How Tom's Hardware Tests, Rates and Reviews Tech Products

Tom's Hardware
(Image credit: Tom's Hardware)

To choose the best technology or keep up with the market, you need expert advice based on solid evidence and copious experience. Sure, you could just read spec sheets and assume that, when a company says its laptop provides "all day" battery life or its GPU delivers great 4K gaming, it must be true. Or you could comb through hundreds of user reviews to see what previous buyers have to say. But at Tom’s Hardware, we believe that rigorous testing is key to understanding the latest gear.

As a leading PC hardware and enthusiast technology site founded in the mid-1990s, Tom's Hardware has been evaluating the latest and greatest CPUs, graphics cards, motherboards, and much more for 25+ years. From Raspberry Pi accessories to SSDs, 3D printers and power supplies, once we get it in our hands, we benchmark the hell out of it using software, sensors and anecdotal testing. The Tom’s Hardware staff and freelancers have more than a century of experience pushing technology to its limits -- and sometimes even breaking it. 

 Tom’s Hardware Promise: Telling it Like it Is

Our recommendations can’t be bought. In an ideal world, every user would have the time, equipment, expertise and access to products they needed to do their own testing and draw their own conclusions about a whole category of hardware. But in the real world, we have the rare privilege of doing this professionally and standing in for our readers. With that privilege comes a sacred responsibility: to give you the unvarnished truth about what we experienced when using and benchmarking a product. 

 Tom’s Hardware Test Methods

Transparency is key. If we conduct a test, you should be able to replicate that experience. In each review, we describe how we tested that particular product, including both synthetic benchmarks and anecdotal use. Some product categories – CPUs, GPUs and SSDs, for example – use highly-scientific methods for testing while others such as gaming chairs can only be evaluated based on the reviewer’s experiences. 

We test too many types of tech to list them all here, and that list often changes as we delve into new and emerging areas of consumer and enthusiast hardware, or when once-hot product categories wither (remember dedicated sound cards or netbooks?) or become stagnant. 

But to give you a sense of what we test and how we test it, below we'll list some of our key categories, explain the basics of how we test those products and, if we have a dedicated how we test page for that product category, we'll link you out to it for much more detail. 

  • CPUs: We use a demanding suite of eight real-world gaming benchmarks in two resolutions, six synthetic gaming benchmarks, 50 desktop PC applications, thermal, power and efficiency tests. We also measure overclocking performance on any unlocked chip. To ensure a level playing field and avoid bottlenecks from other components, we use carefully-maintained test images paired with the best GPUs, motherboards, RAM, SSDs, cooling and power supplies. 
  • Graphics Cards: We test with a collection of recent games. On each game, we run once to "warm up" the GPU and then run at least two passes at each setting/resolution combination. We check for any anomalies and retest as needed, to ensure we get the most accurate data possible. We also go back and periodically retest cards with the latest updated drivers, and we occasionally add new games to the test suite when good candidates are released. 
  • Laptops: Before our reviewers lay hands on them, laptops are benchmarked in our labs where our lab staff test the displays with a colorimeter / light meter to determine vibrancy and brightness. They undergo a custom battery test which involves web surfing over Wi-Fi and a series of synthetic and real-world tests that measure processing power, storage speed and, if it’s a gaming laptop, game frame rates. Our reviewers attempt to open each laptop to see if it is upgradeable and what can be easily repaired. We also measure heat during periods of intense work. All of these tests are combined with real-world impressions of the system’s usability, design and input devices. 
  • Desktop PCs: Prebuilt desktops undergo most of the same tests as laptops, except there’s no screen quality, skin temperature or battery life to test. Our reviewers focus on airflow, future upgradeability and cable management: all long-term features to increasing the system's utility in the future.
  • Gaming Monitors: As we explain in detail in our article on how we test gaming monitors, we use a variety of high-end equipment, including a spectrophotometer and a colorimeter. We also use a DVDO generator to create accurate test patterns. We test brightness and contrast, grayscale tracking to ensure white is neutral at all brightness levels, color gamut accuracy and volume for sRGB and DCI-P3 gamuts, viewing angles  and response time. We also show results both before and after calibration. 
  • SSDs: We use a balance of tests that includes real-world file transfer, gaming workloads, synthetic benchmarks, and power and thermal testing. Each SSD is prefilled to 50% capacity to simulate long-term use (an empty drive runs faster) and tested as a secondary device to avoid interference from the operating system and other programs running in the background. 
  • 3D Printers: We set up each printer and use it for several prints which vary depending on its capabilities and included models, but usually include Benchy, a common test print. We try the printer both with its included software and, if possible, with popular slicers such as Cura.  
  • Keyboards and Mice: The best way to test input devices is to use them in a variety of common situations, preferably for several days. We install any relevant software and tweak the options. If a device is meant for gaming, we use it to play games. If it's meant for productivity, we use it for work.

