XPS 13 and 2 in 1 9310 now with Tiger Lake

I take it all over to be honest. If it’s one thing the iPad does well is be portable. So anytime I’m using it for media consumption it’s great. I watch movies on it all the time. Speakers are surprisingly great on it compared to even larger laptops. I’ll have to keep that in mind.

It’s anytime I need it to be a computer it’s just lacking. For example a friend wanted me to 3D print something for him and I decided to use and iPad app to make the model. My student licenses for autocad and such have long since expired. And I was just not liking the iPad for the entire process and even had to learn more about the app on a forum because it was so limited.

 

Exact same experiences here. Ipads are NOT computer replacements. They are computer complementing devices. Nothing more.

 

I have picked up both the 7390 i5G1 variant and just got the new i5G7 Iris Xe version 9310 as well. Both are 8gb RAM and 256GB SSD.

Oddly - finding the new Iris Xe version performs either the same or worse than the G1 version from last year in most ways. It's astonishing really - not comprehensive yet, but I tested across 3 games

Civ VI - Direct 11 mode (12 runs far worse on iGPU) - medium graphics with shadows bumped down - getting about 65ms render time on both at native 1200 P rez. No advantage of Iris Xe. Now this may just be because Civ VI is pretty CPU bound, so the jump to 80 execution units provides no benefit.

Mad Max - running in 720P, mostly "normal" settings. Older G1 model runs it above 30FPS for a small test sequence driving around the open world. Largely stays above 30FPS most of the time as well. The G7 model has significant dips in the mid to low 20s frequently, and generally felt less stable.

Batman Arkham City - running 900P - mostly lower settings - Older G1 model runs 30-45FPS, but lots of framerate dips/stuttering. Largely not really that playable (maybe need to drop rez down??). New G7 actually seems to run closer to 60FPS, but suffers the same framerate dips/stuttering and is largely unplayable.

So I guess the move to 80 execution units on the new chip has a minor advantage in one title, but otherwise no advantage over the lowest end i5 processor from last year.

Any clues as to what is going on? I updated all drivers on the new Iris Xe, and even pulled the latest driver direct from Intel to see if that solved anything. Is this simply a product of the 8GB ram being too little? I guess if so, not sure why the older chipset can perform better in Mad Max. I was hoping by going to the upgrade that it would allow me to play a lot of my older PC game collection with near native rez at 30-60FPS, but perhaps that's only possible on the i7/16gb ram version? At that cost delta though, I'd be better off just picking up an xbox series S or something.

 

Update on Civ VI - while the Gathering Storm Graphics test runs the same between both machines (CPU limited?) - in game, it seems to run at a lot higher FPS, at least in the mid-game. Also, world leader animations run significantly faster on the Iris Xe GPU. On the G1 model from last year, leader animations are in the teens, whereas on the Iris Xe they were running well north of 30FPS!

I also tried another older game - Fallen Enchantress - runs about 30 FPS on the G1 model, but was 45-60 FPS on the Iris Xe. I need to test Mad Max again later - perhaps I hadn't updated the BIOS/Drivers yet on the first test. I'm not excatly operating in a very scientific manner.

 

Having a i5 in a XPS is degrading the series.

 

Thanks for the substantive feedback. lol

This is a secondary machine for me - have a nice 2060 Super equipped desktop - but always curious how many games of the old PC games library I can run. I retested Mad Max tonight - still shows worse performance than the G1 basic GPU from last year's model. This has to be a driver issue.

 

I know this machine isn't a gaming machine, but since Intel touts the Iris Xe platform - I think worth sharing some of this for anyone considering. I thought by grabbing the new Iris Xe i5-G7 equipped XPS 13 2-in-1 I could play a few older titles with some extra firepower...so far this is not working out at all.

As a general rule, I'm finding 800P is the maximum on this 8GB system - although in Civ VI, i can tolerate mid 20 FPS or even lower sometimes, so I run it at 1200P. But something weird going on with the new "i5 Xe" platform - despite a huge advantage in execution units, it just isn't really showing the delta to last years very basic UHD G1 variant with just 32 EUs. All this to say, it might be worth avoiding this i5G1 and just get last year's model. About the only place I can see an advantage is in Civ VI, and really only a moderate uptick so far.

