Page de post

Idéalement, ce serait le titre du post en question qui serait ici. Bientôt!

Hands-on problems Part 3: What difference does the i9-11900H makes inside the Dell XPS 9510?

In part 3 of my ongoing woes surrounding the much-more-complicated-than-it-seems task of purchasing a new laptop, and in English for a change, I am now reviewing the Dell XPS 9510, once more, now with a working HDMI output. This time, however, I have ordered the configuration that contains a 1TB SSD (a Samsung PM9A1 which will soon star its own post, you’ll want to see this!) and an Intel Core i9-11900H. For more details, please review the following chart:


Old XPS 9510

New XPS 9510


Intel Core i7-11800H

Intel Core i9-11900H

CPU configuration

8x Willow Cove @ 2.3GHz, 10nm, up to 4.6GHz

8x Willow Cove @ 2.5GHz, 10nm, up to 4.9GHz


512GB M.2 PCIe Gen 3x4 by Micron

1TB M.2 PCIe Samsung PM9A1


16GB DDR4 3200MHz, Dual Channel

High Density Memory (1x16)


NVIDIA GeForce RTX 3050 Ti, 4GB GDDR6, 45W TGP.

Battery and charging

87WHr, Fast charging in 1hr


Windows Hello IR Camera, fingerprint scanner (on Power button)


Touch Samsung AMOLED 3154x2160 pixels, 16:10, 400 nits, HDR and Dolby Vision, no pen input.


Corning Gorilla Glass 6, Magnesium alloy body, soft carbonate inside.


Soldered Killer AX1650 (Intel AX201), WiFi ax (6), Bluetooth 5.1


2x USB-C with Thunderbolt 4 + Power Delivery + DisplayPort 1.4, 1x USB-C 3.2 Gen 2 with Power Delivery + DisplayPort 1.4, Headphone jack, lock.

| Quick advice on storage

This new SKU has cost me in effect an additional CAD150, including tax, for twice the storage and a seemingly faster processor. Honestly, I wouldn’t have purchased it, were it not the only configuration that was available to ship close to the beginning of the semester, since I was planning to upgrade the main SSD to a Samsung 970 Pro or even an 980 Pro for those blazingly fast PCIe 4.0 speeds, useful in my various creative workflows. But since I now have it, I can assess how this new set performs to satisfy your curiosity.

For some, the jump to 1TB of storage would be enough alone to motivate the extra expenditure, though I would advise instead to purchase another SSD on your own. The Samsung 970 Evo Plus 1TB could be had for 170$ and, considering most devices have two M.2 slots or more, you will end up with 1.5TB of storage, which means you have 1.5x the effective storage for far less money. You could even save more money by opting for those DRAM-less drives. That is only adequate if you plan to use those drives for data storage only and not for intense programs or as scratch disks. I plan to delve onto the subject in a future post if all of this sound too abstract for you and would just like some help in purchasing a new device or upgrading your old one.

| We want temps!

NOTE: Core temps are the temperatures recorded by the CPU, not the temperature of the device’s body.

One thing at a time, though, of course. Online, you can read a lot of things related to the H-series i9 processors, ranging from marginal performance gains to massive increase in both heat and power consumption. Looking back at my brief hands-on of the i7 version, I wrote that the cheaper SKU ran at an average of 4.4GHz under extended single-core load for a power consumption of about 30W. Well, I am happy to report that the i9 inside the XPS 9510 runs at 4.75GHz for a slightly less 28W of power. Core temperatures are wildly uneven, as they are reported to range from 75 to 90 degrees. There is thus a little bit of thermal headroom where one could imagine the i9-11900H go all the way up to its peak 4.9GHz turbo speed in single-core. This translated into a 6-8% performance gain in various performance tests. Quite frankly, you will have difficulty noticing the performance delta. Same goes for the multicore scores. The i7-11800H was able to sustain 3.25GHz on all 8 cores for approximately 65W of power consumption and a core temperature averaging 97 degrees, whereas the i9-11900H gives me a speed of 3.5GHz on all 8 cores with an average power draw of 63W, all this while the temperature averaged 90 degrees. The effective performance delta was once again of less than 8%.

It is surprising to note how the i9 ran cooler on my system, with higher speeds, than the i7. This should reassure prospective owners that might have been scared due to many unsourced claims made on various platforms that claim those processors run hotter. What should be noted, however, is that there is not a lot of extra performance to gain here. Unless you can get the better processor for cheap, there is no reason to buy it, at least until Dell allows it to run both hotter and faster under load. I should also note that, under load, the device does not spin up its fans at their maximum speed. One may conduct the experiment by setting the device on either Optimized or Ultra Performance thermal profiles in the Dell Power Manager utility, run some intense load for at least five minutes, then stop the load and set the thermal profile to Cool. You will have enough time to both get the system warm and to hear its fans running while your CPU works, and then when you switch the profiles, your CPU should be idling and yet, the fans will be noticeably louder. The only problem is that this profile also throttles the system components in order to maintain a lower temperature, so you unfortunately cannot elect to use that profile to compensate for the low normal fan speeds. Here, the system was throttled at about 35W, for speeds of about 2.5GHz on all eight cores and a core temperature of 60 degrees.

Hopefully Dell will consider adding an Turbo Performance thermal profile that mixes a higher TDP limit than normal with the fan speeds of the Cool profile. Who knows, it might even lead to 4GHz speeds on all cores! It would certainly suffice to get the system to 4.9GHz on one core. Some light overclocking could also be unlocked, as the Tiger Lake-H i7s and i9s support a limited overclocking of 400MHz and 500MHz respectively, which can be allowed or blocked by OEMs. That is merely speculation, however, as I have not conducted any test to support those theoretical speeds.

| And the runtimes on battery?

I will address the battery situation briefly here as I'll have something more detailed in an upcoming post but I have not noticed a higher power draw from the i9 processor in regular office/student tasks and some light image editing in Photoshop while running on battery. They are both adequate for working on the go, as long as you limit yourself to light tasks. Heavy computing or video editing is best left for when a charger is connected, as the device limits its power to about two-thirds of what it was.

| Temporary conclusion

To recap, while the disadvantages are non-existent, the advantages of spending more money for the i9-11900H processor are quite marginal and are easy to ignore. Neither the i7 or the i9 versions of the Dell XPS 9510 are slow, overly hot or otherwise bad choices. In the end, I suggest you purchase whatever model is available to you the fastest and overlook the processor, as long as it is not the i5 model which is actually slower, due to having less cores.

I will have a more detailed write-up about the XPS 9510 soon, after another interim post about the RAM and storage options (which is where I will discuss the aforementioned PM9A1 SSD!) and I will touch on my Alienware x15 as well. Truly, there is no shortage of subjects on my end, unlike those poor technological corporations that struggle to source all sorts of chips.

0 commentaire