Google has been toiling to get Android 11 ready for launch this fall in the midst of the coronavirus pandemic. While it still expects to hit that target, the testing phase will take a little longer than expected. Google is launching a fourth Android 11 developer preview today, and the beta is delayed until next month.
Unlike past Android tests, Google has split Android 11’s preview period into the developer preview and beta. During the developer preview, you can only install the new OS via flashable images from Google’s dev site. That requires a compatible Pixel phone and a computer running the Android Debug Bridge (ADB). When the beta begins, you’ll be able to sign up for OTA updates that install the new OS on your Pixel. Additionally, other OEMs will be able to offer test builds of Android 11.
So, Developer Preview 4 will go live today with another round of minor tweaks and changes geared mostly toward stabilizing the operating system. You’ll be able to flash the new build on Google devices from the Pixel 2 and 2 XL all the way to the latest Pixel 4 and 4 XL. Google has not discussed any specific changes, and we wouldn’t expect anything major beyond improved stability.
The time for new features will come next month, which is when Google will launch the Android 11 beta. Previously, Google planned to hold its I/O conference in May, but it scaled that back to an online event in light of the pandemic. Now, it’s scrapped even that and will just have a “Beta Launch Show” on June 3rd. This won’t be the same wide-ranging discussion of Google technology and products you’d get at I/O. Instead, this sounds like an online version of the popular “What’s new in Android” keynote that takes place at each I/O.
Google’s Dave Burke will host the show, talking about topics like connectivity, security, accessibility, and some things Google is keeping secret until the event. There will also be a live Q and A session following the presentation. There should be one additional Beta release in July, and the next release after that should be final. Google doesn’t have a date for that yet, but it’s still expecting to hit its Q3 2020 timeline. And no, we’re not expecting a fun dessert name this year. It’s just Android 11.
The Deep Space Nine Upscale Project is an unofficial fan effort and not affiliated with Paramount in any way.
After three months of work, the Deep Space Nine Upscale Project (DS9UP) has hit some milestones that I feel comfortable showing off in greater detail. While there’s still more to do to bring Deep Space Nine into the modern era, I’m proud of what we’ve achieved in the past three months.
I’ve included a broader array of videos in this article. If you’ve wanted to see more head-to-head comparisons and examples of how different settings play against each other, you’ll like what we’ve got in store for you. Footage sources compared here are:
Original DVD: Not much point to the comparison if I don’t show you where we started from. There’s a lot I don’t like about Deep Space Nine on DVD, but I’d recommend anyone buy them compared to streaming the show. In-article designation: “DVD.”
Upscaled MKV: This is the toughest comparison for DS9UP. Topaz Video Enhance AI has a 30-day free trial and a $199 – $299 list price depending if it’s on sale. The DVD set costs ~$110. If I can’t demonstrate a better version of Deep Space Nine than you’d see from upscaling a DVD rip, there’s literally no point in what I’m doing. In-article designation: “MKV.” Original MKV from the DS9 rips I did ~15 years ago.
DS9UP Encode Process: My own work. Because this is an ongoing series of articles and I want to be able to refer to my own encode workflows without confusing people, I’ve decided to use codenames for the major “releases,” as it were. Fortunately, there’s a ready supply of thematically appropriate names near to hand. Say hello to Rubicon, a 23.976fps constant frame rate conversion of DS9’s original variable frame rate content. It’s currently created with Handbrake, StaxRip (AviSynth, not VapourSynth), and DaVinci Studio Resolve. Details on why I’m using Handbrake instead of MakeMKV will be forthcoming in an upcoming article.
Rubicon isn’t perfect — it’s got a ~100ms audio synchronization bug in a couple of clips that I still need to fix, but I’ll iron it out. Besides, it wouldn’t be Star Trek if Season 1 didn’t end with a few bugs to work out.
Previously, On Deep Space Nine The Deep Space 9 Upscale Project…
I was inspired to undertake this project by the work of CaptRobau, who published the first screenshots and video of what an upscaled Deep Space Nine could look like. I investigated the possibility of upscaling the show myself, but the performance figures were daunting and Gigapixel AI often crashes if you attempt to load too many images at once. The idea of manually converting each episode in batches of 2,500 to 5,000 frames was anything but appealing.
The release of Topaz Labs Video Enhance AI took this idea from pipe dream to possible. While the application is still very new and in need of some bug fixes, its long-term potential is demonstrably tremendous.
