maflynn
macrumors Broadwell
You can turn the demos off in 3dmark, but for some reason identifying the system was painfully slow. I'm not sure why, it wasn't that bad for the Razer
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thoughts and Tests on the 2.2GHz MacBook Pro
- Thread starter DannEboE
- Start date
- Sort by reaction score
Maybe it couldn't fully identify what the Radeon Pro 560X is. I see validation check errors in the results saying its an incompatible card lmao. Razer would just have a standard Nvidia card.
maflynn
macrumors Broadwell
So, finally I completed all the tests...
The numbers are - 555x : 560x : 560x advantage in % relative to 555x
1. I'm surprised the 560 is that much faster. I mean it should be 43% faster based on number of shaders and clocks, but I didn't expect it to come that close to that theoretical limit in thermally limited MBP chassis, often showing 29% improvement.
2. As I expected, when CPU and GPU are both loaded the gap gets smaller, to the point where 555x is faster - but it could be also a fluke on @DannEboE part and something was wrong during his skydiver test.
3. During GPU only tests I can barely hear the fans, best indication that there is headroom left.
4. Gaming on a laptop without headphones is going to be a new experience for me
5. I still don't know why my TimeSpy CPU score is so low. But - my kid pulled the power cable, I didn't notice initialy and while on battery I got 3699 (12.43 fps), so in the test the 560x was 13% faster and not 29%. @DannEboE uses power supply limited to 60W - maybe that sets a hard power limit instead of using a battery and results in better thermals.
6. Honestly if I didn't already make a choice and knowing the above results I'd test the crap out of 560x to see if it is not too hot/loud for my liking. So I'm glad my 14 days is up and I'm staying with 555x 🙂 It is plenty for my needs and I'm really floored by how quiet it is.
The numbers are - 555x : 560x : 560x advantage in % relative to 555x
- Heaven Score 326 381 17%
- Avg FPS 12.9 15.1 17%
- Valley Score 611 775 27%
- Avg FPS 14.6 18.5 27%
- Superposition Score 2439 3105 27%
- Avg FPS 18.25 23.23 27%
- Firestrike Score 4103 4942 20%
- Firestrike graphics score 4568 5680 24%
- test 1 fps 22.03 28.31 29%
- test 2 fps 18.09 21.91 21%
- Firestrike Physics score 10819 10904 1%
- test fps 34.35 34.62 1%
- Firestrike Combined score 1523 1769 16%
- test 7.09 8.23 16%
- Sky Diver Score 13365 14744 10%
- Sky Diver Graphics Score 13438 16400 22%
- test 1 fps 65.3 79.58 22%
- test 2 fps 57.87 70.72 22%
- Skydiver Physics score 12699 11125 -12%
- 8 thread fps 157.5 150.94 -4%
- 24 thread fps 110.06 94.95 -14%
- 48 thread fps 71.14 61.94 -13%
- 96 thread fps 42.67 37.39 -12%
- Skydiver combined Score 13897 11620 -16%
- test fps 57.19 47.82 -16%
- Time Spy Score 1374 1774 29%
- TimeSpy Graphics Score 1247 1611 29%
- test 1 fps 8.52 11.01 29%
- test 2 fps 6.87 8.88 29%
- TimeSpy CPU Score 3253 4188 29%
- test fps 10.93 14.07 29%
- Final Fantasy Score 2038 2318 14%
- Cinebench GL FPS 105.78 111.43 5%
1. I'm surprised the 560 is that much faster. I mean it should be 43% faster based on number of shaders and clocks, but I didn't expect it to come that close to that theoretical limit in thermally limited MBP chassis, often showing 29% improvement.
2. As I expected, when CPU and GPU are both loaded the gap gets smaller, to the point where 555x is faster - but it could be also a fluke on @DannEboE part and something was wrong during his skydiver test.
3. During GPU only tests I can barely hear the fans, best indication that there is headroom left.
4. Gaming on a laptop without headphones is going to be a new experience for me
5. I still don't know why my TimeSpy CPU score is so low. But - my kid pulled the power cable, I didn't notice initialy and while on battery I got 3699 (12.43 fps), so in the test the 560x was 13% faster and not 29%. @DannEboE uses power supply limited to 60W - maybe that sets a hard power limit instead of using a battery and results in better thermals.
6. Honestly if I didn't already make a choice and knowing the above results I'd test the crap out of 560x to see if it is not too hot/loud for my liking. So I'm glad my 14 days is up and I'm staying with 555x 🙂 It is plenty for my needs and I'm really floored by how quiet it is.
@Thysanoptera Thanks for taking all those tests! I wonder if I could somehow manually control the fans on bootcamp side so that I could take full advantage of both the GPU and CPU. I can thermally limit CPU to 40W maybe to make sure GPU has all the headroom it needs.
Looks like for older version there are modified drivers that allow Wattman to show, with this you should be able to modify GPU clock speeds to what you need, don't know about any fan control software for bootcamp. But for now this doesn't work on the 560x/555x.
That may be more complicated, I guess we could use iStatMenus on Mac to capture 'GPU proximity' sensor (Valley and Heaven work on Mac) while running benchmarks, on Windows 3dmark saves the logs for temps, frequency and utilization, but if you look at the screenshots they don't look realistic at all (spikes to 120C, and flat in general), except for FPS, I think the driver doesn't present the information yet and 3dmark cheats by calculating the values instead of reading the sensor.
