The Acer Swift 3 SF314 Notebook Review: Swift Gets Swifter With Ryzen 4000
by Brett Howse & Andrei Frumusanu on May 5, 2020 8:00 AM ESTSection by Andrei Frumusanu
SPEC: Renoir vs Picasso vs Ice Lake
Kicking off our look at performance, let's start with SPEC and analyzing how AMD managed to increase the performance against Picasso with its first-generation Zen cores. The newer generation Renoir with Zen2 cores should in theory see more significant performance upgrades thanks to the improved IPC of the new microarchitecture – but today’s 4700U also is just a mid-tier part with a lower 15W TDP as well as “just” featuring DDR4-3200 – not the fullest memory configuration possible as the SoC also support LPDDR4X up to 4266 MHz.
For our SPEC harness, We’re testing under WSL1 in Windows due to simplicity of compilation and compatibility, and are compiling the suite under LLVM compilers. The choice of LLVM is related to being able to have similar code generation across architectures and platforms. Our compiler versions and settings are as follows:
clang version 8.0.0-svn350067-1~exp1+0~20181226174230.701~1.gbp6019f2 (trunk) clang version 7.0.1 (ssh://git@github.com/flang-compiler/flang-driver.git 24bd54da5c41af04838bbe7b68f830840d47fc03) -Ofast -fomit-frame-pointer -march=x86-64 -mtune=core-avx2 -mfma -mavx -mavx2
Our compiler flags are straightforward, with basic –Ofast and relevant ISA switches to allow for AVX2 instructions.
Single-Threaded Performance
In SPECint2017, the new Zen2 based Ryzen 7 4700U is performing well, although the score increase compared to the last-generation 3780U aren’t as big for some workloads. As a reminder, the 3780U clocks up to 4.0GHz in terms of boost clocks while the 4700U clocks up to 4.1GHz. SMT being disabled on the 4700U should also give it a slight advantage in IPC over other Zen2 platforms given that it doesn’t have to statically share some core resources anymore in this mode.
Renoir showcases the biggest increases in workloads such as 548.exchange2_r and 525.x264_r which are back-end execution bound workloads, and the microarchitectural improvements here help a lot.
On the other hand, the weakest improvements are seen in workloads such as 520.omnetpp_r – this test is mostly memory latency bound and unfortunately the new chip here barely just matches its predecessor. The same can be said about 505.mcf_r where the improvements are quite meager.
The generational improvements here aren’t enough to catch up to Intel’s Sunny Cove cores in the Ice Lake i7-1065G7. Although that core might be running at higher single-core TDPs and power consumption, it still makes for a big gap in some of the more instruction pressure and cache pressure high workloads such as 500.perlbench_r and 502.gcc_r where the Intel chip still has a considerable lead in.
In SPECfp2017, these are floating point heavier test workloads. The generational increases here are also relatively smaller, with even an odd regression in 527.cam4_r. The Intel chip still has a lead across the board, and with particular large gaps in the more memory heavy workloads such as 519.lbm_r and 549.fotonik3d_r.
Overall, in terms of single-threaded performance, the Zen2 Renoir chip might seem a bit underwhelming, but when looking at the IPC improvement it’s actually not that far off from what’s expected of the new microarchitecture. In both SPECint and SPECfp we see a 12% performance per clock improvement, which is generally in line with Zen2’s generational improvements. As a reminder, the mobile chips here only feature a quarter the CCX L3 cache compared to its desktop counterparts, and that will result in smaller IPC gains, especially in memory heavy workloads.
AMD’s higher-end SKUs such as the 4900HS score slightly better due to the higher boost frequencies, but nothing significant enough to change the single-threaded competitive landscape; Intel still has a substantial advantage here thanks to a much stronger memory subsystem.