Deep Space Nine is, unfortunately, an ideal candidate for this kind of restoration. While Paramount created an absolutely beautiful Star Trek: The Next Generation remaster, the company has claimed the boxed set didn’t sell well enough to justify making an effort for the later shows like Deep Space Nine and Voyager. For reference, this sort of image quality is what major stream providers like Netflix believe is acceptable:
Defending the Alpha Quadrant with the power of moire!
DS9’s lousy streaming quality isn’t news to longtime fans of the show. It’s just that now, there’s something to be done about it. I decided that if Paramount wasn’t willing to treat DS9 with the respect it deserved, I’d take on the challenge myself, despite having no previous experience or education in video editing. Rubicon isn’t perfect, but I think it represents a significant uplift worth talking about.
Up until now, my clips have either lacked audio or had audio badly out of sync with the video feed. That’s no longer the case. So let’s get started:
Tips for Best Viewing, Notes on Quality Comparisons
Be advised that the audio on some of the Rubicon clips is off by a fixed ~100ms. This can be easily corrected for in a video player, but I haven’t had time to figure out why it happened in the first place.
Always set Netflix to play in 4K or the highest available quality, even if you do not have a 4K monitor. This will improve the quality of the stream regardless. The quality differences between Rubicon, an MKV, and the other versions of Deep Space Nine are less apparent at lower quality levels.
Be aware that the largest gap in quality, at least in my opinion, is between the streamed services and the DVDs. That’s not an absolute — there are some scenes where the DVDs are quite poor — but it’s a pretty good rule of thumb.
The gap between the upscaled MKV file we’ve included in this coverage and what Rubicon can achieve is much smaller than the leap from DVD to upscaled footage. When I wrote my first article on upscaling Deep Space Nine, I said that I felt as if running an MKV through Topaz VEAI got you 75 – 80 percent of the improvement that might be reasonably squeezed out of the DVD source, and that still seems like a pretty fair assessment.
Initial Fleet Flyby
DVD:
The DVD looks pretty good here, honestly. There’s not a lot of great detail on the distant ships, but the nearby Excelsior comes through quite nicely.
MKV:
The MKV looks really nice, here. I don’t expect to see the Rubicon encode recover much more detail than we have already.
Rubicon:
The hitch at the front of the video is because of where I cut the frame. Apart from that, Rubicon and the MKV tie here. Not much difference.
Defiant Conversation
Most of my comparisons have been battle comparisons to-date, but not much of Deep Space Nine actually involves combat. This short clip focuses on a range of characters — shows like DS9 have a lot of skin tones, and some AI upscalers handle them oddly.
DVD:
The DVD is dark — Deep Space Nine feels dark overall — but the detail is pretty solid. Dax’s face looks oddly low-detail, though.
MKV:
The GCG preset brings out some nice detail in the clip, but it also creates an odd aberation on the bulkhead over Sisko’s left shoulder at one point. This appears to be an error in the upscaler — the problem isn’t present on the regular version of the MKV — but it shows how important it is to keep every bit of detail, since it isn’t present in Rubicon, either.
Rubicon:
I don’t think there’s all that much difference between the MKV and the Rubicon upscale in this clip, either. That’s not to say I can’t see a difference — if you pause both videos on exactly the same frame, you can usually find a few details that favor Rubicon, and in a few spots, places that favor the MKV. Ultimately, though, I’m not sure how much of the detail is visible. Rubicon has a ~100ms audio delay in this clip that I didn’t notice until it was too late to fix for this article.
First Fleet Engagement
This sequence is one of the great battle shots of Deep Space Nine. The show had been showing us fleet engagements all throughout Season 6, but Sacrifice of Angels was teased hard as something that was going to be extra-special. This isn’t the first fleet skirmish — those have been going on for a while by this point in the episode — but this is the largest pure Federation fleet we’ve ever seen opening up on-screen.
DVD:
The DVD looks like it was recorded off someone’s old VCR tapes. The ships going by in the background look like vague little miniatures. The image is downright ugly and robbed of most of the impact.
MKV
The MKV file is a huge improvement over the DVD. The noise is gone and you can get a sense for just how large an engagement this was intended to look like. Tremendous improvement.
