Beelink GTR7 mini-PC Review: A Complete AMD Phoenix Package at 65W
by Ganesh T S on August 24, 2023 8:00 AM ESTSystem Performance: Multi-Tasking
One of the key drivers of advancements in computing systems is multi-tasking. On mobile devices, this is quite lightweight - cases such as background email checks while the user is playing a mobile game are quite common. Towards optimizing user experience in those types of scenarios, mobile SoC manufacturers started integrating heterogenous CPU cores - some with high performance for demanding workloads, while others were frugal in terms of both power consumption / die area and performance. This trend is now slowly making its way into the desktop PC space.
Multi-tasking in typical PC usage is much more demanding compared to phones and tablets. Desktop OSes allow users to launch and utilize a large number of demanding programs simultaneously. Responsiveness is dictated largely by the OS scheduler allowing different tasks to move to the background. The processor is required to work closely with the OS thread scheduler to optimize performance in these cases. Keeping these aspects in mind, the evaluation of multi-tasking performance is an interesting subject to tackle.
We have augmented our systems benchmarking suite to quantitatively analyze the multi-tasking performance of various platforms. The evaluation involves triggering a ffmpeg transcoding task to transform 1716 3840x1714 frames encoded as a 24fps AVC video (Blender Project's 'Tears of Steel' 4K version) into a 1080p HEVC version in a loop. The transcoding rate is monitored continuously. One complete transcoding pass is allowed to complete before starting the first multi-tasking workload - the PCMark 10 Extended bench suite. A comparative view of the PCMark 10 scores for various scenarios is presented in the graphs below. Also available for concurrent viewing are scores in the normal case where the benchmark was processed without any concurrent load, and a graph presenting the loss in performance.
| UL PCMark 10 Load Testing - Digital Content Creation Scores | |||
| UL PCMark 10 Load Testing - Productivity Scores | |||
| UL PCMark 10 Load Testing - Essentials Scores | |||
| UL PCMark 10 Load Testing - Gaming Scores | |||
| UL PCMark 10 Load Testing - Overall Scores | |||
Configuring the GTR7's Phoenix SoC at 65W gives it plenty of headroom to tackle multiple concurrent demanding workloads triggered by mainstream users (of the type benchmarked by PCMark 10). The GTR7 was already having leading performance in the absence of concurrent loading. Activation of such loads reduces the raw score, but the performance loss is less than the competition, and the GTR7 continues to remain on top.
Following the completion of the PCMark 10 benchmark, a short delay is introduced prior to the processing of Principled Technologies WebXPRT4 on MS Edge. Similar to the PCMark 10 results presentation, the graph below show the scores recorded with the transcoding load active. Available for comparison are the dedicated CPU power scores and a measure of the performance loss.
| Principled Technologies WebXPRT4 Load Testing Scores (MS Edge) | |||
The observations made for the overall PCMark 10 scores continue to hold for WebXPRT 4 also. The GTR7 has the least performance loss, and that allows it to leapfrog the RPL-P systems in performance.
The final workload tested as part of the multitasking evaluation routine is CINEBENCH R23.
| 3D Rendering - CINEBENCH R23 Load Testing - Single Thread Score | |||
| 3D Rendering - CINEBENCH R23 Load Testing - Multiple Thread Score | |||
Intensive background tasks keep the efficiency cores busy, and take up some of the power budget. As a result, while the RPL-P systems came out on top in the absence of the background loads in single-threaded mode, the AMD-based systems manages to outwit them in raw scores after the introduction of the load. Multi-threaded performance continues to be a strongpoint of the GTR7, with the system coming out on top irrespective of the concurrent loading.
After the completion of all the workloads, we let the transcoding routine run to completion. The monitored transcoding rate throughout the above evaluation routine (in terms of frames per second) is graphed below.
The transcoding rate during different segments is also recorded below.
| Beelink GTR7 (Ryzen 7 7840HS) ffmpeg Transcoding Rate (Multi-Tasking Test) |
|||
| Task Segment | Transcoding Rate (FPS) | ||
| Minimum | Average | Maximum | |
| Transcode Start Pass | 2.5 | 14.12 | 46.5 |
| PCMark 10 | 0 | 13.99 | 44.5 |
| WebXPRT 4 | 6.5 | 13.63 | 23 |
| Cinebench R23 | 5 | 13.78 | 48.5 |
| Transcode End Pass | 5.5 | 14.8 | 45.5 |
The comparison is against RPL-P systems such as the NUC BOX-1360P/D5. The GTR7 has a much lower delta in the transcoding rate between different task segments, but the power budget ensures that it doesn't translate to a lower primary workload score as seen in the graphs above. Overall, the RPL-P systems seem to prioritize foreground task better compared to AMD systems, but with the right power budget, the end user may not even notice the aspect.
