While I’m keeping Debian 13 Trixie on my media center computer, I’ve decided to fallback to Debian 12 Bookworm on my laptop and workstation. The more that I used Trixie on those machines, the more I realized some things that I relied on just weren’t working right. Once that software gets updated, I’ll try Trixie again, but for now, especially while I’m frantically getting things ready for classes to begin next Tuesday, I’ll rely on tried-and-true Bookworm.
Tag: Workstation
-
Debian 13 Trixie Full Install Notes
Last week, I upgraded the Debian 12 Bookworm installation on my workstation to Debian 13 Trixie. After updating my sources file and running apt update and apt full-upgrade, the installation went painlessly and quickly. However, I noticed some glitches with video playback (audio stutter) and listing directory contents on a hard disk drive (prolonged delays even though the drive was in active use and not waking from sleep). I wasn’t that interested in spending time to track down the source of these particular problems, so I decided to do a nuke-and-pave full reinstall of Debian 13.
That led to three days of frustration and reinstalling Debian 13 five times. I now have a stable installation that does what I want, but it took a lot of bashing my forehead into the desk to get here.
To be fair, some of the problems were created by me trying to configure software that I don’t have a deep understanding of using shared knowledge online that might be years old and applicable to older versions of those programs.
The one problem that caused the most headaches wasn’t my fault as far as I can suss out. The installer seemed to alternate assigning the “zero” nvme drive between my two identical 2TB nvme drives. This strange behavior led to me wiping both drives and installing Debian 13 on both during the multiple installations. When it comes to partitioning, I’m super cautious, because even though my data was backed up, I still didn’t want to wipe a drive unless it was what I wanted to do.
Besides the Debian 13 installer’s partitioning software, I wondered if it could have something to do with my motherboard’s bios. I am a few versions behind the latest release, but none of the bios’ change logs mention anything to do with code for handling the nvme drives.
I didn’t document my experience like I should, because I got to a point where I just wanted a stable system so that I could get some work done. I accomplished that at least. Though, I wanted to put this up as a potential warning in case anyone else experiences something similar.
-
Remember to Clean Your PC’s Air Intake
Due to the noise my workstation makes during AI inference, I keep it on the floor under an adjacent desk. Down there, it’s in the shadows. So, I was a little surprised how dusty the front air intake was after being in operation just a little over a month. It probably says a lot about how bad the air quality is in my apartment despite running three HEPA air cleaners in a roughly 600 square foot space. I know that it would be better for the PC to be up off the floor–on the desk, for example. Unfortunately, its noise and disco lights on the CPU fan and white light on the NVIDIA RTX 3090 make this an undesirable choice. I’ll have to remember to vacuum it every two weeks or so, and I might add a foam sheet behind the front grill to help catch more dust before it goes into the case and lands on the components’ heatsinks and fans.
I’m reminded of The Crafsman‘s “Don’t Forget Your Dust Mask.”
-
Continuing to Receive New Workstation Components
Over the weekend, I received two more pieces of my upcoming AI-focused workstation build to go along with the CPU and A4000 video cards. They are a Gigabyte B650 Eagle AX model ATX motherboard, which has four PCIe slots–one spaced for a 3-slot card like my 3090 and three spaced for single slot cards like the A4000s, and a Silverstone Fara R1 V2 ATX mid-tower case, which was the least expensive steel case without a glass window and good ventilation. My new Corsair DDR5 RAM won’t arrive until after Christmas, so the actual build will have to wait until then.
-
New AI Workstation Build Continues: Three NVIDIA A4000 Video Cards
After receiving a new AMD Ryzen 7 7700 CPU earlier this week, I received the three NVIDIA RTX A4000 16GB VRAM video cards pictured above in antistatic bags today for my new AI workstation. Brand new, these cards run just over $1000, but I got these refurbished ones from an eBay seller for just under $600 each. These three video cards will work alongside my NVIDIA RTX 3090 Founders Edition 24GB VRAM video card for a total of 72GB VRAM, which will allow me to run low-or-no quantized large language models at a much faster output rate than I currently can using the 3090 with system RAM. The limited PCIe lanes on the Gigabyte motherboard that I ordered shouldn’t be too limiting as far as inference work is concerned.