Rubicon:
This is another area where the improvements from Rubicon over MKV are small, but definite. Watch the leading edge on the saucer on the lead Galaxy-class vessel in the two clips and you’ll see that it’s blurrier. Rubicon is slightly sharper overall, and it’s easier to follow the Galaxy-class vessels as they move from the background to the foreground of the video. Until I upscaled this scene, I didn’t realize that the Galaxy-class ships moving through it near the end had even been visible in the back of the video. When I said you’ve never really seen this footage until you upscale it, I really wasn’t kidding.
Second Fleet Engagement:
The second, climactic battle of Sacrifice of Angels. I’ve showed short clips of this fight sequence before, but this is the first time I’ve shown the whole thing. This was one of the all-time high watermarks for DS9’s VFX team.
DVD:
The DVD is, once again, badly marred by noise. It’s difficult to read the hull letters on the Miranda-class ships and there’s a weird aliased grill on the bottom of the Miranda saucer. Several interlaced frames are prominently visible. It’s a dismal way to experience such a beautiful set of scenes.
MKV:
The MKVs, as expected, dramatically clean up the show. Again, I can’t argue with anybody who says this footage looks gorgeous in upscale. If it didn’t, I wouldn’t be writing this article.
Rubicon:
Rubicon eliminates all but a trace of the aliased pattern on the underside of the Miranda-class hull, smooths out the motion overall, and cleans the noise out of the smoke billowing around on the Defiant’s bridge until it just looks like, well, smoke. Detail levels in the Rubicon clip are just a touch higher.
How About Some Footage From a Different Episode? Any Episode?
No matter how bored you are of looking at Sacrifice of Angels, I promise you, you are not as tired of it as I am. I’m literally better at lip-reading than I was before I started this project. With that said? Your wish is my command:
It was important to me to demonstrate that Rubicon could stand up to footage in other seasons without modification. I used exactly the same process to render “The Die is Cast” that I did for the rest of the show.
Looking Ahead to Season 2
I’m declaring this the end of “Season 1” of DS9UP for several reasons. First, I’ve got a move coming up, and need to turn my attention towards it. Second, I’m long past due to circle back and talk to some folks who have either wanted to help with this project or are already actively working on efforts of their own. Third, I want to pause long enough to hammer some remaining issues out of my workflow, understand some of the problems I’ve had over the past few months a bit better, and return to the idea of improving the color balance of Deep Space Nine through some judicious changes.
As things stand, I’ll be writing a follow-up article to this one over the next week or so, with some additional examples of alternate workflows and outcomes when using applications like AviSynth. In addition to 5Sharp, I’ve got a ~48 fps and ~60 fps version of Deep Space Nine that have their own strengths and weaknesses. We’ll also finally be taking a look at Gaia-HQ.
I couldn’t have completed the work I’ve accomplished to-date without help from a number of people, including Gary Huff, Mark Renoden, Steve Reeve for some deinterlacing solution ideas, and help from several members of the Doom9 forum. Shortstack, I still hope to chat with you about recoloring ideas. Anybody else who deserves to be on this list, I sweartogod I’ll update it as soon as I’ve slept.
What We Brought Ahead
Deep Space Nine is too good of a show to be left rotting on DVD-era source. For all Paramount’s talk about the high costs of remastering, I’d love to see the breakout of recutting all of DS9 and Voyager compared with the cost of a single episode of Discovery. Back in 2017, leaked data showed the budget for Discovery at $8M – $8.5M per episode. Supposedly TNG’s episodes cost $70K each to remaster, but let’s assume VOY and DS9 are more expensive, at $100K each. The $34.8M it would take to remaster 348 episodes of TV works out to… about 4.5 episodes of Discovery?
Paramount could build a better version of the show than I could even hope to create — but since they aren’t going to bother, I figure I’ll keep up my own efforts.
On the night the last episode of Deep Space Nine aired, I carried my IBM K6-233 tower out to the living room, ran a 3.5mm cable from the audio-out port on our VCR into the line-in port on my sound card, and made a recording of the following. It’s been one of my favorite moments of the show ever since it aired, and I can’t think of a better way to end what I feel has been a very successful “season” thus far, than with a little James Darren.
Rubicon’s credits. I’m actually a little more partial to the ones I released back on April 27, but the Defiant’s motion is better here.
To the actors, artists, creators, directors, set crew, sound crew, and anyone else I’ve forgotten: Thanks for creating a show so damn good, people still rally around it 25 years later to see it treated with the respect it deserves.
