same specs different performance
Laptop Benchmarks,  Laptop Comparisons

Buying A Laptop? Don’t Just Compare Spec Sheets!

Shopping for a new laptop? Unfortunately you can’t just compare the spec sheet, laptops with the same hardware may perform differently – let’s investigate.

The Problem

I recently reviewed the MSI Prestige 14 laptop, which has the same key specs as the larger Prestige 15. Based purely on this information, which laptop would you pick?

 MSI Prestige 14MSI Prestige 15
CPUIntel i7-10710UIntel i7-10710U
GPUNvidia GTX 1650 Max-QNvidia GTX 1650 Max-Q
Memory16GB LPDDR3-213316GB DDR4-2666
Storage512GB NVMe M.2 SSD512GB NVMe M.2 SSD
Screen14" 1080p 60Hz15.6" 1080p 60Hz
Size319 x 215 x 15.9 mm356.8 x 233.7 x 15.9 mm
Weight1.29 kg 1.6 kg

You’ll be forgiven if you thought “great, the 14 inch has the same hardware inside but it’s smaller and lighter, I’ll go for that one!”

Unfortunately it’s really not that cut and dry. The graph below shows the Cinebench performance from both laptops.

Cinebench R20 benchmark

This is a measure of CPU performance. The Prestige 15 is scoring 23% higher despite both laptops apparently having the same processor.

The differences are even more severe when we take a look at gaming performance.

Overwatch gaming benchmark

So what’s the deal? There are a number of factors that will affect performance, primarily different limits.

Push it to the limit

Laptops have different limits in place to keep things safe.

Thermal Limits

You wouldn’t want your CPU heating up so much that it melts. Fortunately, this doesn’t happen as thermal throttling will kick in first. This will reduce performance by ensuring temperatures do not rise above some configured limit.

An example may be that a laptop never gets hotter than 90°C. Once this temperature is hit, the clock speeds are dialed back to prevent it heating up further, the goal being to maintain a safe operating temperature.

Power Limits

Power limitations also exist, and usually go hand in hand with thermals. If a laptop has a weak cooling solution, then it will probably have a lower power limit. More power typically equals more heat, so to avoid overwhelming a weak cooling solution, power limits may be restricted.

An example of this is that many modern i7 laptops have a 45W TDP. For many machines with such a limit, the power being used by the CPU will not increase above 45W, even if there is thermal headroom available. Depending on the machine, it may be possible to increase the power limit at the expense of thermals – that is to say, the machine will get hotter but performance should be improved as a result.

What’s happening here

The Prestige 14 and 15 have the following power limits in the previous examples.

  • Prestige 14 – 15-30W GPU / 25-35W CPU
  • Prestige 15 – 35W GPU / 40-45W CPU

The limits need to be different due to the size difference between the machines, check out the internals.

MSI Prestige 14 and 15 internals
Left: MSI Prestige 14. Right: MSI Prestige 15

The larger Prestige 15 has superior cooling, so higher power limits should go down better, meaning a performance boost.

During the Cinebench test earlier, towards the end the Intel i7-10710U CPU  in the Prestige 14 was running at 25W, while the same CPU in the Prestige 15 was running around 40-45W. This explains the performance difference.

The Prestige 14 started out running better with a higher 35W limit, then 30W limit, before finally settling at 25W. This is an example of the power limit changing over time with respect to thermals, and this is why I take averages from multiple runs when reporting test results.

When gaming, the GPU in the Prestige 14 averaged 15W, while the Prestige 15 had no issues running at its 35W limit, so that explains the differences in gaming performance.

But wait, there’s more!

Another issue is that power limits can vary based on the workload being run. It’s not uncommon to see higher CPU power limits being reached when only the CPU is active and the GPU is idle. The CPU may be restricted further when the GPU is also active, such as when playing a game.

Some laptops will even dynamically adjust power limits on the fly as a means of keeping thermals in check. This means it even depends on the specific workloads being run, and don’t forget about thermals!

As thermals can be a limit, depending on how close you are to thermal throttling, even a change of room temperature by a few degrees can be enough to negatively affect performance. The general rule of thumb is that for every 1°C increase to ambient room temperature, expect about the same to the internals.