Multi-Threaded Performance
The multi-threaded performance comparison today is actually quite interesting just for the fact that the 4700U doesn’t feature SMT. Whilst it lacks multi-threading on the cores, it makes up for it by simply having 8 physical cores on the chip, double that of Picasso-based SoCs as well as Intel’s latest Ice Lake CPUs. Again, performance here will be strongly impacted by the TDP as well as cooling of the systems. Both AMD parts are 15W TDP designs, while the Intel chip sustains 25W.
The resulting multi-threaded performance is quite varied across the benchmarks. What immediately stands out is that in purer execution heavy workloads the new Renoir chip pretty much demolishes its predecessor as well as the competition. This makes sense given that Zen2 has good execution resources and Renoir has double the physical core count to scale up performance.
Renoir’s performance here is very good across most workloads, but there’s a few workloads where Intel’s Ice Lake CPU comes closes or beats it even (502.gcc_r) – again these are the workloads that are most memory intensive.
SPECfp across 8 instances again shows similar results. Renoir fares extremely well across execution bound workloads, but this time around loses a lot more often against the Ice Lake chip in memory-bound workloads even though it has a 2x core count advantage.
The chip’s performance in memory-bandwidth heavy 519.lbm_r is particularly odd as it manages to fare worse than single-core results of a desktop Zen2 based system, and again far behind the Ice Lake counterparts.
Overall, AMD’s 4700U’s multi-threaded performance is fantastic in areas where it should matter most in a laptop-based system. The 2x core count advantage more than makes up for the lack of SMT, and workloads such as rendering and other creative tasks are the platform’s strong-points. In more memory-heavy workloads, performance doesn’t scale all that well as clearly the new chip still has a big weakness in its memory subsystem – scientific workloads or code compiling won’t scale nearly as well.
We have to wonder how Renoir will fare on an LPDDR4X configured system – unfortunately we haven’t been able to test such a platform yet. Overall, the 4700U seems like an excellent chip given its 15W TDP, but let’s see how it performs in our system benchmark suite…
Facebook
Twitter
RSS
191 Comments
View All Comments
lightningz71 - Tuesday, May 5, 2020 - link
It appears that the Acer Swift design philosophy just doesn't translate for anything that requires a high steady-state power draw and thermal load. The Swift had the same issues with the 2XXX and 3XXX series chips as well, so this is nothing new. This is something that Acer has deliberately chosen to make a design trade-off for: sacrifice some thermal dissipation ability to keep the product in the size class that it is intended for.It will be interesting to see the benchmarks on the 4500U in this platform. It was shown in benchmarks of previous versions of the swift that lower end APUs actually performed better in gaming than the top end parts because the system was better able to manage the thermal output and the APU was better able to keep consistent clocks. While the absolute performance was still lower than notebooks with better thermal management implementations, it was a better gaming performer than the top end SKU.
neblogai - Tuesday, May 5, 2020 - link
Generally, you are not complaining about Acer, but about U-series chips from both Intel or AMD. The philosophy of laptops with ~15W chips is that these chips are used in ultrabooks that are responsive and fast in short boosts, and not made for steady power (even if there are some premium devices that offer both).fmcjw - Tuesday, May 5, 2020 - link
The 2019 Swift 5 and LG Gram series were matched to the thermal promises of Intel 10nm chips that took 2 generations to arrive, and maybe 5nm AMD SoCs in 2021 since AMD decided to rush to market with last generation GPU architecture in early 2020. Perhaps with software or AI based optimization, there can be a more optimal mix of processes split between the CPU versus the GPU for best performance within thermal constraints. Not every task is as clear cut as gaming routines, where more of the work is performed more efficiently on the GPU. I'm not sure that will happen though, as such software optimization has the least return on investment outside cloud and data center applications. Not even Apple wants to do it for the 2020 Macbook Air, thermally crippling a fine Intel chip and resolving the issue by sticking a more powerful cooling solution in a Macbook Pro (and charge more).For 2020, Acer managed to get the Swift 3 down to 1.2kg from 1.45kg of the 2019 model through the use of aluminum AND magnesium (not just aluminum as the article states). The 2020 Swift 5 maintains a 1kg weight while including a rare-breed matte touch screen. The Swift 5 is the model you want to get for 100% sRGB at a $300 premium. I think the only reason these fine machines sell for $600 to $900 for a mid-range configuration is the thermally constrained performance of the more stubbornly ambitious SoCs.
