Google last week said that this summer Chrome will remove resource-hogging web advertisements from websites, including ads that mask unauthorized crypto-mining operations.
Arguing that a very small number of online ads – three-tenths of a percentage point of all on the web – disproportionately account for major portions of total network and CPU consumption, Google plans to scrub sites of such ads starting with a Stable build of Chrome near the end of August.
Chrome 85 is scheduled to release on Aug. 25, and would be the most likely version to debut the feature.
“These ads (such as those that mine cryptocurrency, are poorly programmed, or are unoptimized for network usage) can drain battery life, saturate already strained networks, and cost money,” Marshall Vale, a Chrome product manager, wrote in a May 14 post to the Chromium blog.
When Chrome detects one of the über-aggressive ads, the browser will unload the content from the ad’s frame – the portion of the page in which it’s displayed – and refill the space with an error message stating “Ad removed,” along with a link to more information.
After copious measurements, Google decided to strip out any ad that consumed 4MB of network data, used the CPU during half of any 30-second span or tallied a total of 60 seconds of CPU usage. Those bars were so high that they affected only 0.3% of all ads, but, Google contended, such ads accounted for 27% of all ad-generated network traffic and 28% of all ad-related CPU usage.
Other browsers have addressed bad actors, including crypto-miners, using different approaches. Mozilla’s Firefox, for example, blocks crypto-miners by targeting domains known to harbor such scripts. (Mozilla relies on domain lists produced by Disconnect.) It’s not surprising that Google took a metrics-based route here; it typically bases decisions, or says it does, on data collected by Chrome and/or its search engine.
Chrome users can try out this “Heavy Ad” detection and removal prior to version 85 via the chrome://flags option page. Set the Heavy Ad Intervention flag (which also goes by #enable-heavy-ad-intervention) to Enabled and relaunch the browser.
Site developers and ad content creators should use the time between now and late August to test and, if necessary, alter first-party advertisements, Google said. “Our intent with this extended rollout is to give appropriate time for ad creators and tool providers to prepare and incorporate these thresholds into their workflows,” said Vale.
Detailed instructions on how to track removed ads using an API as well as how to test the resource usage of ad content were provided in this support document.
Google is working on a new “Tab Freeze” feature for Chrome, which will pause (freeze) tabs you’re not using. That means lower CPU usage, a faster browser, and longer battery life on a laptop or convertible.
The Problem: Too Many Tabs
If you only had a single tab open at all times, Chrome would only need to render one web page at once. But you probably have more. Even while you’re not using them, each tab you have open in Chrome contains an open web page. That web page uses system memory. Any scripts and other active content on it continue running, too, which means the web page can use CPU resources in the background.
In some ways, this is good: Even if you switch tabs, a tab can continue playing audio or updating itself in the background. When you switch back to it, you don’t need to wait for the web page to reload—it’s instant.
But it can be bad. If you have a large number of tabs open—or even just a small number of tabs containing heavy web pages—they can use a lot of system resources, filling up your memory, taking up CPU cycles, making Chrome less responsive, and draining your battery. That’s why Chrome’s engineers created Tab Discarding and, now, Tab Freezing. They’re related features, but do different things in different situations.
How Tab Discarding Saves Your RAM
Tab Discarding was added back in 2015. This is a “memory-saving” feature, as Google puts it. In short, if your computer is low on memory, Chrome will automatically “discard” the contents of “uninteresting” tabs. Chrome won’t automatically discard a tab if you’re interacting with it, but that background tab you haven’t interacted with in hours is a prime target.
When a tab’s contents are discarded, it’s removed from your system’s memory, and the state is saved to disk. Nothing changes in Chrome’s interface—the tab appears on your tab bar and looks normal. But, when you click it and switch to it, you’ll see Chrome take a moment to quickly reload the page and get you back to where you were.
This slight delay is why Chrome only discards tab when your system’s memory is “running pretty low.” It’s good to use your RAM for caching. But automatically discarding a tab and quickly reopening it is better than forcing Chrome’s user’s to bookmark and close tabs manually.
