Review: Lenovo ThinkPad T14 Gen 4 AMD
Background
I'm coming from an first gen Framework Laptop with the i7-1165G7 and PopOS installed.
As a student I was looking for 14" 16:10 Laptop with resolution above 1920x1200 to install Linux on. Excellent Keyboard, little to fan noise during web browsing, good build quality. Battery life of 6 to 7 hours would be enough.
I finally decided for the Lenovo ThinkPad T14 Gen 4 AMD. In the following I want to share my experiences with this Laptop after one week of usage.
My unit is configured with 7840U processor, 16 GB RAM, 2880x1800 OLED panel, 52,5 Wh battery. I installed Ubuntu 23.10 and Windows 10 in dual boot. The following review is based on the experience of using the Laptop with Ubuntu as this is my primary OS.
Build Quality
Both Laptops have a good build quality. Both of them feel very tanky. The weakspot on the Framework is the wobbly lid, which on the Thinkpad feels al solid as the bottom. The aluminum casing on the Framework feels more premium compared the plastic? of the Thinkpad.
Installation Process PopOS, Ubuntu 22.04, 23.10.
First I tried to install Pop!_OS, as this is the distro I was using before. It used already Kernel 6.6. Installation went smooth. During usage I unfortunately had several issues:
- 60 Hz mode selected in Gnome Settings didnt't apply
- usage of an external monitor would turn off the built in screen, some times hangups screen would only turn on again after closing and reopening the lid
- suddenly stuttery animations, especially after wake up from sleep
In the end I decided to my luck with Ubuntu 22.04 and 23.10. Here everything worked. For installation I used the graphics safe mode. On 22.04 I upgraded the latest OEM Kernel, which was 6.5. Kernel 6.5 was already used by 23.10.
After installation they would automatically download several 100 MB's of firmware updates. I ran into no issues.
Inputs
The Keyboard is great. Key travel is comparable to Framework but they keys are much stiffer and provide better feedback while typing.
Although the touchpad on the Thinkpad is smaller compared to that of the Framework, I overall prefer it. I'm able to move the mouse with greater precision. It reacts more precise to suble finger movements.
Screen
My unit has the 2880x1800 90 Hz OLED screen. Compared to 13,5" 3:2 2256x1504 IPS screen of the Framework Laptop, this screen is maybe 0.5 cm less tall but several centimeters more wide. I prefer the screen of the Thinkpad in every aspect. The colors are much more intense due to the OLED screen. Text clarity is improved mainly due to drastically reduced reflections and somewhat due to the higher PPI. In the end I'm comfortable using a smaller text size on the Thinkpad compared to Framework Laptop. I use Laptop mostly indoors and found a brightness of around 25 % to be enough to view everything comfortably, which also benefits battery life. Unfortunately automatic brightness control isn't offered in the Gnome settings.
With having such a high PPI, at default the GUI was rendered very small. To improve readability I increased the size of the GUI by setting the font scaling factor to 1.4 in the Gnome Tweaks app. A more processing intensive alternative would be to use fractional scaling.
Fan Noise
I had no issues with fan noise. During my every day activities like web browsing, coding in VScode, E-Mail and note taking in Obsidian the fan usually wasn't spinning. The CPU temperatures hovered between 30 °C and 35 °C. The fan only starts spinning somewhere beyond 40 °C and stays unnoticeably quiet for a long time. I noticed that the cooling solution of the T14 Gen 3/4 AMD uses two heatpipes (like the P14s AMD shown here) while other Thinkpads like the T14 Intel or T14s only use a single heatpipe, which is bigger. I'm not sure how this impacts the overall cooling capability.

System Performance
The biggest surprise to me was how good suspend and wakeup from suspend works on this Thinkpad. The T14 Gen 4 only supports the S0, S4, S5 sleep modes, which I confirmed by entering this:
$ sudo dmesg | grep -i "acpi: pm: (supports"
[ 0.498921] ACPI: PM: (supports S0 S4 S5)
I confirmed the sleep method s2idle
with the following command:
$ cat /sys/power/mem_sleep
[s2idle]
When opening the lid the lock screen lights up immediately and one can start using the Laptop. Never had a Laptop that felt so close to Smartphone in terms of wakeup latency.
During active usage the T14 feels very responsive. Starting, interacting or switching between apps all happens with notably quicker coming from the i7-1165G7 in the Framework Lapop. Its awesome. Using Speedometer 2.1 to measure system responsiveness, the Thinkpad T14 scores roughly 50 % better (300 vs 200 in Firefox/Wayland) when compared to the Framework.
The boot time of the Laptop is not great, not terrible. Using the default, Grub, with Ubuntu being the only OS and immediatly hitting the enter-key, I measured the following time from cold start:
$ systemd-analyze
Startup finished in 7.761s (firmware) + 2.774s (loader) + 1.532s (kernel) + 7.772s (userspace) = 19.840s
graphical.target reached after 7.715s in userspace.
