|
LionArcher posted:I’m saying it felt very suspect to me. I would want to see him doing it with two machines side by side, (one running FC one running P) and see if that made a difference. Also, could he have a bad unit? The whole thing. Especially after his terrible video about hating Apple last week did not come off like he entirely knew what he was talking about, or being honest. Explain how FCX vs Premiere makes a difference if his point is that putting any computation on the i9 CPU starts making it throttle? It doesn't matter what video editing app you use in that case. He showed a 2017 i7 MBP vs the 2018 i9 MBP with the i7 beating out the i9 in terms of rendering the same file. Then put the i9 into the fridge and showed that once that happened it beat the i7 easily. Literally in this video and the "hating apple" video he says he loves the Apple ecosystem. And also said he would have loved to use the i9 if not for the issues. His "Apple Hating" video was about how Apple is lagging behind by not using 8th gen CPUs. Then he says if you're using Premiere on a Mac it's a bad choice because the PCs all have 8th gen CPUs. He critcizes all the soldered one parts and the touchbar. But states that he prefers to use MacOS. Then his last complaint was that he cracked his iPhone X and was expected to pay pretty much half the cost of the phone to replace it. And stated that it sucks that people have to pay that much for it. What exactly was he being not "honest" about? Edit: It's indeed possible he has a "bad unit" so I guess we'll see if other reviewers notice the issue. But given how little it seems Apple did to the case I don't doubt that the thermals for an i9 were not handled properly. Strong Sauce fucked around with this message at 05:53 on Jul 18, 2018 |
#
?
Jul 18, 2018 05:48
|
|
|
|
| # ? May 25, 2024 09:20 |
|
|
You're using it wrong! Macs were designed for _________ not _________. 
|
|
#
?
Jul 18, 2018 05:51
|
|
|
LionArcher posted:I am not shocked it’s Dave 2d, who uses premiere and made no attempt to test it with final cut. Totally fair use case 🤔 This is so far from actually paying attention that it almost counts as changing the subject. The only time he even flirted with presenting Premiere as an objective benchmark was when he showed that the same render took longer on the new i9 than it did on the old i7.
|
|
#
?
Jul 18, 2018 05:56
|
|
|
I am honestly surprised that they are not taking thermals more seriously considering the resources they have at their disposal. What the hell Apple, you can bond an LCD to a piece of gorilla glass flawlessly but you still can't figure out how to apply some loving thermal paste and heat sink to a CPU. 'Sorry guys we just can't figure that one out!'
|
|
#
?
Jul 18, 2018 05:58
|
|
|
The most embarrassing part of the video was the dongle needed to plug his mouse in.
|
|
#
?
Jul 18, 2018 06:03
|
|
|
Pakistani Brad Pitt posted:The most embarrassing part of the video was the dongle needed to plug his mouse in. Using a wired mouse to avoid the latency of wireless mice in twitch gaming: Not the Apple way
|
|
#
?
Jul 18, 2018 06:14
|
|
|
Wulfolme posted:Using a wired mouse to avoid the latency of wireless mice in twitch gaming: Not the Apple way Maybe it's a Logitech issue, but using bluetooth for my MX Master results in a lot of intermittent microstuttering which is borderline 'throw laptop out of window' irritating when you need even the slightest amount of precision or finesse in something like Photoshop or InDesign, so there is no alternative to using the USB A RF receiver. A USB C to A dongle for that piddly little unifying receiver sounds like a special kind of hell.
|
|
#
?
Jul 18, 2018 08:25
|
|
|
My MX master over Bluetooth works flawlessly and is the best mouse ever. Best $100 mouse ever.
|
|
#
?
Jul 18, 2018 08:26
|
|
|
KingEup posted:I am honestly surprised that they are not taking thermals more seriously considering the resources they have at their disposal. The throttling is neither surprising nor new, Macbook Pros are just not designed for this kind of duty cycle. Apple knows that the vast majority of buyers is going to use the device for facebook browsing and posting emojis on twitter so they optimize the hardware for those use cases, arguably rightly so. It's not just Apple either, almost all "Ultrabook" laptops behave this way under heavy load. My perfectly dust free and aftermarket repasted 2013 15" MBP throttled to 1.3-1.5 GHz when I once attempted run a game that was both CPU and GPU intensive. Increasing the resolution (=GPU load) would throttle the CPU further because the two chips share a heatsink and a power envelope. That's just what happens when you run a 40W CPU, 45W GPU plus screen, RAM, SSD and other components off a 85W power supply. I don't think this will ever change as long as MBPs run on x86, though it'd be awesome if videos like this go viral and make them to reintroduce the 17" which would have room for a 99,9Wh battery and much larger heatsinks. PS: This also illustrates why CTO CPU upgrades in MBPs are almost always a bad idea.
|
|
#
?
