir - a lot shorter time, and better power to boot, derived from the eight cores built on TSMC 7nm. Tiger Lake still represents a good jump over Ice Lake, offering 10% better performance at only 79% of the power, or a 13% increase in performance efficiency.

Power Consumption: Comparing 15 W TGL to 28 W TGL CPU ST Performance: SPEC 2006, SPEC 2017
POST A COMMENT

252 Comments

View All Comments

  • tipoo - Thursday, September 17, 2020 - link

    “Baskin for the exotic”
    I see what you did there...
    Reply
  • ingwe - Thursday, September 17, 2020 - link

    I didn't get it until I read your comment. Reply
  • Luminar - Thursday, September 17, 2020 - link

    RIP AMD Reply
  • AMDSuperFan - Thursday, September 17, 2020 - link

    "Against the x86 competition, Tiger Lake leaves AMD’s Zen2-based Renoir in the dust when it comes to single-threaded performance." - But I am hoping Big Navi can compete well against this Intel chip. Reply
  • tipoo - Thursday, September 17, 2020 - link

    What does Big Navi have to do with a laptop CPU? Reply
  • AMDSuperFan - Thursday, September 17, 2020 - link

    You care about games don't you? This Intel Tiger won't have an answer for Big Navi. We can look forward to that showing who is the boss. Reply
  • blppt - Thursday, September 17, 2020 - link

    Based on preliminary data, they'll both be about 2 years behind Nvidia, what with Big Navi only matching a 2080ti, and not available for another month at the earliest. Reply
  • hecksagon - Friday, September 18, 2020 - link

    Crazy how you can make that prediction, the only preliminary data that is out is a photograph of the card. Are you a wizard? Reply
  • blppt - Friday, September 18, 2020 - link

    Incorrect.

    https://wccftech.com/amd-radeon-navi-gpu-specs-per...
    Reply
  • HarryVoyager - Friday, September 18, 2020 - link

    I'm not really seeing where you are getting that from. We know that RDNA2 can hit 2.23Ghz from the PS5 implementation, and we have solid rumors that it the top end one will be an 80CU chip, rather than a 40 CU chip. That implies on the order of a 230% improvement over the 5700XT, if their are no other performance improvements. That alone puts it in the 30-40% improvement range over the 2080 Ti. Given we've already seen at least a few AMD benchmarks of unidentified cards showing a 30-40% improvement over 2080 To performance, that sort of lift does seem likely.

    If I had to guess, that RDNA2 that recently showed up with a near 2080 TI performance is probably a 6700 competitor to the 3070, not the top end card. Those do have to be developed and tested too, after all.
    Reply

Log in

Don't have an account? Sign up now