Speakers
The speakers are a weakpoint on the Thinkpad. After installation of Ubuntu they lacked bass and sounded very tinny. Fortunately I found this recommendation on the Arch Wiki to improve the sound quality by downloading some presets calibrated for the speakers. It does improve the sound quality notably. The sound gets more bass and sound overall rounder and warmer. I still think the Framework speaker sound a bit better, also I prefer their down firing positioning which allows me to close the lid without obstructing them.
Power Consumption / Battery Life
My energy settings were managed by the default power-profiles-daemon
. To estimate battery life, I measured power consumption using the tool powerstat
.
Active Usage
My everyday workload looks like the following:
- Foreground User Apps: Firefox, Thunderbird, Obsidian
- Background User Apps: OneDrive, Nextcloud
- Radio: WLAN, Bluetooth on
- Gnome Power Mode: balanced
I use the Laptop mostly indoors and found brightness level of 20 % to 30 % to be enough for comfortable usage. I set the refresh rate to 60 Hz in Gnome Settings and use dark modes where ever available.
Out of the box the power consumption was quite high at around 9 W to 13 W while browsing through the web which meant a battery life of 4 h to 5 h. At first I thought this was because of the OLED panel which for example Notebookcheck pointed out to be quite power hungry.
To investigate the problem I installed tlp
. While running tlp-stat
I noticed that the energy-performance-preference (EPP) for the processor was set to performance. After changing this to balanced_power
system power consumption improved to 7 W to 8 W, which equates to a battery life of 6 h to 7 h. There is also the EPP power
, but other than making the system sluggish I couldn't measure any reductions in power consumption.
The following command changes the EPP to the more efficient balanced_power
mode:
echo "balance_power" | sudo tee /sys/devices/system/cpu/cpu*/cpufreq/energy_performance_preference
In Firefox I had to manually enable hardware decoding to reduce processor load when streaming YouTube videos. To achieve this I opened about:config
and set the flag media.ffmpeg.vaapi.enabled
to true
.
Suspended
Suspended is minimal. Powerstat measured 0.17 W which would equate to a loss of 1 % points of battery every 2 to 3 hours. This is also what I observed.
These are the the results of my measurements using power-profiles-daemon
on balanced
and having EPP set to power
manually.
Usage | Power Draw | Battery Life (52,5 Wh) |
---|---|---|
Sleep / S2idle | 0.17 W | |
Idle, 25 % brightness | ~5 W | |
Light (Browsing, E-Mail, Coding, Note Taking, brightness 25 %) | 7 W to 8 W | 6 h to 7 h |
YouTube Streaming, full screen, HW decode on, 25 % brightness | 10 W to 11 W | ~5 h |
Additional Measurements
To get a better feeling on the impact the OLED panel has on power consumption I made several measurements.
Usage | Ubuntu 23.10 Kernel 6.5 | Dark Desktop 25 % base line |
---|---|---|
Idle, 25 % brightness, solid black background | 4,6 W | -0,2 W |
Idle, 25 % brightness, dark desktop | 4,8 W | 0 W |
Idle, 25 % brightness, dark desktop 90 Hz | 4,9 W | 0,1 W |
Idle, 25 % brightness, Wikipedia Wayland | 5,8 W | 1 W |
Idle, 100 % brightness, dark desktop | 4,8 W | 0 W |
Idle, 100 % brightness, Wikipedia Wayland | 9,7 W | 4,9 W |
My key take aways are:
- displaying solid black vs dark grey has a small impact on power consumption of only 0,2 W
- at my default brightness level of 25 %, displaying mostly white content like websites or PDFs has a manageable increase in power consumption of 1 W
- displaying white content at max brightness is vastly more expensive that at lower brightness levels
- displaying dark grey content at max brightness doesn't seem to increase the power by much compared to lower levels of brightness
- for static content, increasing the refresh rate from 60 Hz to 90 Hz has minimal impact on power consumption
Conclusion
In the end it was a easy decision to switch from the Framework Laptop Gen 1 to the Thinkpad T14. My favorite aspects are the screen, keyboard and the overall system responsiveness. The things that I will miss are the better speakers and the slightly thinner chassis on the Framework. After manually configuring some settings battery life is now comparable to the Framework. Fan noise for every day tasks is on both systems no issue. However under load the fan inside the Thinkpad will get much less annoying compared to that of the Framework.
For reference this content was displayed during the measurements:



Conclusion
In the end it was a easy decision to switch from the Framework Laptop Gen 1 to the Thinkpad T14. My favorite aspects are the screen, keyboard and the overall system responsiveness. The things that I will miss are the better speakers and the slightly thinner chassis on the Framework. After manually configuring some settings battery life is now comparable to the Framework. Fan noise for every day tasks is on both systems no issue. However under load the fan inside the Thinkpad will get much less annoying compared to that of the Framework.