Firefox Quantum vs Chrome TESTED! Things Just got seriously close

Firefox has always been my favorite browser. Its rich choice of add-ons (native), its stringent privacy policies and its strict stands for an open web have always been the reason I used it. But it was slow. Not very slow, but slow nevertheless. But the new Quantum engine aims to change that! I’ve used since day 1 and only one thing I can say, it’s fast! Man, it’s fast!

Before posting my biased opinions and unbiased benchmarks, I would like to tell that, yes, I prefer Mozilla. Mozilla has always been one of my favorite companies (tesla also shares that now), due to its philosophy. I doubt that any other company has ever built a lasting impression (and revenue) being totally free, open source and transparent. I use Thunderbird, pockets, Firefox and tried its Boot2Gecko platform (Firefox OS) as well. Its products aren’t the best, aesthetically speaking, but they’re still functional and respects the privacy I want.

Mozilla Claims Firefox to be 2x faster on Speedometer

Firefox Quantum: What’s new

Well, Firefox Quantum has been making ripples in the tech industry for the last few days. And it definitely has some good reasons to brag about. Since 2016, the entire Mozilla and Open Source community has continuously worked and re wrote some fundamental parts of Firefox! And now some very essential parts like Stylo is the new CSS Layout Engine, the Rendering engine isĀ  WebRender (not WebKit), The compositor, DOM and flow everything is revamped and fresh.

Which means, the JavaScript Engine remains SpiderMonkey, which is a con. SpiderMonkey is slower than V8 and a JavaScript benchmark should confirm that. But Mozilla is taking steps for that. Like a blog post says that Mozilla will switch to HolyJIT (now still in infancy) for JIT compilation. (Currently SpiderMonkey has 15000 hand crafted lines of code vs 36000 of V8). TL;DR: In JavaScript performance, Chrome still keeps the edge, but Mozilla seems determined.

TEST 1: REAL WORLD

Without knowing the testing process, the test isn’t good! In the browsers, there is an Inspector where we can monitor the network. To test it, I have picked a few websites, they are

  1. Google: http://google.com
  2. Alexa: http://alexa.com
  3. Facebook: http://facebook.com
  4. Twitter: http://twitter.com
  5. Quora: http://quora.com
  6. IAMSohan: http://iamsohan.in
  7. YouTube: http://youtube.com
  8. Neematic: http://neematic.com
  9. LGDB: http://lgdb.org
  10. Mozilla: http://mozilla.org

However, sites like Facebook, Twitter, Quora are constantly loading content via Ajax and I can’t measure time if I want them to load fully. I noted the finish time as soon as the visible content was functional. The first test was best of three. NOTE: My connection was erratic. Sometimes a resource loaded in 300ms, sometimes it took 700ms. So, they contain some error due to network. Which is fine because the error is random, and affects both browsers (Firefox and Chrome), also it is present pretty much everywhere. I’ll do the second test with cache enabled so it won’t present any issues there.

RESULTS:

Firefox 57 Quantum on 64bit Linux. 2.4GHz Dual core/4GB DDR3
Chrome 62.0.3202.94 on 64bit Linux. 2.4GHz Dual core/4GB DDR3

As you can see that most of the sites have indeed loaded faster in Firefox, with the exception of Twitter, Neematic, LGDB and Mozilla. It’s ironic that Google loaded Mozilla faster and Firefox loaded Google faster. Neematic was much slower on Firefox Quantum. Mostly because it depends on heavy JavaScript, which remains weak on Firefox.

But I can’t stop admiring how Firefox has closed the gap. 5/10 wins that too having a lead in major sites like Google, Facebook and Quora and YouTube. I intentionally put unpopular sites, speculating that Firefox might have optimizations in specific sites/applications, which is still a win for the user. Or the sites may be optimized for Firefox. Either way, user wins and Firefox wins.

VERDICT

By Finish Time: Firefox 5-6 Chrome

By DOM Load Time: Firefox 4-7 Chrome

Opinion: TIE. Really there’s nothing of a significant difference in the real world. Maybe if I had 91Gbps internet, I might notice some difference, but they both look smooth and that’s a significant improvement for Firefox.