With students and business personnel alike forced to work at home for the time being due to the COVID-19 outbreak, it has become essential to have a capable computer for work. For a limited time, Lenovo’s well-built ThinkPad laptops are being offered with sizable discounts. Any of these systems would be excellent options for clearing your in-tray or finishing up your child’s homework.
This lightweight system weighs in at just 2.4 pounds, and it comes equipped with a modern Intel Core i5 processor as well as a 1080p 14-inch display. In addition to the solid build quality, this system also features a fingerprint sensor and exceptional battery life; it can last for up to 18.3 hours on a charge. Right now you can get this system marked down from $2,789.00 to $1,533.95 from Lenovo with promo code THINK45.
Lenovo ships this ThinkPad T490 laptop with a Core i5 processor and a 1080p display. The system also features a high-quality battery that can last for up to 16.11 hours on a single charge. Rapid charging technology enables the battery to regain 80 percent of its charge in just one hour — that’s enough for up 12 hours of use. With promo code THINK45, you can get this system marked down from Lenovo marked down from $2,009 to just $1,104.95.
If you’re looking for a powerful gaming laptop with a 4K display, have a look at Lenovo’s ThinkPad X1 Extreme. This system features a Core i7 processor and an Nvidia GeForce GTX 1650 graphics card, giving it decent power for running modern games. This system is also exceptionally light for a gaming computer at just 3.76 pounds. Add to that the 4K OLED HDR500 display and a battery that can last up to 14 hours and you end up with quite an incredible laptop. Currently, you can get this system from Lenovo marked down from $4,037 to $2,220.35 with promo code THINK45.
Featured Deals
Lenovo ThinkPad X1 Carbon Gen 7 Intel Core i5-10210U 14″ 1080p Ultra-Light Laptop with 16GB RAM, 512GB SSD for $1533.95 at Lenovo (use code: THINK45 – list price $2789)
Lenovo ThinkPad T490 Intel Core i5-10210U Quad-core 14″ 1080p IPS Laptop for $1104.95 at Lenovo (use code: THINK45 – list price $2009)
Lenovo ThinkPad X1 Extreme Gen 2 Intel Core i7-9750H 6-core 15.6″ 4K OLED Laptop with GTX 1650 Max-Q, 32GB RAM, 1TB SSD for $2220.35 at Lenovo (Select i7-9750H, 32GB RAM, 1TB SSD, 15.6-inch UHD OLED Display and use code: THINK45 – list price $4037)
Lenovo ThinkPad X1 Yoga Gen 3 Intel Core i5-8250U Quad-core 14″ 1080p IPS Touchscreen Laptop for $949.05 at Lenovo (use code: EXTRAFIVE – list price $2659)
Lenovo ThinkPad X280 Intel Core i5-8350U Quad-core 12.5″ 1080p Touch Laptop for $806.55 at Lenovo (use code: EXTRAFIVE – list price $2079)
Lenovo IdeaPad S540 Intel Core i7-10510U Quad-core 13.3″ QHD Laptop for $759.99 at Lenovo (use code: EXTRAFIVE – list price $999.99)
Lenovo Yoga C740 Intel Core i5-10210U Quad-core 14″ 1080p IPS Touchscreen Laptop with 512GB SSD for $788.49 at Lenovo (use code: EXTRAFIVE – list price $829.99)
Lenovo IdeaPad L340 Intel Core i7-9750HF 6-core 15.6″ 1080p Gaming Laptop with GTX 1650 for $835.99 at Lenovo (use code: EXTRAFIVE – list price $1019.99)
Lenovo ThinkBook 14 Intel Core i5-1035G1 Quad-core 14″ 1080p Laptop for $723.87 at Lenovo (use code: THINKDEAL – list price $1149)
Lenovo Yoga C930 Intel Core i7-8550U Quad-core 14″ 1080p Touch 2-in-1 Laptop for $1049.99 at Lenovo (use code: YOGA4U – list price $1599.99)
Lenovo IdeaPad S940 Intel Core i7-8565U Quad-core 14″ 4K IPS Laptop for $1139.99 at Lenovo (use code: EXTRAFIVE – list price $1799.99)
Lenovo ThinkCentre M75q AMD Ryzen 5 PRO 3400GE Tiny Desktop with 512GB SSD for $585.22 at Lenovo (use code: THINKDESK42 – list price $1009)
Lenovo ThinkCentre M90n Nano Intel Core i5-8265U Quad-core Desktop for $626.05 at Lenovo (use code: EXTRAFIVE – list price $1089)
Note: Terms and conditions apply. See the relevant retail sites for more information.For more great deals, go to our partners at TechBargains.com.
