Puget Systems print logo

https://www.pugetsystems.com

Read this article at https://www.pugetsystems.com/guides/1250
Article Thumbnail

Lightroom Classic CC 2018: Core i7 9700K & i9 9900K Performance

Written on October 19, 2018 by Matt Bach
Share:

Introduction

Lightroom Classic CC is much better at using a higher number of CPU cores than its predecessor, but for many tasks the speed of each core is often more important than the total number of cores. With the launch of the new 9th Gen Intel Core Processors, Intel has made a number of improvements including a small frequency bump and an increase in core count. Where the current top-end consumer Core i7 CPU (the Core i7 8700K) has only 6 cores and a max Turbo of 4.7 GHz, the Core i7 9700K and i9 9900K both have 8 cores and a 4.9-5.0 GHz max Turbo frequency. Since Lightroom tends to value both core count and frequency, these new CPUs should do great but the only way to know for sure is to actually benchmark them and find out.

If you would like to skip over our test setup and benchmark result/analysis sections, feel free to jump right to the Conclusion section.

Test Setup & Methodology

Listed below are the systems we will be using in our testing:

While benchmarking the i7 9700K and i9 9900K against the i7 8700K is likely the most direct comparison we could make, we also wanted to see how these new CPUs stack up against a number of other processors. AMD vs Intel is always a popular discussion, so we included the Ryzen 7 2700X - which tends to be cheaper than either of these new CPUs - as well as the Threadripper 1920X which is similar in price to the i9 9900K. To get an idea of whether or not purchasing a more expensive Intel CPU would give you a notable increase in performance, we also include the i7 7820X and the i9 7900X.

To thoroughly test each processor, we will be using two sets of images: one set of 22MP.CR2 RAW images taken on a Canon EOS 5D Mark III and a set of 45MB .NEF RAW images taken on a Nikon D850.


While our Lightroom testing is still evolving, we are currently able to accurately benchmark the following tasks with both sets of images:

Bulk Tasks

  • Import
  • Export to JPEG
  • Build Smart Previews
  • Convert to DNG

Module Tasks

  • Library Module Loupe Scroll
  • Develop Module Loupe Scroll
  • Library to Develop switch

Photomerge

  • Panorama Merge
  • HDR Merge

We are currently working on putting up an alpha version of our benchmark for public download (similar to our Photoshop Benchmark) which will have a much more in-depth description of each of these tests. Keep an eye out in the coming weeks (or months)!

Benchmark Results

While our benchmark presents various scores based on the performance of each type of task, we also wanted to provide the individual results in case there is a specific task someone may be interested in. Feel free to skip to the next section for our analysis of these results.

Benchmark Analysis

The scores shown in the chart above are relative to the best possible performance for each task with a Core i7 8700K CPU and a NVIDIA GTX 1080 Ti 8GB GPU. In essence, a score of "90" would mean that it gave 90% the performance of the reference system while a score of "110" would mean it was 10% faster.

In the module tasks (scrolling through images and switching between the Library and Develop modules), there was surprisingly little difference between all the Intel CPUs we tested, although the AMD CPUs lagged behind just a bit. The new Core i9 9900K was at the top, but it was only a few percent faster than the i7 8700K which is likely within the margin of error for those tests. However, both the i7 9700K and i9 900K were definitely a bit faster in photo merge tasks, coming in at 4-6% faster than the i7 8700K

The bulk tasks like exporting and generating smart previews are where we expected these CPUs to shine and while they were ~10% faster than the i7 8700K, we honestly thought we would see a larger difference. They certainly compare favorably against the more expensive i7 7820X, but if exporting is a major consideration at this price point than you may be better off with the AMD Threadripper 1920X.

Are the 9th Gen Intel Core Processors good for Lightroom Classic?

The new i7 9700K and i9 9900K are certainly good for Lightroom Classic CC, but they are only about 5% faster than the i7 8700K on average. In fact, for most users there is little reason to use the more expensive i9 9900K as the i7 9700K is only a tiny bit slower.

9th Gen Intel Core i7 9700K & i9 9900K Lightroom Classic Benchmark
If you are looking for a new workstation at roughly this price point, there is no reason not to use the i9 9700K over the i7 8700K, especially considering how well it performs in other applications like Photoshop. However, unlike Photoshop, there is probably no reason to upgrade if you already have an 8700K since you are unlikely to notice a difference unless you are actively benchmarking Lightroom.