When a tab is discarded, its process actually vanishes from Chrome’s built-in Task Manager, and you won’t see its memory used by Chrome anymore. When you click it to reload it, it starts up again.
How Tab Freezing Will Save Your CPU (and Battery)
Tab freezing is different from tab discarding. When a tab is frozen, its contents stay in your system’s memory. However, the tab’s contents will be “frozen.” The web page in the tab won’t be able to use CPU or perform actions in the background. For example, let’s say you have a heavy web page open in a tab somewhere, and it’s continually running scripts. After a while, Chrome will automatically “freeze” it and stop it from performing actions until you interact with it again. Those are the basics, and Google will likely explain how it works in much more detail soon.
Tab Freezing is an experimental feature. It’s built into current stable versions of Chrome 77, but can only be initiated manually. In Chrome Canary builds of the upcoming Chrome 79, Chrome will be able to automatically freeze tabs just like it can automatically discard them.
In Chrome Canary, several options are available for tab freezing if you head to chrome://flags and search for “Tab Freeze.” With this option enabled, Chrome will automatically freeze “eligible” tabs after they’ve been in the background for five minutes. Depending on which option you choose, Chrome can either leave them frozen or unfreeze them for ten seconds every fifteen minutes—just enough time to sync with a server or get a bit of work done if they need it. Google is clearly testing which option is best.
While tab freezing is an experimental feature, it’s almost certainly coming to stable versions of Chrome sometime soon—in some form, at least. The options in Chrome Canary were spotted by TechDows.
How to Play With Tab Freezing (and Discarding) Today
The current stable version of Chrome lets you play with both features if you want to know how they work. Just typechrome://discards in Chrome’s Omnibox and press Enter.
You’ll see a diagnostic page with a list of your open tabs and whether they can be frozen or discarded. On the right side of the page, you’ll see action links to “Freeze” and “Discard” each tab.
You can test it out to see the difference yourself. For example, if you launch YouTube and start playing a video, clicking “Freeze” for that tab will pause the video playback but not remove the YouTube tab’s contents from memory in the Task Manager. Clicking “Discard” instead will pause video playback and remove the tab’s contents from memory—you’ll see it vanish if you open Chrome’s Task Manager. Clicking “Load” will reload the tab’s contents to memory.
Why Discarding and Freezing Are So Useful
In other words, if your system’s memory is becoming full, Chrome will discard tabs you’re not using to free up space. It’ll silently reload them when you click the tab, but you’ll notice the page loading for a split second. There’s no need for Chrome to discard tabs while you have plenty of memory—Chrome is using that memory as a cache rather than leaving it empty. This speeds things up.
But, even if you have a lot of memory, Chrome will soon look at freezing tabs you’re not interacting with to save CPU time and battery power, potentially making Chrome and the other applications on your system more responsive. It will still keep them in memory—that way, when you reactivate a frozen tab by switching to it, the web page in the tab is ready to use as quickly as possible.
If Chrome needs to free up some memory, it might discard a frozen tab. But you can’t freeze a discarded tab: It’s already been removed from memory and isn’t truly open, so it can’t perform any actions in the background.
Now that the upcoming version of Microsoft Edge will be based on Chromium, Google’s work on Chrome will also make Windows 10’s default web browser better. Expect future versions of Edge to start freezing tabs automatically, too.
Your computer processor has a home: the socket. The CPU socket is rarely mentioned because it doesn’t help or hinder performance. Rather, it provides a standardized shape for a specific generation of CPUs.
Why, then, should you care about CPU sockets? Well, if you want to upgrade your CPU, you need to know the socket type. Your motherboard socket type dictates which type of CPU you can use, whether your CPU upgrade is worthwhile, or if you should consider upgrading your entire system.
So, what are CPU sockets, and why are they important?
What Is a CPU Socket?