TEST 2: JavaScript Resource usage

Well, I know the JS Engine haven’t been updated, but I still wanted to give it a shot, because if the JS Lags behind, other differences will matter less. A significant drawback in JS will even out the advantages. So, I chose JetStream as the benchmark. And the priority was resource usage. The differences in time will be minor, very minor so, I thought of testing the resource usage instead. Here’s the results.

Chrome JetStream Result
Firefox JetStream result

Here the difference is minor, Firefox has 91.6s score while Chrome scores 89.9, and it isn’t quite significant, and honestly, It might seem that V8 isn’t superior to SpiderMonkey after all. But no. Here’s the resource report

Sensor data while running the benchmark

RESULTS

Here, you can see that Firefox used significantly more resources, resulted in heating that reached a high of 76C. Not just that, if you take the image, draw a straight line from the peak temperature of chrome, you can find 3 peaks that crosses that line. And in CPU usage, Chrome used almost 30% less CPU. That’s not it, Just look at the FREE RAM. At times, Firefox consumed the entire available RAM. while Chrome had around 800MB RAM FREE all the time.

And even with reduced consumption of resources and 1C less CPU heat, Chrome scored only 2 points less. That’s V8 for you, folks!

VERDICT

Chrome just blew Firefox out of the equation here. The JIT of Firefox probably kept firing, consuming RAM like crazy, which means the crazy ass optimizations of the JIT in V8 easily overpowers that of SpiderMonkey. That’s good because, whatever Quantum has changed, was indeed significant because it practically closed the gap. But if you need raw performance, Firefox reigns!

TEST 3: Cached results

While, the uncached results are good, in reality, everyone would have cached content. So, I used all these sites extensively to make sure the visible content is cached as much as possible. The sites, on average loaded 2x faster consistently. The results still had variance, but since I took 3 reloads, the second and third run were identical as the cache was fresh.

RESULTS

Firefox 57 Quantum, Cache Enabled
Chrome 62.0.3202.94 Cache Enabled

VERDICT

By Finish Time: Firefox 5-6 Chrome

By DOM Load Time: Firefox 8-3 Chrome

I would give it to Firefox. Because as the DOM Loads were faster, and Firefox did retain lead in major sites as Google, Facebook, Quora, YouTube, etc. Which means, in real world usage, the apparent speed will be better in Firefox. Here both loaded 2x faster with cache enabled and the difference was noticeable and in favor of Firefox.

TEST 4: RAM Usage 5 Tabs

5 Tabs being simultaneously open, they were

  1. Facebook
  2. Twitter
  3. Quora
  4. YouTube
  5. Google Docs

Each of them are heavy on their own right. So, after opening 5 of them, I checked the System Monitor, averaged the results over 20 seconds, and here it is.

RESULTS

FIREFOX: 421MB

CHROME: 484MB (~15% more)

VERDICT

Chrome uses separate processes for separate tabs, which means it’s only prioritization is thread priority, and it can’t do a lot of tricks like tab prioritization etc. So, even when we see a smaller memory footprint (30%) in our Benchmark run, in real world usage, Firefox trumps Chrome in RAM Consumption. As the results were same with 10 tabs as well, so, I thought of omitting those results.

Conclusion

Firefox Quantum is a new technology and most sites have no optimizations for Firefox, there’s too much new features and changes that the internet will take about 6-8 months to optimize for it (that is, if it gains market share, else the time might be higher) but the changes are excellent, speeds have caught up to Chrome, resource usage has decreased, and a difference is noticeable only in benchmarks.

Chrome still feels smooth in a lot of sites, but so does Firefox in others, and due to Firefox’s superior privacy policies and Free and Open Source nature, I will not hesitate to recommend Firefox Quantum.

Recent Posts

Recent Comments

Archives

Categories

Meta

Be First to Comment

    Share your valuable feedback