Core i7 9700K vs Core i7 8700K for Lightroom Classic

On average, the Core i7 9700K is about 4% faster that the Core i7 8700K in Lightroom Classic. You should notice the biggest difference in tasks like exporting and generating previews, but when navigating around the Library and Develop modules there is very little difference.

Core i9 9900K vs Core i7 8700K for Lightroom Classic

The Core i9 9900K is approximately 20% more expensive than the Core i7 8700K, but we only saw about a 6% performance increase in Lightroom Classic. It is technically the fastest CPU we have tested for navigating around the modules and photo merge tasks, but if you are looking for the best export performance you may be better served with a Threadripper 1920X or a Core i9 7900X or higher CPU instead.

If you are interested in how the 9th Gen Intel Core Processors perform in other applications, be sure to check out our recent Processor articles as we have a number of other articles for looking at the i7 9700K and i9 9900K.

Tags: Lightroom Classic, 9700K, 9900K, 8700K, 7820X, 7900X, 2700X, 1920X
Behive

I really wonder if the Z370 motherboard you used negatively affected the 9th gen results? Thanks for pushing so quickly!

Posted on 2018-10-19 14:04:10

It shouldn't. We did some comparison testing with a Z390 board and the results were pretty consistent. We just already had a bunch of results already on Z370 before that board came in so we stuck with it rather than having to re-run a bunch of testing.

Posted on 2018-10-19 15:12:15
Behive

Thanks Matt. I watched some other 9900K reviews and it seems like a Z390 makes very little difference or none.

Posted on 2018-10-19 21:32:58
Bryan

Matt, I know that it is impractical for you guys to test at overclocked speeds, but how do you think the 8700k will compare against the 9900k and 9700k when all are overclocked? I currently have a 8700k running at 5.0ghz all core. When I ran your Photoshop benchmark a few months ago, I achieved a score of 1062, which is not far behind your score for a stock 9700k.

The 9th gen chips have a higher stock turbo compared to the 8700k, but other reviews have found that the 9th gen 8-core chips also run significantly hotter, which reduces maximum overclocking potential. In real world use, the 8700k can be clocked 100-200mhz faster for the same temperatures, so a fair comparison might be a 8700k @ 5.0ghz and a 9700k/9900k @ 4.8ghz.

I think that your findings of a 16% and 6% performance benefit for the 9900k for Photoshop and Lightroom may be overstating what real world results are going to be. I was excited when the 8-core chips were announced, but I can't justify making an upgrade due to the launch price and thermals.

Posted on 2018-10-20 17:30:24
Reid

Bryan,

Good point about the overclocking and thermals. I'm looking to upgrade my system and am looking at the i7-8700K, i7-9700K, or i7-8086K. I'm wondering if the 8086 is the sweet spot as it may overclock better without the thermal issues, yet is faster than the 8700. I hate noise, I only have one fan in my system and that's the CPU fan on an oversized heat sinc. Open to suggestions.

As this article states, the "the Core i7 9700K is about 4% faster that the Core i7 8700K" so the 8086 would be even less of a difference, but cost $30 more than the 8700. Maybe the 8700 is still the best bet?

Posted on 2018-10-21 16:43:31

I don't recommend overclocking, but if that is your goal you will have much better results with the 9000 series because of the better thermal interface material it has compared to the 8000 series (including the 8086).

Posted on 2018-10-21 17:35:50
Reid

I'm not too concerned about the overclocking. I've let the Asus software do it and haven't had issues with it. I understand your reasoning about the new CPUs having better thermal interface, which I guess they need because the run hotter. Given the three I'm looking at and considering heat, thread, and clockspeed, would you still recommend the i9700? https://www.newegg.com/Prod... :

Posted on 2018-10-21 19:37:44
Bryan

The 8086k is the exact same chip as the 8700k, so I would not buy it unless it is the same price or cheaper than the 8700k. Again, my personal take on these results is that the performance differences are largely due to the different stock turbo frequencies of the 8th gen and 9th gen chips. Once you overclock and take these differences out of play, the performance difference will decrease or disappear all together.

The thermal differences are an additional confounding factor, as you may be able to clock an 8700k or 8086k higher than a 9700k or 9900k. This further reduces any potential performance gains by the 9th gen chips, as you are trading clock speeds for extra cores.