Microsoft announced Windows 10X last year as the key to its dual-screen ambitions, but the company has changed course since then. Confirming a previous rumor, Microsoft now says Windows 10X will instead target single-screen devices first. The latest tidbits from the rumor mill suggest this is more than targeting traditional laptops — Microsoft may be planning to use Windows 10X to compete directly with Chromebooks.
Google’s first Chromebooks were sluggish, ugly, and poorly built. Over time, Chromebooks have expanded to every niche in the laptop market, even the high-end. Still, it’s the budget-priced and midrange Chromebooks that have become popular, and that’s taken a big bite out of Windows-based laptops. Similar efforts on Microsoft’s site like Windows RT have crashed and burned, but the company hopes it can strike the right balance with Windows 10X.
According to a leaked document, Microsoft reportedly adjusted its plans when the coronavirus pandemic swept the globe. Launching expensive new dual-screen laptops in this climate is a significant risk, but demand for traditional laptops has exploded as people seek to get work done from home. Inexpensive Chromebooks, in particular, have been in high demand, but Microsoft has also seen a 75 percent jump in Windows 10 usage this year. So, the company feels it’s the right time to launch a streamlined version of Windows with a focus on traditional form factors.
The current Windows 10X UI. Image by Windows Central
Windows 10 is a feature-packed operating system with a lot of baggage from decades of ongoing Windows development. With 10X, Microsoft will feature UX improvements across the board, including a new Start Menu and window management system. It will also feature a new application container system that promises better security and performance. There’s also talk of cloud-based app virtualization technology that could make light-weight Windows laptops more capable.
This is a tough environment to launchanyproduct, particularly new form factors that inevitably demand a premium price. We’ve seen the same problems in the mobile space with foldable phones that cost well over $1,000. It’s just a bad time to ask people to spend that much. With all the work already done on Windows 10X, repositioning it for traditional laptops might be the best option available to Microsoft. It’s clear it needs something to go toe-to-toe with Chrome OS. Hopefully, this goes better than all the other times Microsoft tried to pare back Windows.
If you assumed the increased flood of work from home employees in the wake of the COVID-19 epidemic was taxing companies’ ability to safeguard their digital systems, you’d be correct. But it isn’t just a surge of unprotected personal computers accessing potentially sensitive company information that’s got security personnel worried.
The truth is that many of those specialized cybersecurity experts aren’t even patrolling their usual beat these days. To help handle the needs of their now remote workforce, many companies have reassigned their security experts to general IT duty to help keep up.
CyberTraining 365 Online Academy is a treasure trove of security training, offering over 3,800 up-to-date modules explaining the latest technologies and industry security standards. Whether you’re interested in analyzing malware, conducting penetration testing, identifying persistent threats, researching threat intelligence, or the latest in reverse engineering, you can stay on top of the rapidly changing industry through their growing library of video courses.
Taught by some of the industry’s best and most recognized certified Cyber Security instructors and infrastructure experts with years of teaching experience, you can take this training at your own pace. It’s currently a stockpile of over 660 hours of content, all aligned with the National Cybersecurity Workforce Framework developed by the National Initiative for Cybersecurity Education (NICE).
Students can learn all the latest on ethical hacking, real-life vulnerabilities, and threat examples and get familiar with the latest hacker tools while working toward global-recognized skill sets and certifications.
Regularly $499, you can now get a year of access to all of CyberTraining 365 Online Academy’s courses for only $29.99. You can also keep up with your training with similarly discounted 3-year ($39.99) and lifetime ($79.99) service plans.
Note: Terms and conditions apply. See the relevant retail sites for more information.For more great deals, go to our partners at TechBargains.com.
Nintendo is notoriously secretive about its development process, but we just got a glimpse of how the sausage is made thanks to a massive leak of data from the Nintendo Wii and N64. Among the thousands of files posted online in recent days are the source code for the Wii operating system, detailed design documents, and software demos for the Nintendo 64.