Your CPU socket is similar to a light socket. A light socket makes your light bulb part of an electrical network, giving the bulb the power it needs to work. Your CPU socket makes a processor part of your computer, providing power and offering a way for the CPU to communicate with the rest of your system hardware.
.) In the past, there were other CPU socket configurations, including slot-mounted processors that you insert like a modern PCI card. Today, however, you place your CPU into the socket, on the motherboard, and secure it using a latch of some sort.
CPU sockets are decades old. Intel’s famous first processor, the Intel 386, used a 132-pin PGA socket (I’ll explain this acronym in a moment). The original Intel Pentium CPU used Socket 4 and later, Socket 5.
CPU sockets are not ubiquitous. There are differences between the CPU sockets developed by Intel and AMD, relating to the differences in CPU pin configurations between the two CPU manufacturing giants.
Why Are There Different CPU Sockets?
Unlike a light socket, CPU socket design does change frequently. Why?
Well, changes to the CPU architecture are the reason. New processor architectures arrive every few years and often come with a new set of requirements, including shape, size, and motherboard compatibility. Plus, there are two major x86 processor manufacturers: AMD and Intel. AMD and Intel CPUs have separate processor architectures, and compatibility between the two is impossible.
That last statement hasn’t always been true, mind. Back in the early days of computing, if you were lucky enough to own a high-end Socket 7 motherboard, you could use an Intel Pentium, an AMD K6, K6-2, or K6-3, a Cyrix 6×86, an IDT Winchip, or a Rise Technology mP6. And while dual-CPU motherboards do exist, there aren’t any that facilitate AMD and Intel concurrently.
What Type of CPU Sockets Exist?
Over the years, many types of CPU sockets have come and gone. Only three are relevant at the current time: LGA, PGA, and BGA.
LGA and PGA
LGA and PGA can be understood as opposites. “Land grid array” (LGA) consists of a socket with pins that you place the processor on. PGA (“pin grid array”), on the other hand, places the pins on the processor, which you then insert into a socket with appropriately placed holes.
(short for Threadripper 4), which is an LGA socket. TR4 is only AMD’s second LGA socket. Earlier Intel CPUs, such as the Pentium, Pentium 2, and Pentium 3 all used a PGA socket.
There is also a BGA socket, which stands for “ball grid array”. The BGA technique permanently attaches the processor to the motherboard during production, making upgrades impossible. A BGA socket and motherboard can potentially cost less, but there are very few equivalents between consumer BGA products, and LGA and PGA.
Furthermore, BGA technically is not a socket because it is a permanent motherboard feature. (You can easily replace an LGA or PGA CPU.) BGA sockets are still worth mentioning since it serves the same function.
Several years back there was a rumor that Intel was going to sunset the LGA socket. Intel LGA sockets would phase out after the 4th generation Intel Haswell CPUs. It never came to pass, and Intel still develops CPUs for LGA sockets.
That said, with the increase in system-on-a-chip (SoC) hardware, Intel has increased its BGA socket use. Similarly, ARM, Broadcom, Qualcomm, Nvidia, and other SoC manufacturers all rely heavily on BGA.
Does CPU Socket Type Matter?
A processor using a particular socket type will fit into any motherboard with that socket, right? Wrong!
Socket types like LGA are a category, rather than a specific model. There are many socket variations built on the basic specification.
Intel gives its LGA sockets a name based upon the number of pins. LGA1155, for example, has 1,155 individual socket pins. A processor built for that specific socket type will work only with that socket. Sometimes the numbers are incredibly similar, like LGA1155 and LGA1156, but you cannot force one into the opposing socket. A single Intel socket variation can cover multiple CPU generations.
AMD takes a slightly different approach. It labels its sockets with broad names, like AM3 or FM1. Compatibility is still strictly enforced, though AMD occasionally upgrades a socket while retaining compatibility. You can spot an upgraded AMD socket with the “+” symbol, such as AM2+ and AM3+.
Will CPU Sockets Become Extinct?