Which if they can think outside the box can easily resolve by selling a fan-assisted cooling dock and unleash the full potential of the SoC we already paid for (and charge more).
Fulljack - Tuesday, May 5, 2020 - link
while Vega here is based on GCN 5 and not RDNA, it doesn't mean it's an outdated architecture. AMD did enhance the Vega arch for Ryzen 4000, with 56% improvement over Vega arch found in Ryzen 3000. overall, with enhanced arch, reduced core count, and higher clock, AMD did deliver 2020 Vega 8 that performs 28% better than 2019 Vega 11.https://www.anandtech.com/show/15324/amd-ryzen-400...
Oxford Guy - Thursday, May 7, 2020 - link
All irrelevant when the laptop's cooling is so pathetic:"The laptop really struggled with its thermals, dropping the framerate into single digits often. The device attempted to run at around 18 Watts of power draw, slightly over the 15 Watt TDP, but in fact only averaged around 8 Watts during this run."
csp4me - Tuesday, May 5, 2020 - link
Acer Swift 3 and 5 are in the market for lightweight, cool & quiet laptops within thermal constraints, thus ~ 18W tdp, and throttling under stress test.For the same budget of Acer Swift 5 ~ $900 you can find laptop models with the same quality display or better and also better thermals ~ 28W-35W at the expense of weight 1.3-1.4kg and noise/heat during heavy loads. Examples Lenovo Ideapad S540-13 both AMD or Intel, or Yoga Slim 7 both AMD or Intel.
psychobriggsy - Tuesday, May 5, 2020 - link
When you buy low power laptops - Y or U series TDPs - you are really looking out for these things1) Sustained Single Core Clock
2) Sustained All-Core Clock
3) Race-To-Sleep Single Core Turbo Boost (and the time it can sustain this)
4) Race-To-Sleep All-Core Turbo Boost (and the time it can sustain this)
1 and 2 are what your gaming sessions will occur in. 3 and 4 might help in some particularly CPU-heavy parts, but only for small periods of time.
This is the problem Intel have with their 14++++ chips - 1 and 2 cannot be raised in the TDP they are restricted to (unlike desktop, where they can simply lie about the figures, in a laptop this will affect battery life and be easily detected), so they hype 3 and 4 to compensate.
AMD on 7nm does well in all four measures, but you should never think you'll get long-term turbo clock performance from any mobile chip. I don't know if Renoir has a 4C turbo that can last longer than the 8C turbo for lightly threaded loads.
fmcjw - Tuesday, May 5, 2020 - link
Can a Y or U series SoC be considered the equivalent of an H series SoC internally, with a beefier cooling solution and reduced I/O capabilities externally? I imagine not just the I/O or cooler but also the capacitors and power circuitry need to be higher specified for the higher sustained load.It's just that few makers even try to create a balanced system around the U or Y series, letting Turbo Boost go wild to impress for the first minute or two, or restraint the system thermally to achieve longer battery life even when you can plug it in.
T1beriu - Tuesday, May 5, 2020 - link
>the mobile chips here only feature half the L3 cache compared to its desktop counterpartsSmall correction. Ryzen 4000 actually has a quarter.
ads295 - Tuesday, May 5, 2020 - link
Acer has consistently impressed me with their attention to detail. There are so many moronic OEMs that put in a single module of RAM but even my 2016 Acer E5-553-T4PT came with 2 modules of 2GB DDR4 RAM to enable the A10-9600P to run in dual channel mode.I suppose they don't get paid to debilitate AMD setups.