What Our Ratings Mean

All reviewed products are rated on a scale of 1 to 5, with 5 being the best. Each product may also receive an Editor's Choice badge, which designates it as the best within its niche. The ratings mean the following:

5 = Practically perfect
4.5 = Superior
4 = Totally worth it
3.5 = Very good
3 = Worth considering
2.5 = Meh
2 = Not worth the money
1.5 = Buy for an enemy
= Fails horribly
0.5  = Laughably bad

How We Choose and Obtain Products for Review

There's a nearly-infinite universe of tech products, ranging from the thousands of generic cables on Amazon to the two dozen PC processors made by Intel and AMD. While we'd love to thoroughly benchmark and rate them all, we have to be selective. Our editors decide which products will be of the greatest interest to our audience and request companies lend us samples for testing purposes.  In some cases, where a company is unwilling or unable to send us an important product, we will use our budget to purchase it from a retailer such as Amazon or Best Buy. 

While companies often pitch us products to review, deciding which to review or cover in any way is an exclusively editorial decision, based on what we think our readers need to know. We do not accept payment of any kind for reviews nor do manufacturers receive preferential treatment for being advertisers. We also do not show our reviews to manufacturers prior to publication. 

Avram Piltch
Avram Piltch is Tom's Hardware's editor-in-chief. When he's not playing with the latest gadgets at work or putting on VR helmets at trade shows, you'll find him rooting his phone, taking apart his PC or coding plugins. With his technical knowledge and passion for testing, Avram developed many real-world benchmarks, including our laptop battery test.
With contributions from
  • bit_user
    @apiltch thanks for the article.

    I followed a link from one of your product reviews, hoping to find some detail on how CPU & GPU power consumption is measured. It would be nice if you could update this article with additional such details.

    One reason I'm interested is that I'm wondering if there's a way I can take isolated CPU (or, at least motherboard) measurements, in my own system, to evaluate the efficiency of my cooling solution.

    The other reason is just to know how to interpret the power measurements in your reviews, especially when it comes to comparing your results to those published elsewhere.
  • rabbit4me1
    Sorry guys I really don't believe that your tests are completely unbiased You're always going to support your sponsors you might do a little negativity on them but you always tend to be pushing Dell products and most of them are junk.
  • bit_user
    rabbit4me1 said:
    you always tend to be pushing Dell products
    I guess you're talking about laptops or monitors, because I think that's the only Dell-branded stuff they review. Here's a recent Dell laptop they rated at only 4 stars:
    Meanwhile, here's a Framework laptop they reviewed about the same time, granting it 4.5 stars and an Editors Choice award:
    It's easy to cry foul, but you should be prepared to substantiate your allegations.

    Regarding sponsors, I assume their advertising is handled through a 3rd party intermediary. Do you know differently?
  • apiltch
    We rate products based on how good we think they are and whether we think they are a good choice for readers who are in the market for that type of product at that budget. A lot of companies buy ads on our site, but the editorial team are not in sales. What matters to us is giving you good advice. I know that sounds trite, but it's the truth.

    Nobody here got into tech journalism to get rich. We're lucky to make a living doing it and we never forget how lucky we are to get to test out tech and share our observations with you. Our obligation, always, is to tell the truth.