The files first appeared on 4chan, a largely unmoderated online forum that has given rise to Anonymous, QAnon conspiracy theories, and uncountable memes. We won’t link to the leaked Nintendo data, but it’s easy enough to find with a Google search. We don’t know who posted the files (4chan is anonymous by default) or if it has anything to do with the recent hack of Nintendo’s online service. The timing is suspect, though. However, online chatter points the finger at a company called BroadOn, a company that worked with Nintendo repeatedly over the years and did much of the technical design work to make the Wii a reality.
Emulation fans and developers will probably be overjoyed about this leak. It should be possible to improve game emulation significantly with access to the full source code for the Wii’s operating system and SDK. The leak even contains block diagrams and Verilog files for every part of the Wii. Verilog is a text format for describing electronic circuits and systems — someone with the right know-how could basically recreate every component of the Wii from this data. Of course, Nintendo’s lawyers will probably be racking up billable hours to ensure that it doesn’t happen.
Also contained in the 3 gigabytes of leaked data are several documents and files from the Nintendo 64, which dominated gaming in the late 90s. There are functional demos for the N64 that show off what, at the time, were groundbreaking 3D rendering techniques (see above). This is a fascinating glimpse of this pivotal time in gaming history, and there may be a lot more to come. There are even a few documents relating to the development of the 2001 cult-favorite Gamecube console.
Some sources on 4chan claim this cache of data is just a taste of what’s out there — there may be as much as 2TB of stolen Nintendo data in the hands of hackers. If true, it’s only a matter of time before it’s laid bare on the internet.
Speeding up a slow PC can be a challenge, particularly when dealing with older hardware that may be on the cusp of needing an upgrade or replacement anyway. Sometimes, a system simply needs a fresh OS install or driver update to perform significantly better. In other cases, upgrade or wholesale replacement are necessary.
This article is designed to help you troubleshoot a slow machine. Specifically, it walks through the process of determining whether your problem is more likely to be caused by software, hardware, or simply the age of the machine. I’ll address the topic of what to upgrade (and when a machine is worth upgrading) in a separate article.
Top Level Questions
Whether you are doing this for someone else or for yourself, here are a few “big picture” questions to consider:
Has the system’s performance changed recently? If the answer to this is “Yes,” did it change when you installed a specific application or OS update?
Has the entire PC slowed down, or just specific applications? If the problem is isolated to one app or a specific group of apps, have those applications been recently updated or modified?
Does the machine feel sluggish and slow to respond to feedback like typing? Does it have this problem in all cases, or only when certain programs are running? Does it have the same lag if booted in Safe Mode?
Does the system begin to run slowly after a few minutes of work or gaming, or is it slow from boot?
Is the system just slow, or is it also unstable? If it crashes, is there a specific application or application(s) that makes this happen reliably?
I ask these questions even before I ask for the make and model of the PC or specific information about its components. Knowing the age, model, and intended purpose of the machine are all important. But age, especially in desktops, just doesn’t mean what it used to. My media center PC downstairs is built on a Core i7-920. It’ll turn 10 this year and still works perfectly.
It’s not enough to say that the system is running slowly. Where it’s running slowly (or displaying other issues) tells us something about why these things are happening.
How Do I Tell If My PC Is “Just” Slow?
A PC that’s “just slow” shouldn’t have any other real problems apart from the lack of performance. You should have a feel for how long it’ll take to open a webpage or launch a game, and you shouldn’t see any unexplained variation in those times or particular instability when performing them. A slow PC may thrash the hard drive if you ask it to open too many browser tabs simultaneously, but there should be a reasonable and predictable (to you) relationship between how much work you are asking your PC to do and how quickly it can do it. A machine that’s “just” slow shouldn’t have problems with pop-ups or persistent trojan and malware infections, and it shouldn’t be unstable.
If you’re a gamer on an old gaming PC, said system should still run the same games you bought it to play back then (with allowances for OS compatibility or driver issues). If you could run BioShock Infinite at 60fps on your current computer in 2013 you should be capable of running it today. If BioShock Infinite (or any other old game or application) plays and runs perfectly, but newer software doesn’t, you may be looking at a situation where you need to upgrade as opposed to dealing with an underlying hardware or software problem.