Facebook
Twitter
RSS
56 Comments
View All Comments
Gm2502 - Friday, August 25, 2023 - link
Games, productivity or anything else the VAST majority of the world does when they don't have to run on a proprietary, non-upgradable glorified mobile phone. This Miniforum is superior in EVERY metric mate, so take your fan boy rubbish elsewhere.darkswordsman17 - Friday, August 25, 2023 - link
"Glorified mobile phone" despite this not being in a phone, and the reason the AMD chip isn't is because its literally incapable of being put in such a form factor due to the power and thermal limits still exceeding what a phone could deal with. And the M2 chip is still competitive enough that its not ridiculous to compare them.There's definitely flaws (and plenty of legitimate criticisms of Apple/Apple product), but anyone acting like there aren't flaws with the PC side is just delusional. Acting like you can't do anything more than run a few proprietary apps on Apple stuff is so ridiculously nonsensical at this point that you really need to look in the mirror. Yes, lemurbutton is ridiculously nonsensical, but going the total opposite direction is just as pointless and biased.
Heck, is no one going to call out the proprietary power connector on this? That alone stopped me from considering this over its competition, and is just absurd in this market space.
Further, Apple could make the Mini even smaller (there is a YouTube where someone does exactly that, I think it ended up roughly 3/5 the size, maybe even smaller) with no real performance loss. That'd put it roughly 1/4 the size of this box. Try putting Phoenix in such and the M2 would probably outperform it.
Gm2502 - Saturday, August 26, 2023 - link
Roflol, so ARM architecture isn't synonymous with mobile technologies? Literally accounts for over 95% of ARM sales, so my statement about it being a glorified mobile phone isn't far off. The M2 Chip is not even close to be competitive, with cpu performance difference from Tomshardware - "The Ryzen 7 7840U was anywhere from 15-71% faster than the M2" and The Ryzen 7 7840U was anywhere from 35 - 180% faster than the M2" in graphics testing. So again, if it wasn't for the stupid propietry OS optimised for the iPhone reject chip, it wouldnt even be a comparison.As for size, again your lying about a mac mini being 1/4 the size. M2 Mini - 197mm x 197mm x 36mm, this GTR7 box is 168mmx120mmx49mm, so cut the crap there. This thing is roughly the same size as the Mac Mini. You complain about design choices, well you literally have hundreds of designs for tens of companies to choose from, each with different hardware (like use a DIN plug is a big deal, but fanboyz have to clutch at straws), all of which can be upgraded and user repaired. Want more RAM or storage on a M2, go buy a new PC.
Want freedom of choice to run any application and game, go windows. If you go apple, pray it works if its not on the very specific optimised for MacOS list, and good thing most games don't run in it becuase the god awful IGPU of the M2 won't be able to produce much more then a slide show, unless I look to double or quadruple the price and look at M2 Pro or ultra equiped M2 machine, which will be roughly 5-10 x the price is this machine...
So again, calling you and your "facts" as fanboyism not grrounded in any reality. Especially stuoid when you comment on a single PC manafacturer design choices knowing full well that PC allows choice across multiple manafacturers while Apple if wholly proprietary. Stupid argument across the board.
Samus - Thursday, August 24, 2023 - link
You can upgrade it to Windows 10 and it'll be solid.mode_13h - Thursday, August 24, 2023 - link
The M2 has 4x P-cores and 4x E-cores. Is that really faster than 8x dual-threaded Zen 4 cores?> has AI inference
Phoenix features AMD's new "Ryzen AI" accelerator. I don't know how they compare, but if you're running Windows or Linux, you'll probably find AMD's solution better supported than Apple's.
I think that's the key. If you *want* to run MacOS, then you'll do better with the M2 mini. If you'd rather run Linux or Windows, then it's really no contest. Recent benchmarks of the M2 running Linux have shown that its Linux support has a long ways to go, before it's remotely competitive.
https://www.phoronix.com/review/apple-m2-zen4-mobi...
dontlistentome - Thursday, August 24, 2023 - link
Have you any concept of how bored we are of these comments about Macs?Qasar - Thursday, August 24, 2023 - link
keep in mind lemurbutton is paid by apple to post how great their products are, while posting no proof or anything.you wouldnt want him to get punished by his apple overlords, would you???
bji - Thursday, August 24, 2023 - link
Do you have proof of that or are you just talking out of your ass?Qasar - Thursday, August 24, 2023 - link
have you seen his other posts ? nothing but pro apple, amd and intel suck and cant compete with m1/m2/m3/ etc on any levelhe mostly talks out of his ass when it comes to apple
PeachNCream - Friday, August 25, 2023 - link
Not a paid shill, I would think, but someone just having fun with the bags of hammers that read Anandtech that fall for trolling every. single. time.Wouldn't you keep that up if you didn't need to change tactics and could "outsmart" your audience while gaining imaginary internet attention to feel good about yourself?