Computer development continues to feature a socket as a core design feature. Most components, including the processor, are upgradable or serviceable. Home and business users have the opportunity to build a system to whatever specification they desire, knowing that in time, they can make improvements.
The rise of mobile devices has seen a slight paradigm shift. The PC is not going extinct, far from it. But it is changing significantly to cope with the demands of the mobile hyper-networked world. The extinction of sockets could well be part of that change. CPU sockets add bulk and manufacturing complexity to products striving to reduce costs and size.
Predictions of the demise of the CPU socket in the near future are premature. You only have to look at Intel and AMD developing smaller, faster CPU manufacturing processes, as well as the development going into upgrading existing sockets or producing new socket variations.
It makes sense, too. Even though there more mobile devices than ever, enthusiasts and IT specialists will always look to a motherboard with a socket so that upgrading a single part is an option, rather than replacing an entire system, server, or otherwise.
How To Build Your Own PC It’s very gratifying to build your own PC; as well as intimidating. But the process itself is actually quite simple. We’ll walk you through everything you need to know. Read More
Gaming hardware tends to take a backseat at the yearly E3 trade show, with new titles getting the bulk of news coverage. But AMD is making a splash by showing off its latest Ryzen desktop CPU designs, aimed squarely at PC gamers.
The third generation of Ryzen dedicated CPUs come with more efficient 7-nanometer designs, and the usual boosts in speed, cache, et cetera. Prices range from just $200 for the Ryzen 5 3600X all the way up to $750 for the Ryzen 9 3950X, which boasts no less than 16 cores, 32 threads, and 72MB of cache.
But that’s all pretty par for the course. What’s most impressive about these newly-announced chips is that, despite the boost in performance and a shift to a new manufacturing process, all of the work on the existing AM4 socket design. AMD’s AM4 standard has been popular with budget builders, and in service since the first-gen Ryzen chips hit the market in 2016. The newest Ryzen chips stay compatible with the old standard due to some ingenious and very intentional fabrication design.
Now those same builders have access to the latest chips, and quite a lot of flexibility in terms of price and performance. AMD’s desktop chip line isn’t entirely limited to the AM4 socket—the ultra-powerful “Threadripper” chips still need a different standard. But the practical upside is that, if you want, you can use the $80 AMD motherboard you bought for a tiny budget machine three years ago for a polygon-pushing powerhouse gaming PC later this year with very few compromises.
The latest batch of Ryzen chips will be available between July and September of this year. At E3, AMD said it intends to keep the AM4 socket standard going into 2020 at the very least. Take that, Intel.
Every computer contains at least one processor, also known as a CPU or central processing unit. Your computer’s CPU is probably made by Intel or AMD. Here’s how to see what CPU you have and how fast it is.
You don’t need a system information utility to find this information. Windows shows it in several different places.
To find this information in Windows 10’s Settings app, navigate to Settings > System > About. Look under “Device specifications.” The name of your computer’s processor and its speed are displayed to the right of “Processor.”
You can press Windows+I to open the Settings app quickly. You can also press the Windows key, type “About” to search your Start menu for this settings screen, and click the “About This PC” shortcut that appears.
Windows 10’s Task Manager shows detailed CPU information, too. Right-click your taskbar and select “Task Manager” or press Ctrl+Shift+Esc to launch it. Click the “Performance” tab and select “CPU.” The name and speed of your computer’s CPU appear here. (If you don’t see the Performance tab, click “More Details.”)
You’ll also see real-time CPU usage data and other details, including the number of cores your computer’s CPU has.
Windows 7—or Windows 10—users can find this information in the Control Panel. Specifically, it’s on the system pane. Head to Control Panel > System and Security > System to open it. You can also press Windows+Pause on your keyboard to instantly open this window.
Your computer’s CPU model and speed are displayed to the right of “Processor” under the System heading.
If Windows isn’t booting on your system, you can still find this information in several other ways. Your computer’s documentation likely includes system specification details like this. You may also find this information displayed in your computer’s BIOS or UEFI firmware settings screen.