Jul 18, 2018 08:45
|
|
|
Boris Galerkin posted:My MX master over Bluetooth works flawlessly and is the best mouse ever. Best $100 mouse ever. It is very good, but the OG Razer Mamba was a better shape imo, fite me.
|
|
#
?
Jul 18, 2018 08:51
|
|
|
I'd like to see some Intel Power Gadget logs/graphs and not just video rendering times before coming to the conclusion that it's throttling.
|
|
#
?
Jul 18, 2018 09:02
|
|
|
There are plenty all over macrumors. If you ignore the people that run Prime95 small FFT with FMA3 (lol) and complain about throttling to 800 MHz, the 2.9 base throttles to 2.4-2.5 in Cinebench which I personally consider a fairly realistic max load. e: * Oops, only just now realized that his video rendering times on the 6 core are actually longer than on the 4 core.  e: It looks like the CPU throttles too aggressively under certain workloads like Premiere. It seems to bounce between 800 and 3200 MHz rather than just holding a steady frequency a few hundred MHz below baseclock. That's something that should be easily fixed via SMC/firmware update and would explain the overall performance loss vs. the old quadcore. eames fucked around with this message at 09:37 on Jul 18, 2018 |
|
#
?
Jul 18, 2018 09:14
|
|
|
It's kind of hilarious, really. Drop a hex core i9 K series chip into something as Thin And LightTM as a MacBook Pro - who is the bigger dummy: the people that actually made it happen or the people that bought it an expected Apple to have rewritten the laws of thermodynamics? Not gonna lie though, I'd scalp somebody to be able to drop an i9 in my cheesegrater Mac Pro. 
Theophany fucked around with this message at 10:52 on Jul 18, 2018 |
|
#
?
Jul 18, 2018 09:25
|
|
|
This is why they went to all usb-c, Apple is gearing us towards mineral oil submersion docking stations. You dunk your Mac into their Apple Juice tank and it slots into a usb-c connector at the bottom to connect to your monitor and keyboard. It uses their new Apple Cool refrigeration module to chill the oil and allow the cpu to work at full speed. Presto, no one wants a Mac Pro anymore. edit - they'll also put an Apple Watch crown on the laptop that you spin to pump the oil out when you need to take the laptop Starbucks.
|
|
#
?
Jul 18, 2018 09:52
|
|
|
That would actually be rad as hell if mineral oil wasn't an absolute nightmare to clean off.
|
|
#
?
Jul 18, 2018 09:59
|
|
|
Strong Sauce posted:Edit: It's indeed possible he has a "bad unit" so I guess we'll see if other reviewers notice the issue. But given how little it seems Apple did to the case I don't doubt that the thermals for an i9 were not handled properly. You don't doubt that because you are just assuming that it must be so much hotter than an i7. But it's not. The i9-8950HK is rated by Intel as a 45W TDP (Thermal Design Power) processor. That's exactly the same TDP as every CPU Apple has ever shipped in a 15" retina MBP, going all the way back to the original 2012 models. For Intel, 45W TDP is a promise to systems integrators (eg Apple) that, so long as they design their cooling system to handle a continuous 45W thermal load, a 45W TDP CPU will not throttle below its base clock frequency. (*) This means that actually, Apple didn't need to change a thing. That's the point of Intel's last 8-ish years of designs, particularly in the notebook space. Intel doesn't want to force systems integrators to do very much in order to keep up with Intel's desired ~1yr refresh cycle, it's to Intel's benefit if they can just keep basically the same chassis and slide a new CPU/logic board in, so Intel keeps TDP values for each class of notebook fairly constant. It's on Intel to figure out how to improve the performance possible inside that fixed thermal budget. Over the years, carefully done tests (eg Anandtech when Anand was still running it) have shown that Apple has done quite a good job of building a thin and light chassis which cools a 45W TDP CPU well enough that it shouldn't throttle. Simultaneous heavy discrete GPU load is another thing, but just the CPU by itself should be fine, despite all the people trying to dunk on Apple ITT because haw haw, so thin AND USELESS AMIRITE??? TLDR maybe don't freak out over one youtube video * - Except for a recent complication. Ever since Intel started introducing newer bigger wider vector extensions such as AVX2 and AVX512, they also started restricting clocks when those instructions are actually in use. This is true even in desktop and server CPUs. Speaking of which: eames posted:There are plenty all over macrumors. If you ignore the people that run Prime95 small FFT with FMA3 (lol) and complain about throttling to 800 MHz, the 2.9 base throttles to 2.4-2.5 in Cinebench which I personally consider a fairly realistic max load. If Cinebench is using AVX512, 2.4 to 2.5 is probably about right.