Any PC still using a magnetic HDD is going to feel slower than a PC equipped with a SATA or M.2-based SSD. Any PC with less than 4GB of RAM is going to struggle at least some of the time under even moderate workloads. 8GB should be considered a minimum for any modern gaming. Dual-core CPUs without Hyper-Threading or SMT and CPUs based on Intel’s Pentium, Celeron, or Atom product lines are more likely to “feel” slow, even if the system is well-configured. Systems based on old AMD CPUs from the Phenom II era or the Bulldozer/Piledriver product families do not offer particularly high performance.
A system that was brand-new in 2013 and hasn’t been upgraded since shouldn’t feel like the latest-and-greatest in 2020 — but it shouldn’t be riddled with unexplained performance drops or errata, either.
Hunting Software Problems
Performance problems are more likely to be software-related if they are associated with an underlying change to the OS or target application. If you’ve been gaming with no problems but recently bought your first new title in months, your GPU drivers may need to be updated to play it appropriately. It’s never a bad idea to check the minimum specifications for a new title to make sure you meet them: If you’ve got a GTX 480 and the game recommends a GTX 770 as a minimum card, you may well have found the cause of your problem.
Similarly, if you have an older system and you are having issues related to a specific subsystem (networking, sound, video), it may be worth updating your chipset and component drivers as well. The chance of this solving a problem is much higher if there’s an actual identified issue with what you’re using already, but I’ve had reinstalls clear up issues before from seemingly unrelated causes. It’s rare, but the chance isn’t zero.
Tools like Task Manager can sometimes be used to troubleshoot basic software problems with responsiveness and performance. Consistently high CPU usage may indicate a problem with an application, especially if it’s a browser and especially if killing the process improves system responsiveness. Some browser extensions can cause higher CPU usage under certain circumstances.
When you check Task Manager for CPU usage, you can also check the Startup tab. Make sure you recognize the applications that are loaded and running. You may not recognize every process listed in the “Processes” tab, but look for suspicious names or random, nonsense strings of text. In some cases, this may be a sign you’ve been infected by malware — and malware can steal CPU and GPU time for cryptocurrency mining and harm overall perf.
Event Viewer may contain data related to a performance issue or slowdown, but honestly, it’s a crapshoot and interpreting Event Viewer isn’t all that easy for the common user. I’m not touching on it much here as a result.
If you are concerned that old 3D drivers might be hanging around on your rig and cluttering up its performance, tools like Display Driver Uninstaller can be used to completely remove these traces. I use DDU for reviews when testing AMD and Nvidia cards. Booting into Safe Mode, running the tool, and then booting back to Desktop adds some steps compared with just installing a newer driver. But if you’re trying to troubleshoot performance issues, DDU is a good way to ensure you’ve cleansed your metaphorical palate. If you are having issues with your sound card or network solution, it may be worth visiting your motherboard vendor to see if they’ve released updated drivers for these components, for example.
When all else fails, a full Windows reinstallation will almost always nuke any software-related problems and let you determine if a hardware upgrade is necessary. I try very hard not to take this step for obvious reasons — “Just reinstall Windows” is the kind of advice that cost US citizens millions when deployed as part of a scheme to lie to people about the state of their software. Most people do not want to go to the hassle of reinstalling Windows and losing their current data installations in the process. But if you are facing weird performance issues that you can’t lock down any other way, and you’ve tried the other steps discussed here, it’s one way to deal with the problem.
There is an exception to this. Sometimes, a Windows Update is responsible for changes to user systems that cause high CPU usage or break Start Menu functionality. Upgrading components or an entire PC is no guard against these sorts of failures. In these instances, even a full Windows reinstall may not fix the issue.
We’ve discussed the software side of the equation. Let’s talk about what which hardware issues are most likely to slow a PC. I’ve tossed in a fair bit of discussion about what component failure looks like in each case to illustrate the difference between the problems that sap performance and the problems that turn your computer into an expensive paperweight.
Hardware Failure Characteristics (and Which Ones Hurt Performance)
RAM: Failing RAM can make a game run more slowly, but only if the CPU is still having some luck pulling data out of it at all. More commonly, applications just crash. The distinguishing characteristic of RAM failure is that there isn’t automatically a distinguishing characteristic. If the failure has occurred at a high memory address, you may only see the issue intermittently when you have a lot of programs loaded. It may happen mostly in games if games are what you typically run, but a RAM failure will pop up anywhere given enough time. Applications like MemTest86+ can be used to test RAM. MemTest86+ is the only RAM test I recommend — I’m not saying it’s the only good application available, but I’ve seen other RAM testers claim that memory was good when it actually wasn’t. I’ve never seen MemTest86+ throw a false positive or false negative.