Another note, I'm seeing on the 9310, when charging, crushes iGPU performance, whereas the i5G1 7390 XPS 13 doesn't seem to have issue maintaining performance while charging. I actually unplugged from AC during Civ VI and watched framerates jump from the teens to above 30FPS in mid-game moderate zoom, and leader animations when from teens to 30-45 FPS as well (on the 9310). That's really bizarre. I'm charging to 100% to see if performance improves across the board. I had set win 10 to max performance and Dell's thermal tool to ultra as well, so color me confused.

Total War Warhammer II
i5G1 XPS13 - 25 FPS in battle benchmark, at 800P LOW settings
i5G7 Xe XPS13 - 18 FPS (while charging)

Mad Max
i5G1 - 800P - averages well north of 30-45FPS at Normal settings, so far plays pretty well with ocassional dips into high 20FPS
i5G7 Xe - utterly unplayable most of the time (while charging) - plays much better after charging is complete

Batman Arkham City
I originally tried this at 900P, which seems too high for this 8GB system
i5G1 - 800P - 45-60FPS, minor hitching, very playable
i5G7 - about the same performance

UPDATE - I can verify - performance while charging is abysmal, but once at 100%, performance improves on the newer Iris Xe laptop. In very narrow circumstances, performance outpaces the older model - in select titles.

 

So here is a fun question. If you don't use the keyboard for what seems like 30 seconds the keyboard will go dim regardless what brightness you have it on. The seems really short to me. Is there any way to disable this or atleast increase the time the backlight stays on?

UPDATE
Ah HA!!!
So what you have to do is press f12 at boot and go into the bios. You can change it in there. Really don't know why this would even be on by default. If I want the keyboard lights off I'll turn them off.

 

Just a battery saving measure. No big deal imo. Go in and change it.

 

The screen brightness I get being on battery power. The keyboard just seems odd. If it was a minute by default I would be so much less inclined to change it. Besides its rather dim anyway to begin with.

 

I agree, but that's why it's set. Squeeze every second out of it.

 

Lol. Yeah yeah. I'm very surprised they have a bios that you can do anything inside of it. I figured it would be so locked down that you would be able to turn off hyper threading and change the time. So it was the last place I expected to find that.

 

does anyone know if you can pop keycaps off XPS 13 7390 or the 9310? As usual, I want to apply a little electronic grease to the space bar mechanism to tone down the clacky sound over other keys. I like the maglev keyboard overall but the space bar feels a little loud.

 

I don’t think so with maglev

 

Seems a bit odd. What's the charger rated for? And what does it pull from the wall at 100% vs charging?

Also what is your RAM usage when playing those games?

 

I'm using the factory 45 watt charger - go figure?

 

Wow a whole 45w. Seems a bit meek for a 15-28w + system? You could try a higher wattage PD charger.

You try testing CPU and GPU workloads between the two machines one at a time. Any different in a combination of the two would be due to power or thermal limitations. Try comparing furmark between the two machines.

With the slowness during charging vs battery it sure sounds like it's being power limited. In the notebookcheck review under power consumption you can see the charger is being maxed out.

You can see this german review of the Asus ZenBook 14 UX425EA the system can go up and over 50w on max load. It's very close to your processor and the same TDP afaik.

 

I mean it's the factory charger - and as stated, works fine with last year's model in terms of load use and reaching performance. I also have a 60w USB-C PD charger - and same behavior - just doesn't do well when charging the machine under gaming load.

But you may be right, that beacuse of Iris Xe, Dell may not have updated PL limits to operate well while charging and gaming versus last year's model..not sure the power requirements of the i5G7 Iris Xe chipset vs the i5G1 from last year. I see a new Intel driver, so I'm installing to test...also getting the 2004 windows update installing.

 

Hmm. Too bad no change with the larger power supply. Wonder how much power the charging is taking

You on the latest bios? Looks like there is possibly some related stuff in there.

Could be someting to do with the performance mode changing software too. Seen weird stuff before with that on gaming laptops.

 

Let's give it up for our little buddy here. goes toe to toe with the new Mjesus chip from apple in speed. all without the so called "optimization" of hardware and software that makes apple "magical".