|
|
#
?
Jul 18, 2018 10:51
|
|
|
MSI reckon a peak TDP for the very same i9 chip is closer to 150W than 45W. The (similar, slightly lower clock) i7 part peaks out at closer to 90W; add an extra 400MHz for the part that Apple is actually using and you're at maybe 120W peak load TDP? https://www.msi.com/blog/why-8th-gen-core-i9-i7-and-i5-are-real-deal-and-faster-than-core-i7-7700hq If MSI's own testing (which they are confident enough in to print) is accurate, Intel's TDP figures are about as useful as Volkswagen's fuel consumption figures. edit: that's ignoring the 'look how much cooling we're throwing at this' chest-puffing MSI is doing. How can anybody realistically expect that same heat death CPU part to go in a MBP and not throttle itself? edit2: Intel's product page posted:Thermal Design Power (TDP) represents the average power, in watts, the processor dissipates when operating at Base Frequency with all cores active under an Intel-defined, high-complexity workload. Refer to Datasheet for thermal solution requirements. Well that sure smells like non-specific bullshit. Theophany fucked around with this message at 11:16 on Jul 18, 2018 |
|
#
?
Jul 18, 2018 11:13
|
|
|
BobHoward posted:
Cinebench doesn't use any AVX or FMA instruction set. Theophany posted:
All of this is really Intel's fault, I posted some numbers over in the Intel CPU thread if you are interested. Looking at temperatures and power consumption in Cinebench it's pretty clear that Apple actually improved the thermal design of their 2016 15" MBPs while simultaneously making the devices thinner, but Intel just keeps hiking actual power consumption without updating the rated TDP.
|
|
#
?
Jul 18, 2018 11:26
|
|
|
eames posted:All of this is really Intel's fault, I posted some numbers over in the Intel CPU thread if you are interested. Looking at temperatures and power consumption in Cinebench it's pretty clear that Apple actually improved the thermal design of their 2016 15" MBPs while simultaneously making the devices thinner, but Intel just keeps hiking actual power consumption without updating the rated TDP. They seem to be pretty much what I'd expect. The the i9 is a 6C/12T chip, just like the i7 but with higher clock and boost frequencies so the idea that they have the exact same TDP just doesn't make sense. My understanding is that the major difference between the two is that i9 has a much richer platform feature set, not a physical difference that makes it more thermally efficient for a given workload.
|
|
#
?
Jul 18, 2018 11:40
|
|
|
Boris Galerkin posted:My MX master over Bluetooth works flawlessly and is the best mouse ever. Best $100 mouse ever. 
|
|
#
?
Jul 18, 2018 11:56
|
|
|
German site showing the 13" doing the same thing in Cinebench https://www.notebookcheck.com/Test-Apple-MacBook-Pro-13-2018-Touch-Bar-i5-Laptop.316002.0.html The 2016's weren't known for this, right? I can't find a good review like Anandtech used to do. I don't see how Apple would just go OH YEA WE NEVER ACTUALLY TESTED THIS
|
|
#
?
Jul 18, 2018 13:00
|
|
|
Bob Morales posted:German site showing the 13" doing the same thing in Cinebench It's almost like Apple doesn't really care about making a professional laptop anymore.
|
|
#
?