Most laptops with the same CPU/GPU inside have different cooling solutions, so differences are expected. Generally, thicker laptops tend to have better cooling. One may thermal throttle, while another may be fine and perform better as a result. The one with better cooling may be configured with higher power limits, resulting in better performance at the expense of some extra heat – it’s a trade off.

When you factor all these variables in, it’s easy to understand why performance for the same machine can vary between different reviewers.

The solution?

Unfortunately there isn’t an easy one, there are just a number of different factors that affect performance that are impossible to be aware of when reading a spec sheet.

The only way to really understand what’s actually going on with each machine is to check out independent reviews. This is why I try to perform all sorts of different testing in my detailed reviews, however this means the buyer needs to spend a lot of time researching.

The majority of people have no idea about the different limits discussed here. They will simply assume that if two machines have the same specs, then performance is probably quite comparable.

Unless there is some sort of standard testing methodology to show this information that all companies start using, then third party testing will be the only way to know for sure how a laptop will actually perform.

To be fair, in many instances the differences are far less than what I’ve shown here. The MSI Prestige 14 and Prestige 15 are more of a worst case example, but one that helps illustrate the point clearly.

Hopefully the next time you’re looking to buy a new laptop, you know why it’s worth checking reviews containing this information rather than just shopping the spec sheet.

 

 

10 Comments

  • Andre Faelnar

    Hi sir. First of all, I love your channel and your in depth reviews of different laptops. Also, I was gonna ask what can be a good suggestion for your next topic which is about cleaning the surface of your laptop. Take the 2 laptops for example: In Exhibit A, you have the Acer Helios 300 that has a matte surface and in Exhibit B, you have the Lenovo Y540. In one side, the Helios 300 has a smooth surface where its easy to clean but on the other end, the Y540 is prone to being a fingerprint magnet and it will look gunky and over time, it might worry some people using it long term. Which leads us to ask that can this also be a factor in terms of choosing for your next laptop. I’m just curious what you think. Plus keep it up bro. Love your content and looking forward to your future videos soon enough.

    • Jarrod

      Hmm the textures are definitely different, generally though it just means you might need to wipe it a little extra to get to the same level of cleanliness.

  • Paul Pham Phu

    Hi Jarrod,

    Thanks for the clear overview, this is to be expected considering the cooling design of both machines.
    I just want to add one thing as I do own the Prestige 14 and the heavy thermal throttling can be mitigated by quite a bit by undervolting/underclocking both CPU and GPU (mainly the latter as 75°C thermal cap is really easy to reach).
    The performance gap will be much reduced (with the Prestige 15 always having an upper hand) and that can justify the smaller form factor of the 14.

    • Jarrod

      Cool thanks for sharing, I did test some undervolting in my review, albeit only for the stress tests rather than actually for gaming, but yes if you can lower the temps you should be able to prolong the period before throttling or possibly avoid it entirely.

  • Ollie

    Something is wrong with your gaming results. You might be running on the iGPU, but much more likely; you’re probably experiencing the GPU throttling issue, Basically, if the GPU exceeds 75c, then it throttles to iGPU-tier performance (400mhz GPU core).

    With my P14, stock, this issue quite problematic when gaming in Sport Mode, as this lets the CPU produce too much heat. However, when using Comfort Mode (e.g switch to Balanced Profile for gaming), this limits CPU power/heat, and I found that it completely prevented GPU throttling.

    (But the best solution IMO is to repaste CPU+GPU and glue another heatpipe from the GPU to the fan shroud, since this should lower temps enough to avoid overheating. For posterity; I put TG Conductonaut in my P14).

    Anyway, stock, no mods, etc, the P15 does perform better than P14, gaming performance should be a lot closer than your results, though ~20% CPU difference seems about right. For a typical user, the P15 is definitely the better choice.

    That being said, both the P14 and P15 have a 30W GPU power limit and an 81W sustained system power limit and you can bypass CPU power limits with advanced BIOS. Basically, they have the same performance ceilings. Out of the box, P!5 hits the ceilings, P14 does not. But with liquid metal, heatpipe mod, and BIOS tweaks, P14 can also hit these ceilings.