[doublepost=1534028842][/doublepost]For what is worth, these are screenshots after Heaven benchmark runs, Extreme preset, had to run it 4 consecutive times to get stable fan and temps. After the first run the fans barely moved from the idle settings. Screenshots made around the end of benchmark, 255 second mark. Someone with 560x can easily replicate this.
@Thysanoptera
Attempted to recreate the same procedure as above, ran it in extreme 4 times and took a screen shot at 255 seconds of the last run. This is without using my monitor so it shouldn't be more stressed than normal too.
It has higher temperatures for sure. Though, it does look to eek out a bit more performance at the cost of higher temps. 4% higher temperature for 17% higher performance is not the worse tradeoff in the world.
Attempted to recreate the same procedure as above, ran it in extreme 4 times and took a screen shot at 255 seconds of the last run. This is without using my monitor so it shouldn't be more stressed than normal too.
It has higher temperatures for sure. Though, it does look to eek out a bit more performance at the cost of higher temps. 4% higher temperature for 17% higher performance is not the worse tradeoff in the world.
@Thysanoptera
Attempted to recreate the same procedure as above, ran it in extreme 4 times and took a screen shot at 255 seconds of the last run. This is without using my monitor so it shouldn't be more stressed than normal too.
It has higher temperatures for sure. Though, it does look to eek out a bit more performance at the cost of higher temps. 4% higher temperature for 17% higher performance is not the worse tradeoff in the world.
Thank you both, this thread has been very helpful in helping me decide what to go for...I don’t do any 3D rendering etc. and as even the 560x is not that great for the latest games, I think I’ll stick with the 555x, the extra £90 in my pocket and a slightly cooler machine.
The temps are a function of fan rpms, and it look like to dissipate a mere 4W of additional power you need 1000rpm more. Looks like they trying to keep 70C. To get an idea of actual temp difference you'd need to run the same fan rpm, give me sec, I'm just curious at this time.
Assuming the same room temperatures, that's not only higher temperatures. The fans are running at 1k rpm more.
Even better if the difference between the 555x and 560x in terms of temps is more than 4%...looking forward to your update 🙂
I've set the fans to 4626 and 4263 rpm in Macs Fan Control and run the benchmark 4 times. Ended up with 59 C on GPU, and CPU cores around 60C. 560x was at 72C GPU and 74 on CPU cores. So thats actually 22% temp increase for 17% gain, I mean in this benchmark, but it gives a general idea. I actually like the fans at those settings, they're not too loud but the keyboard remains cool to the touch. I think apple is too skittish with fan curves.
Again thanks for this, think I’m finally convinced on going for the base 15” with the 555x and upgraded SSD 🙂
RumorConsumer
macrumors 68000
Does anybody else want the 2.6 to just not have the base model? I can't shake somehow that Im supposed to have it but this performance ceiling on the 2.2 is pretty interesting. Im also coming from the i7 3.5 13" 2017. Its just unreally faster huh?
@Thysanoptera Thanks for the secondary round of testing to actually confirm the difference. I was a little surprised to see such a low temperature difference. That honestly makes more sense than just a little 4% difference.
I actually had the 2.6 model for a week or so actually! I found myself hitting the thermal limit of the chassis a lot quicker than my current 2.2 GHz model. From averages, the 2.2 GHz model seems to run about 3-5 C cooler at any given task. The one nice thing with the 2.6 model was that I consistently was able to hit 1000+ on CineBench results. Plus, I felt like if the 2.2 GHz was going to run cooler, than that would give me more headroom when playing games with the 560X.
Yeah, I notice a difference in my day to day. For some reason, my 2017 laptop began slowing down pretty dramatically and battery life tanked to 3 hours lmao. I took it to the Apple store who said everything was normal so what can you do. The same workflow on this laptop is easily getting me 8 hours compared to the 3 I was getting before. That upgrade in itself made it worth the price tbh.
I actually had the 2.6 model for a week or so actually! I found myself hitting the thermal limit of the chassis a lot quicker than my current 2.2 GHz model. From averages, the 2.2 GHz model seems to run about 3-5 C cooler at any given task. The one nice thing with the 2.6 model was that I consistently was able to hit 1000+ on CineBench results. Plus, I felt like if the 2.2 GHz was going to run cooler, than that would give me more headroom when playing games with the 560X.
Yeah, I notice a difference in my day to day. For some reason, my 2017 laptop began slowing down pretty dramatically and battery life tanked to 3 hours lmao. I took it to the Apple store who said everything was normal so what can you do. The same workflow on this laptop is easily getting me 8 hours compared to the 3 I was getting before. That upgrade in itself made it worth the price tbh.
RumorConsumer
macrumors 68000
My usage isn’t that intensive but I like s*** to fly when I ask for it to. Is your workflow cpu intensive?
@RumorConsumer Workflow isn't too intensive (I believe), but with many browser tabs, streaming video on the side, and querying databases slows my old laptop to a crawl. Excel on the Mac slows everything down lmao.
Sounds like you don’t need the 2.6 so get the 2.2 and in a few years, use the saved $ towards a new laptop imo.
@DannEboE @Thysanoptera do you guys notice a big difference in temp using an external monitor...I only just learned that the dGPU drives any external displays you connect so obviously that’d cause it to heat up?
I personally haven't really gamed much on the computer, too many things happening for me to settle down and play games. Maybe in a week or so.
I drive a 4k monitor so I assume it pushes the GPU a bit more. When I plug it in, the GPU temperatures will spike about 15 F.
Though, when I plug in, I manually increase my fans to ~3500 RPM to help compensate.