Jul 18, 2018 13:13
|
|
|
Same chip (i9) throttles in the XPS 15, video by the same guy: https://www.youtube.com/watch?v=18PA3WFFuAo Redditor posting some graphics https://www.reddit.com/r/Dell/comments/8u3ir1/dell_xps_15_9570_with_core_i9_thermal_throttling/
|
|
#
?
Jul 18, 2018 13:25
|
|
|
Bob Morales posted:Same chip (i9) throttles in the XPS 15, video by the same guy: Dell also kept the heatsink design from the previous gen XPS which had throttling issues. Which doesn't seem smart.
|
|
#
?
Jul 18, 2018 13:28
|
|
|
FWIW, I have a Lenovo T480S at work I just downloaded Cinenbench, and ran it twice Multithreaded I got a 505, second run it went down to 449. That's about 10%. It held a steady 2.35GHz clock speed, max turbo on this model is 3.4GHz Single-threaded it holds 3.38GHz But in neither test does it bounce around all over the map. I don't know what CPU's our P51's have in them or I'd check them. i7 of some sort though.
|
|
#
?
Jul 18, 2018 13:48
|
|
|
Really disappointed with what Apple is offering in their new Macbook pros. I was looking to get one relatively soon but the specs are so weak in the base models, especially for what you pay. I see these new Dell, Asus, and Chinese laptops coming out with top-of-the-line specs for $300-$800 less than the new MBPs, and with similar build qualities. I bought the MBP 2016 and was severely disappointed with its overheating issues and I thought the touch bar wasn't for me, so I returned it after about a week. I don't feel like these new gen MBPs really offer much more, especially for the (at this moment) crazy price point. I'm holding out hope for that new low-cost Macbook that is rumored to be coming, but unless I really, REALLY need what the new MBP is offering (I'm currently satisfying my needs with a 2013 MBA) I guess I'm gonna have to keep waiting. Verisimilidude fucked around with this message at 14:04 on Jul 18, 2018 |
|
#
?
Jul 18, 2018 14:00
|
|
|
Verisimilidude posted:Really disappointed with what Apple is offering in their new Macbook pros. I was looking to get one relatively soon but the specs are so weak, especially for what you pay. I see these new Dell, Asus, and Chinese laptops coming out with top-of-the-line specs for $300-$800 less than the new MBPs, and with similar build qualities. What specs are 'weak' on the new one? 32GB, i9, super loving fast SSD up to 4TB... Asus/Chinese laptops (and Dell to a point) are poo poo.
|
|
#
?
Jul 18, 2018 14:03
|
|
|
Bob Morales posted:What specs are 'weak' on the new one? The 800Mhz i9 
|
|
#
?
Jul 18, 2018 14:04
|
|
|
edited to reflect that I was speaking about the base models. I don't have $3000-$7000 to spend on a laptop.
|
|
#
?
Jul 18, 2018 14:05
|
|
|
Bob Morales posted:German site showing the 13" doing the same thing in Cinebench I can't read German but the screenshot in that article indicates it isn't throttling.  Brief 50W TDP and Turbo Boost until it hits Tmax, then back down to TDP and what seems to be above-nomimal clocks.
|
|
#
?
Jul 18, 2018 14:06
|
|
|
~Coxy posted:I can't read German but the screenshot in that article indicates it isn't throttling. You're right - it's just slower, though. The result in the Cinebench R15 single test is still in the expected range, but when we saw the result of the multi-test, we were surprised. Based on our previous experience with the Kaby-Lake-R quad-cores we had expected a value of more than 700 points, but there are just 608 points (tested under macOS). Thus, the test device falls behind systems such as the XPS 13 from Dell or the ThinkPad X1 Carbon . Since many manufacturers also control the TDP dynamically, even many models are faster with the very normal Core i5-8250U 
Bob Morales fucked around with this message at 14:11 on Jul 18, 2018 |
|
#
?
Jul 18, 2018 14:07
|
|
|
you guys can gaze at literally any other laptop manufacturer's computers and think thegrass is greener all you want but like. for the hundreds of things i'd change about apple laptops, they're still miles ahead of everyone else in terms of manufacturing computers i wouldn't be embarassed to be seen somewhere with, in addition to being like pretty much the only company to have the balls to do away with all external plastic.
|
|
#
?
Jul 18, 2018 14:08
|
|
|
Can anyone tell me a 'will not burn down my house' version of an Apple 45w Magsafe 2 power adapter?
|
|
#
?
