Some Minecraft Mini Stress Tests I Did for Fun

I had trouble saving and publishing this post.

I'm tired, so I'm gonna edit this later this week.

DISCLAIMER: Results may vary between Minecraft and Hytale.

After waiting since late 2019, I finally got my gaming laptop on June 13, 2020. Thank God that, after about an hour of searching tech stores, we found this in the only store that had this laptop in stock.

This is a massive morale boost since, number 1, I don’t have to feel guilty about trashing my parents’ laptops with my own work (legit or dirty). Number 2, I can do any kind of work much faster and I have new ways to work, like game development. Number 3 and most importantly, GAMING!!! Especially HYTALE!!!

Yeah, I kind of improvised that introduction. I suck at them, remember?

Plus, I didn’t take this post super seriously, which is why I posted it on Wednesday instead of the usual Sunday.

Here are some of my gaming PC’s specs:

• Brand/Model: Acer Nitro 5 AN515-54-56VR
• CPU: Intel Core i5-9300H @ 2.40 GHz (8 cores)
• RAM: 4 GB*
• Graphics Card: NVIDIA GeForce GTX 1650 (4 GB VRAM)

*I’m eligible for a free RAM upgrade to 8 GB, but Acer doesn’t have RAM in stock yet. If only D&D paladins were real, their Lay on Hands feature can probably exterminate the Wuhan coronavirus.

I played with the following settings in Minecraft:

• Modded Minecraft 1.7.10 with Tinkers’ Construct (no OptiFine)
• Amplified world type (creates super tall mountains which supposedly require a “beefy” computer)
• 12 chunks render distance

Clouds were turned off since they were obscuring the scenes and they had no noticeable impact on my PC anyway.

In my standard no-damage runs, my FPS ranges between 90 and 130. I’m glad it never dips below 70 because, holy dang, I’ve had so many near-failures such as when hitting zombies spawned zombies dangerously close to me.

I did my first stress test by spawning 300 villagers, but first, I wanna talk about Hyrule Warriors and Fire Emblem Warriors to illustrate my point.

In Hyrule Warriors and Fire Emblem Warriors, when scoring a lot of kills with a Musou attack (yes, they’re still called Musou attacks in the Japanese versions), the game displays how many kills were made in that one attack, which typically number from 30-120, often enough to clear an entire fort. That KO count sure does save me time that would otherwise be wasted by manually taking a headcount. Surprisingly, having so many soldiers on-screen does not cause noticeable FPS lag.

While I think it’s rare that a Hytale Adventure Mode situation would involve a hundred enemies on-screen, I tested it in Minecraft anyway just in case I REALLY get in Varyn’s nerves.

One Pair Up Musou attack from Ryoma and Caeda took over this entire fort and left the enemy Chrom badly injured. No frame rates were harmed in the making of this footage from Linkmstr.

I think villagers have a slightly higher block count than the standard humanoids (like players and zombies) since they have a big nose and folded arms as opposed to players having only a single cube for a head and one block per arm.

I just wanna do Link's Spin Attack on 'em, man.

I’m quite frankly impressed by this. I like to be Crazy-Prepared, so having this much FPS for a Dynasty Warriors-size headcount keeps me thinking fast in combat while opening opportunities for me to make scripted war scenes.

Next, I surrounded as many villagers as I could with hundreds of blocks of TNT since TNT is a notorious PC killer. I did not count the number of TNT blocks I placed, but if I had to take a guess, it’s probably around 500.

I know this is yet another impractical situation in Hytale since even 10 well-placed blocks of TNT in a structure that’s the scale of Roddan’s custom builds can cause a lot of difficult-to-repair damage with minimal FPS lag, but gotta stay prepared in case some superboss has 3 full-grown dragon minions spewing out full-power dragon breath at the same time.

Say goodnight.

Astoundingly, my FPS stayed above 24, which is the FPS of movies. The entire explosion scene stayed relatively smooth and consistent. Normally, I expect it to dip to single digits at best or freeze for a few seconds at worst since those are what usually happen in YouTube videos and in my previous tests with my parents’ PCs which only had weaksauce Intel HD Graphics. I wouldn’t consider this playable in a boss fight or PvP situation, but in a fight against easy foes like zombies, this is still more than enough.

Some even less serious tests I did:

Using Windows 10’s built-in Video Editor, it took 3 minutes and 25 seconds to export a 10-minute-long 1080p video.

I blew up a slightly more practical bundle of 100 TNT blocks. FPS dipped to 50, which is manageable though not ideal in PvP.

More than 1,000 villagers on-screen. Dynasty Warriors 2 for the PS2 is advertised on the back of the box as having "over 1,000 combatants per stage".

More than 2,000 villagers. This is where things get broken for real...but not in the way I expected.

The mobs moved normally, but my inputs were heavily delayed, perhaps by 2 to 5 minutes. This affected my attacks, block breaking, and the sound effects of placing blocks. Trying to lower the number of villagers with TNT was a pain because of the waiting time—1.7.10 doesn’t have @e which targets all entities in a command.

The game re-stabilized when I lowered the number of entities to 800, which is still a massive number for a video game army.

Oh well, so much for me attempting Thomas Frick’s 10,000 mob “challenge”... It’s not worth it when I break the game at just roughly 20% of that number.

This was...the most disappointing of the tests.
I imported Powerbyte7's Kweebec model into Unity, put box colliders on all the individual body parts, and wrote a script that spawns a Kweebec upon pressing Space. I was only able to spawn 17 before FPS dropped to slightly above 60.

I think I got 20-ish FPS with 30 Kweebecs.

But take this test with a huge grain of salt. I'm still very new to the game development scene and I just experimented like a curious little kid. I don't know about this optimization stuff.