Contrary to William's comment about the 9th gen being able to overclock higher due to having better thermals, testing has shown this not to be the case. Watch this YouTube video for a full explanation of why soldering the IHS has not helped as much as people have been expecting: https://youtu.be/JTAAXCpNhxM

Posted on 2018-10-21 20:04:09
Reid

Well, that video was quite an earful. Correct me if I’m wrong, but the Takeaway seems to be that there is not a huge advantage to the 9700 over the 8700 especially when you include the factors of heat and price

Posted on 2018-10-21 20:59:18

So far as I could tell his testing in that video was limited to the 9900K, and without hyper-threading the 9700K should actually run quite a bit cooler. Considering the additional clock speed on top of that, if you're not overclocking, I think it comes back down to hyper-threading. If you're using software that doesn't utilize hyper-threading well then the 9700K's extra cores and clock speed will make it a better value. If your software does use hyper-threading effectively, it'll be a more even match.

One additional thing to note though, the new 9000 Series CPUs may be somewhat hard to find for a while. This is frequent with new hardware generations, but appears to be more of an issue right now than I remember in years past.

Posted on 2018-10-22 06:06:59

Interesting, Jayz2cents had much better results oc'ing the 9900K compared to the 8700K: https://youtu.be/9yQRmbe2QPU

I don't OC myself, and there isn't much headroom on any of these chips. I think stock speeds are solid these days, especially because of Turbo Boost. The all-core and single-core turbo speeds on these 9000 series processors match or exceed the 8000 series, while having two more physical cores. If your software benefits from Hyperthreading, the 8700K / 8068K might still be faster than the 9700K - but if not, that is what I'd go for (out of those three options).

Posted on 2018-10-21 21:55:57
Bryan

I think that Jay may have received a golden sample, and he said so himself towards the end of the video. 5.0ghz @ 1.31v is very good, as most copies will require 1.35v to be stable at 5.0. The TIM is definitely better on the 9th gen, but the thicker silicon with the extra two cores result in overall worse performance. If you look across all reviews, you will find that most 8700k's will hit 5.0ghz on a standard air cooler (like the NH-U12S that Puget uses on their benches), while most 9700k's and 9900k's will run into a thermal limit at 4.8ghz on the same cooler.

The 8700k has a stock all-core turbo of 4.3ghz, which means that it actually has significant overclocking headroom. It can easily be clocked 700mhz faster. In comparison, the 9700k has a stock all-core turbo of 4.7ghz, so it is nearly maxed out at stock. Keep in mind that base clocks and turbo clocks are arbitrary. On the architecture level, a 8700k core is exactly the same as a 9700k core (edit: with the exception of hyperthreading), so there is no objective reason why the 9700k should be 400mhz faster out of the box.

Puget's testing methodology is a bit problematic because they are comparing a 4.3ghz 8700k against a 4.7ghz 9700k. Regardless if you choose to manually overclock or not, the 9700k is overclocked higher out of the box than the 8700k, so the comparison is unfair. At stock, the 9700k will be using more voltage, more wattage and will run hotter than the 8700k. To make a fair comparison, you have to control for these variables and test either at the same clock speeds, or same voltage, or temperatures. In real world use, the limiting factor will be thermals for all of these chips, so that will be the most valid comparison. However, I do understand that Puget are system integrators, and their primary interest is system stability, so it's not likely that they will test under these conditions.

@Reid: yes, that would be my personal opinion. All else being equal in your system, I believe that there will be no meaningful difference between the 8700k and 9700k once you optimize and overclock both CPUs.

Posted on 2018-10-21 23:09:31

I could see how it might be unfair if you were trying to compare purely the maximum performance you could get across different CPUs when overclocked to their limits. And I know that some of the folks in videos like the one you posted above have that in mind. However, I personally think that the ideal situation will be when Intel is putting out CPUs that offer maximum performance without needing any more overclocking. I think we're getting close with this generation, and you can see that by the fact that there is so little overhead to overclock further without having to crank the voltage up and deal with massive amounts of heat. That may take the fun out of things for some people who enjoy overclocking, but the best case situation to me is one in which everyone can get the maximum performance without having to fiddle with motherboard settings :)

Posted on 2018-10-22 06:02:38
mclaren777

Your team just saved me $488. ;)

Posted on 2018-10-23 05:04:02

Hi, may I ask how big was the set of the 5D Mk III RAW files you used for export? Are these 24 files we can see on the LR screenshot? I'm an event photographer and I'm primarily interested in the fastest export time within sensible price range of course. I don't use Lightroom, but Adobe Camera RAW only which I'm accessing via Adobe Bridge. My current config is i7 6700K oc to 4.4GHz / GTX 750 / 16GB RAM / m.2 ssd for system / sata ssd for Adobe cache / hdd for photos. I just tested the time needed to export 24 raw files to jpeg from some of my recent jobs and got results between 55 sec to 65 sec - so very similar to what you scored for the 9700K - 9900K. So, my question is - how much I will benefit (in terms of exporting time) from upgrading to one of theses cpus? Can you give me a rough estimate? Thanks!