Jul 18, 2018 14:13
|
|
|
landgrabber posted:you guys can gaze at literally any other laptop manufacturer's computers and think thegrass is greener all you want but like. for the hundreds of things i'd change about apple laptops, they're still miles ahead of everyone else in terms of manufacturing computers i wouldn't be embarassed to be seen somewhere with, in addition to being like pretty much the only company to have the balls to do away with all external plastic.  I think that's why people care so much. They know that there's no decent alternative so if there's something (anything — keyboard, touchbar, ports, now throttling) "wrong" with the MBP they'll be stuck with it for the next few years. By the way that screenshot of the 13" in Cinebench is almost 100% identical to my 15" from 2013, the only difference is the TDP line at 28W instead of 47W. I think the current 13" is a quite impressive package, to me more so than the new 15". eames fucked around with this message at 14:18 on Jul 18, 2018 |
|
#
?
Jul 18, 2018 14:16
|
|
|
Yes Apple hardware has a huge upcharge but apparently that's what it costs to get a laptop that isn't a creaky plastic hunk of trash. Assuming you're okay with keyboards losing keys to dust or 800mhz processors.
|
|
#
?
Jul 18, 2018 14:17
|
|
|
xzzy posted:Yes Apple hardware has a huge upcharge but apparently that's what it costs to get a laptop that isn't a creaky plastic hunk of trash. i don't even know if it's that much of an upcharge in the context of like... high quality glass and anodized aluminum are expensive. look at any custom PC case - if you want something as robust as a cheesegrater or imac you're basically paying an additional thousand bucks anyway.
|
|
#
?
Jul 18, 2018 14:23
|
|
|
landgrabber posted:you guys can gaze at literally any other laptop manufacturer's computers and think thegrass is greener all you want but like. for the hundreds of things i'd change about apple laptops, they're still miles ahead of everyone else in terms of manufacturing computers i wouldn't be embarassed to be seen somewhere with, in addition to being like pretty much the only company to have the balls to do away with all external plastic. agreed completely I hate windows with a passion, and my job requires me to use a macbook, so when I see these changes I think about how my productivity is going to be affected. My biggest worry, and this doesn't seem to be fixed this generation, is with overheating. The overheating on the MBP 2017 I purchased was loving terrible. My battery would die after 4 hours at a relatively low brightness while doing basic work and internet use, and the chassis would get so uncomfortably hot that I'd have to put the laptop away. This is why I get so annoyed with Apple products. That premium is getting further and further beyond my reach, and there is no other dog in town.
|
|
#
?
Jul 18, 2018 14:29
|
|
|
Dell's website is stupid so you can't really build a Dell like you can a Mac. Apple tax gets fucky if you upgrade past 1TB SSD. i9 upgrade is $400 Dell + $350.00 32GB upgrade is $400 Can't choose it on it's own apparently 512GB SSD storage + $200.00 Dell + $200.00 1TB SSD storage + $600.00 Dell + $400.00 2TB SSD storage + $1,400.00 Dell + $800.00 4TB SSD storage + $3,400.00 Dell N/A (Samsung MSRP $1800)
|
|
#
?
Jul 18, 2018 14:30
|
|
|
Bob Morales posted:Dell's website is stupid so you can't really build a Dell like you can a Mac. Apple tax gets fucky if you upgrade past 1TB SSD. To be fair tho, the Dell XPS 15 at its highest baseline is $2150 and comes in at 1 TB SSD, 32 GB RAM and 8th Gen i7 6-core You definitely do get more bang for your buck with Dell, but at the cost of build quality (and windows)
|
|
#
?
Jul 18, 2018 14:38
|
|
|
|
| # ? May 25, 2024 09:20 |
|
|
Verisimilidude posted:agreed completely yeah, i should probably clarify - i don't think these computers are really worth it or the premium. it's just that apple gets accused of this robbery bullshit - which they do - but all the other PC and laptop manufacturers do the same poo poo while being chinsey-er and even more embarassing. i mean jony ive is hailed as the Design Genious or whatever when at the height of his career he was just doing the "change my homework slightly so the teacher doesn't find out" meme to dieter rams's poo poo
|
|
#
?
Jul 18, 2018 14:50
|
|