    (BiOS tweaks = IMON Slope/Offset. There are guides for this with MSI GS65. Method is the same for Prestige 14/15).

    • Jarrod

      iGPU definitely wasn’t in use during the test, I checked that first. Yes, the Nvidia GPU is throttling – that’s what’s wrong, and I guess part of the point of the post. It works WAY better for the first 10-15 minutes, I demonstrate this in the video review with the following screenshot: https://i.imgur.com/SEb8Yyt.png
      You can see the FPS of the game was high, then the FPS dips, the clock speed disappears, and the 1650 Max-Q wattage caps to 15W all at the same time. I admit I did not attempt gaming in the other modes as I only test in the best performing mode available, sounds like that’s a problem MSI should fix if the highest mode is shooting itself in the foot 🤔 but at least there is a work around.

  • Axi

    Hi Jarrod,

    I really appreciate your youtube channel with your highly detailed and reliable laptop reviews. I’d like to know your insight if possible. I have been a hardcore gamer for 15 years, used only PC until now, however because of my current work I can’t spend enough time on it. So i decided to buy a gaming laptop, this is kind of a “big deal” switch for me but I tend to be a heavy reasearcher so after a month of searching I have found one. I would like to ask your opinion about it, checked your youtube channel, but didnt find this exact model in your reviews.

    Since this will be my first laptop, i am afraid to invest 2-3000€ into something I might dislike, you know like changing from mechanical keyboard, from ultrawide big screen to 15.6, will defenitely need to get used to it so thats why I want to keep the price around 1000€.

    I would like to buy an Asus ROG Strix III G531GU, with 120hz panel, Intel Core i5-9300H cpu and NVIDIA GeForce GTX 1660 Ti gpu for 1100€. I have 2 main serious concerns with it, the keyboard and the panel. You had the opportunity to try a ASUS Scar III (G531GW), they basically have the same keys. Based on your experience could you tell me how did the keys feel for you? Like enough feedback, tactile or satisfying like on a mechanical keyboard or something different? The other question is the 120 hz panel, I guess it will be way different than your 240 hz scar III panel, but do you think it will be acceptable? And one last thing, what do you think is the asus rog strix III g531gu worth for that price?

    Thank you!

    • Jarrod

      Thanks! I haven’t had a chance to test that model out yet as ASUS never sent it over for me to test, however the spec configuration should be pretty decent, the 1660 Ti can handle all modern games with decent settings 1080p. Of course FPS will be down vs laptop (I actually have a laptop vs desktop 1660 Ti comparison coming in a week or two) but yeah still capable. From memory, the keyboard in the Scar III felt fine/average/ok/nothing special/maybe a bit mushy. I don’t really think I’d have problems using it, but of course mechanical is nicer, that’s just much harder to come by (the only options that come to mind, from lowest to highest, are Eluktronics MAG-15, Aorus 17, MSI GT75). I think 120Hz is acceptable in terms of a refresh rate, however most of the 120Hz panels I’ve tested have lower colour gamut/brightness compared to the 144Hz options. This might differ with the ASUS model depending on the panel that’s actually in use there, so you might want to look into that – again I haven’t tested that model so I don’t have data for it at the moment 😥 price is hard for me to compare, I don’t know Euro at all plus it varies by region. It seems alrightish when you consider in the US $1100 USD for the Helios 300 with i7 and 1660 Ti is decent value.

      • Axi

        Cheers,

        Thank you for your advice and insight. Yeah I would have chosen helios 300 however for some unknown reason here the helios with the same specs as the above mentioned asus costs 500 bucks more at any retailer which is insane, so thats why its off my list.

        Anyway I went with g531gu and i love it, keyboard feels great (taking into consideration that its a laptop) and you were right on point, as u have stated your concern with the panel it has a bit low bright, but its not annoying, bad or anything.

        Thanks again man for your fast reply, precise and correct, im enjoying my laptop now without any surprises, appreciate it!

        Peace

Leave a Reply

Your email address will not be published. Required fields are marked *