GPU: A failing GPU may display odd colors and textures whenever you run a 3D application, or function perfectly for basic desktop work but fail if asked to render a video or play a 3D game. Using a third-party utility like MSI Afterburner may allow you to resolve the issue by either lowering the GPU/RAM clock or increasing the chip’s voltage, but these fixes tend to be temporary. Games may crash at load or may run for short periods of time. The problem may begin in a specific title, but it probably won’t stay isolated. A failing GPU may also refuse to install its own driver.
A GPU that runs perfectly for a long period of time before you start seeing errors or texture flickering may be overheating, but may not be damaged yet. Dust the system and see if that improves things. Many games offer benchmark modes for testing GPUs, and running these tests in a loop will often produce a failure, though it may take multiple loops through the test to see it happen.
GPU failures typically will not slow down the system, unless lowering the clocks yourself (temporarily) fixes the problem.
CPUs: CPU failures are difficult to categorize because CPUs almost never fail. If your CPU’s performance is dropping, chances are it’s a thermal issue — either a mismounted heatsink or a thick build-up of dust. Make certain that your heatsink is properly mounted with an appropriate amount of TIM (thermal interface material) between the CPU and its heatsink.
Power Supply: One way to discover your PC PSU needs replacement is for it to burst into flames without warning. (Ask me how I know!) A far less terrifying method is for your machine to simply flip off or reboot during a gaming session or rendering run. In some cases, your motherboard may warn you that power delivery from the PSU was disrupted. Power supply failures can superficially resemble RAM failures, but MemTest86+ won’t return errors and the PSU won’t trip while running basic desktop tasks unless it’s truly having issues. In these cases, your motherboard may also report that one or more of the power rails is low. Bad PSUs often don’t cause slowdowns; they cause shutdowns.
Hard Drives: Mechanical spinning drives may click quietly and repeatedly when attempting to access certain parts of the drive, or I/O performance may drop badly when attempting to access data. Event Viewer may also log hardware failures if the OS can’t read data across the bus properly. SSDs will not click (no moving parts), but they may display the same longer-than-expected access times or sharply reduced performance as the CPU tries to read data from damaged parts of the drive. In many cases, these types of failures will kill the drive altogether rather than leaving it limping along at a lower level of performance.
Low Storage Space: Not a “failure” as such, but it’s a hardware issue, not a software one. Windows generally does not like to run without hard drive space. The less free space you have, the less the OS likes it. This can lead to errors, application crashes, and cache thrashing, all of which harm performance.
Heat: I’ve decided to list “heat” as a common characteristic of hardware instability and lost perf, rather than breaking this out by CPU and GPU. Dust is an amazing insulator. Pack in enough of it and fans won’t even spin. I’ve lost count of how many gamers and readers I’ve met who were afraid they needed to buy new hardware, only to discover that dusting what they owned restored the performance they were missing. Laptops can be a little trickier to dust than PCs, but if you have a desktop and are experiencing instability while gaming, pop the side of your case off and see if things don’t improve. It’s no guarantee — but it’s not a hard thing to test, either.
A PC with issues in just one game or a handful of games may only need a driver update, especially if the titles are newer. A machine that never crashes during desktop or browser work but slows down and crashes during gaming may need to be dusted. When I run into people who are suffering PC crashes while gaming, one of my first pieces of advice is to take the side panel off their computer. The growing popularity of laptops for gaming has made this fix a little less applicable than it used to be, but you’d be amazed how many “unstable” PCs become stable if you just improve airflow. Dusting is obviously the best way to do this, but you can’t always take time from a raid instance or Fortnite game to hunt down a can of compressed air. Just taking a side panel off is a quick way to test this theory.
Systems are less likely to crash outright from high temperatures than they used to be, but throttling can still play merry hell with game performance. I always check thermals when I’m evaluating system behavior.
Of all the issues here, I’d say dust, driver updates, and low disk space are the three most common factors likely to cause slower-than-desired performance in the absence of hardware failure. Broken software can absolutely cause all manner of problems, but these cases tend to be specific and particular to the user in question (or the hardware in question) and are much harder to resolve in a general-purpose guide like this.
Questions? Tricky cases? Got a machine with an issue you can’t seem to fix? Drop it below.
