Great review, thanks Allen! Just ordered one of these for a client and excited to receive it and play with it during setup. ;-)
-Steve
"are shipping with 1000-nit display options and 5 MP webcams whereas the Latitude 9330 is stuck at 500 nits and 2 MP"
Yeah, you forgot to mention that 1000-nit display is the shitty privacy one that HP uses which is terrible.
2 MP webcam is, I think, enough for a laptop. You won't be recording 8K video with that.
Another primitive logic overpriced system. There are no excuses for a laptop with no discrete GPU to have cooling issues, noise and related ailing's.
Quote from: LL on November 20, 2022, 00:21:37There are no excuses for a laptop with no discrete GPU to have cooling issues, noise and related ailing's.
I think you forgot this is an Intel system. Look at the stress test energy consumption. 36W sustained under load? Sorry remind me what TDP of the i7 1260U is? 9W? Does that number mean anything? Might as well call it an i3-1220P (28W) right? Is that a cooling issue? Or does it look to me like "AMD is wiping the floor with our chips at 15W so we will run our chip at 30W to beat them" issue?
"A higher refresh rate would have improved touchscreen responsiveness as well."
Not necessarily. You need a combination of
1) high refresh
2) low response times
3) low polling rate
4) software touch input optimization
for a touchscreen device to feel "responsive". In 2022, there are no laptops with all 4, whereas most mobile tablets and smartphone displays already do so.
Dell products are horrible. This company does not even have live staff residing state side. It outsourced all its customer service to Asia and this staff doesn't seem to know much about the operation of Dell products. Don't take my words for it, just verify it on your own, before purchasing a Dell product. Also be careful with their warranty, as they don't honor it. I speak from experience. I had their staff give me the run around, scheduling services, then marking them as completed even though they never completed the work order. So, just be careful.
Quote from: Hernandez on December 31, 2022, 08:14:18Dell products are horrible. This company does not even have live staff residing state side. It outsourced all its customer service to Asia and this staff doesn't seem to know much about the operation of Dell products. Don't take my words for it, just verify it on your own, before purchasing a Dell product. Also be careful with their warranty, as they don't honor it. I speak from experience. I had their staff give me the run around, scheduling services, then marking them as completed even though they never completed the work order. So, just be careful.
Get ProSupport and you'll have no issues. They even send a person out to replace parts. Been a great experience for me.
This site is awful when it comes to everything DPC, and considering that's the main attraction here I feel vindicated in being harsh about this. This entire website is objectively misinformation when it comes to DPC, but I'll explain why so people can learn:
1) The testing methodology is terrible here. Every single LatencyMon test MUST be ran for exactly the same amount of time or you CANNOT rightfully compare one computer to another. This is basic troubleshooting and benchmarking 101 information, eliminate the variables. In other words, every test should be 5 minutes, not a minute more or less.
2) This site sorts the tables according to the highest SPIKE in DPC, which is not at all indicative of overall performance. Yes, spikes matter, but what matters far MORE is the AVERAGE DPC latency which is on a different tab in LatencyMon than what the owner's constantly post screenshots of. That is how the tables should be sorted, and this information is more critical than the spikes.
3) Another huge testing methodology flaw is there is way, way, way too much crap running on these computers, evident in the task manager. Handles over 100,000 ROFL, good grief. ALL tests must be performed on a CLEAN install of Windows, without ANY drivers installed. That's how you get a baseline for the HARDWARE. Or you could ONLY install the required drivers (no optional drivers or other bloat) and then record that data.
4) ALL tests must be done on exactly the same build of Windows. You also must not have an active internet connection because the background processes in Windows will use this and interfere with the benchmarking. This is again, basic benchmarking 101 information. You also do not test DPC latency by opening up a browser, playing videos, etcetera, because that has way too many variables. Idling at the desktop with or without drivers is really all you need, don't overcomplicate it because once you start running other software like a browser you then add a ton of variables, such as coding that might not play nice with very specific hardware or drivers, etcetera. That's why we just record data while idle at the desktop. In addition, always perform a reboot before benchmarking, then wait 5 minutes before starting to record so the bulk of the background activity can die down.
There's more I can get into, but clearly there is more wrong on this site regarding DPC than there is correct. All of these DPC tests are completely invalid and cannot be used at all to compare one laptop to another. It's completely inaccurate, I cannot stress just how bad the tests here are. I've made this comment on both the lowest and highest DPC laptops to help get this information seen so the viewers know that the most important aspect of this site isn't valid, and hopefully the owners will not knee-jerk at me, and instead take this constructive criticism to heart and fix their site.
What DPC Expert stated makes sense IF it's an offline pro-audio workstation performance that is of interest. (And I agree, that is of interest but 5 minute test is not enough! Statistical results should be posted instead of just average or spikes)
What would be very interesting testing-wise, is spikes AND input lag in games (or game-like load test) with audio,gpu,mouse,keyboard,network all being utilized simultaneously while also streaming textures from SSD during the test. And the buffer size of the audio should be such there's no detectable lag from mouse/keyboard input to the moment audio is heard. On correctly working computer, you can play a drum machine fine from keyboard or play shooter games, with internet active and disk active. Audio should be on USB bus. And complete statistics should be posted, spikes, quartiles etc.
Two additions to previous post.
1) I meant frame time spikes and audio latency spikes regarding to game-load.
2) What buffer size should be used for audio latency testing?
"The human ear detects half a millisecond delay in sound.
10.6.2021 — Researchers discover how accurately the human ear can detect time delay in sounds in different frequencies." From Aalto university.
At a sample rate of 44100 Hz 1ms corresponds to 44.1 samples.
That means testing should use smallest available buffer size (usually 64 samples at 44100 Hz) and monitor drop outs while every I/O port on the computer is active.