Posted on 2018-10-27 11:09:06

We make copies of the photos so that we have 100 images to export. We've done some testing with more photos and while it took longer to complete, it seems to be pretty pretty linear after you have more than just a few to export. If you got the same time with 24 files, then there is probably a difference in export settings. We're hoping to get our benchmark polished up for public download at some point, at which point you could more closely replicate our testing.

That said, I would expect your system to export the same images/settings we used in somewhere around 70-80 seconds. So a 7900X should be roughly twice as fast for exporting.

Posted on 2018-10-29 17:30:28
Ludo

Thank you so much for this!!!

Posted on 2018-11-03 19:35:19
Ukraine

Is it possibble to know average CPU utilization during the test? I mean import/export images, especially export. Because it's very, very important - In the time, when Lightroom 5.7.1 utilize I7-2600K on 70-80% during export, Lightroom 7-8 utilize it on 99%. And I just wondering, does the Ryzen R72700X have the same CPU utilization - about 100%, or less (may be 90%..or may be even 80%?). Because on my I7 I can't do anything like Photoshop during export, because it's too slowly; when I have to do something heavy during export, I have manually reduce core utilization for Lightroom for 1 or 2 core.

Posted on 2018-11-08 21:14:44

CPU utilization we typically don't log during these benchmarks since from a performance perspective, it is often more misleading then helpful. If you are hitting peak CPU utilization, that is actually good from a performance perspective, but I totally understand how it can cause issue with multitasking. What I would recommend is using a piece of software like System Explorer http://systemexplorer.net/ . With it, you can set the affinity (how many cores it can use), but it also has the option to make it permanent. That way, anytime you launch Lightroom it can automatically have the affinity set to leave 1 or 2 cores unused for multitasking.

Posted on 2018-11-08 21:22:46
Ukraine

Thank you for response and advice! It will be helpful.
But my main thought, or assumption was such: May be CPU utilization in AMD Ryzen 2700X during export (which may take many hours for thousands of pictures) not 99%, as in I7 processors (definitely for I7-2600K), but just 60-70-80 % ? I have seen issue with Ryzen 1800X, where CPU utilization during export was just about 30%. https://feedback.photoshop....
I understand, it's rather atypical issue, but may be really architecture of this Ryzen processors such, that they show this results, as in your benchmark, and in the same time remains 10-20-30% unused CPU power.
if this assumption (partially) true, it can change dramatically CPU preferences.

Posted on 2018-11-09 14:03:52

It looks like that issue was before Lightrooom Classic was launched - they really improved performance in that new version of Lightroom. Between AMD and INtel, if you have a similar number of cores I don't think you should see all that different of CPU load unless there is an issue with your system.

Posted on 2018-11-09 18:14:45
Cantharis

I am currently looking for a new computer and don't know which CPU to take. At first I wanted to take the i7 8700, but then considered the i7 9700k, because it's so much faster at building smart previews. But I am worried, that it's not that future safe, since it doesn't offer hyperthreading. Next time I will be upgrading is in 5-7 years.
Why is the i9 9900k slower at creating previews than the 9700k? It is almost at 8700k level.

Thanks for your reply

Btw: I am coming from an it 2500k...

Posted on 2019-02-13 03:10:57
Michael

doubble post - sorry

Posted on 2019-02-13 05:27:22
Michael

I am currently looking for a new computer and don't know which CPU to take. At first I wanted to take the i7 8700, but then considered the i7 9700k, because it's so much faster at building smart previews. But I am worried, that it's not that future safe, since it doesn't offer hyperthreading. Next time I will be upgrading is in 5-7 years.
Why is the i9 9900k slower at creating previews than the 9700k? It is almost at 8700k level.

The 8700 is currently 100 € cheaper than the 9700k, is it worth 100 € more?

Thanks for your reply

Btw: I am coming from an it 2500k...

Posted on 2019-02-13 07:01:35