Benchmark your computer @4K with Handbrake 1.1 and H265! NOW WITH ZEN3!

[rgba]

i7-8700 non-K, 16 GB DDR4-2666 CL14
encoded 1806 frames in 387.57s (4.66 fps), 11820.04 kb/s, Avg QP:29.09

 

[Nathan Gall]

encoded 1806 frames in 775.24s (2.33 fps), 11820.04 kb/s, Avg QP:29.09

CPU: Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz, Ram: 16270 MB,

My little long in the tooth pc

 

[Nathan Gall]

encoded 1806 frames in 1338.15s (1.35 fps), 11820.04 kb/s, Avg QP:29.09

CPU: AMD FX-8320E Eight-Core Processor, Ram: 8156 MB,

and my really really long in the tooth pc

 

[Fanatical Meat]

Just for giggles:

HandBrake 1.1.1 (2018061800) - 64bit
OS: Microsoft Windows NT 10.0.17134.0 - 64bit
CPU: Intel(R) Core(TM)2 Quad CPU Q9650 @ 3.00GHz
Ram: 8191 MB, (edit: DDR800 Patriot)


encoded 1806 frames in 3439.20s (0.53 fps), 11820.04 kb/s, Avg QP:29.09

 

[Fanatical Meat]

@ElFenix looking feedback on my result, where do I rank?

 

[ElFenix]

@ElFenix looking feedback on my result, where do I rank?

i need to get my computer with excel on it back online (it's been a long process), but just eyeballing it, i'd say you're near the top. if you sort by slowest, that is.

 

[Fanatical Meat]

i need to get my computer with excel on it back online (it's been a long process), but just eyeballing it, i'd say you're near the top. if you sort by slowest, that is.

Think I should go water cooling and OC?

 

[ElFenix]

Think I should go water cooling and OC?

i may try that with a pentium 4 i've been meaning to test.

wonder if it'll break a quarter of an fps (pretty sure it's single core)

 

[PeterScott]

Think I should go water cooling and OC?

You could OC a bit. My Q9400 (lesser version of your chip) is scoring a bit more than yours, earlier in the thread, with a OC to 3.2Ghz.

The sad thing. This isn't some old computer I have around, it's my only computer.

 

[Fanatical Meat]

You could OC a bit. My Q9400 (lesser version of your chip) is scoring a bit more than yours, earlier in the thread, with a OC to 3.2Ghz.

The sad thing. This is some old computer I have around, it's my only computer.

Same here, however I am amazed at the life of my machine, 9 years old, still games adequately for my needs only recently have I started to have random crashes likely due to the power supply.

 

[PeterScott]

Same here, however I am amazed at the life of my machine, 9 years old, still games adequately for my needs only recently have I started to have random crashes likely due to the power supply.

Awesome. Another long in the tooth, well used gaming PC.

Mine with Q9400/8800GT is from 2008. It was great gaming for at least 5 years, though in recent years, I have to stick to older games, but I only get games from GoG anyway.

But otherwise it has had been a great PC for 10 years. OC to 3.2GHz and stable the whole time, and on nearly 24/7 serving as DVR recording OTA and compressing Video, gaming and general usage.

 

[ChronoReverse]

HandBrake 1.1.1 (2018061800) - 64bit
OS: Microsoft Windows NT 10.0.15063.0 - 64bit

encoded 1806 frames in 658.39s (2.74 fps), 11820.04 kb/s, Avg QP:29.09
i7-4790K running at 4.4GHz for all cores.

encoded 1806 frames in 644.52s (2.80 fps), 11820.04 kb/s, Avg QP:29.09
i7-4790k running at 4.5GHz for all cores.

Seems to scale exactly to the frequency bump.

 

[kjboughton]

i need to get my computer with excel on it back online (it's been a long process), but just eyeballing it, i'd say you're near the top. if you sort by slowest, that is.

I look forward to seeing my 3-year-old (plus) dual Haswell-E system at the tippity-top of the list.

 

[Nathan Gall]

Anyone rush out and order the Threadripper 2990WX to benchmark against this... come on, someone must have too much money handy? inquiring minds want to know!

 

[Fir]

Anyone rush out and order the Threadripper 2990WX to benchmark against this... come on, someone must have too much money handy? inquiring minds want to know!

Handbrake isn't optimized for this! Avg CPU load was approx 50% the entire run:


encoded 1806 frames in 216.35s (8.35 fps), 11820.04 kb/s, Avg QP:29.09

2990WX @ 41x100
128GB DDR4 @ 2400
1080Ti FE
Asus ROG Zenith Extreme Bios 1402
Samsung 830 256GB SSD

 

[The Stilt]

Handbrake isn't optimized for this! Avg CPU load was approx 50% the entire run:


encoded 1806 frames in 216.35s (8.35 fps), 11820.04 kb/s, Avg QP:29.09

2990WX @ 41x100
128GB DDR4 @ 2400
1080Ti FE
Asus ROG Zenith Extreme Bios 1402
Samsung 830 256GB SSD

Nothing to do with Handbrake, its due to the characteristics of modern video encoders.
To get close to 100% utilization you need the workload to utilize all 64 threads.

At 4K resolution you can utilize 33.75 threads (2160 / 64), in theory.
33.75 / 64 = 52.734375% maximum utilization.

Practically the scaling takes a nose dive well before the theoretical limit.
At 2160P you can expect around 75% scaling on 32 cores.

 

[Fir]

So we need some 8K60 footage.

 

[kjboughton]

My 2014-based dual Haswell-E system still #1. Where's that leaderboard again?

 

[kjboughton]

And we would expect that your 36 cores could easily beat 16 cores, but there are some threadripper results in here anyway,

My 36 cores also beat TR2 32 cores. Shocker.

 

[Markfw]

My 36 cores also beat TR2 32 cores. Shocker.

OK, so you have more cores, so you are faster, so whats your point ?

The cheapest I can find on ebay is $1650 EACH, and you have 2 of them, and a $500-600 motherboard vs TR2 is $1800. So $3800 (Xeon) vs $2120 (TR2) and yours is faster. Gee, no surprise here. Money quite often wins.

 

[Fir]

My 2014-based dual Haswell-E system still #1. Where's that leaderboard again?

Try the new Blender benchmark (long version) and post results in that thread.
I'm curious.

I can post CB15 screenies from a very expensive four socket 8180 system (hint, it's over 11,000!) but I won't because it's not mine to "play" with.

 

[The Stilt]

My 36 cores also beat TR2 32 cores. Shocker.

No surprises there, as HSW is significantly faster clock to clock in HEVC due to AVX2.

 

[kjboughton]

OK, so you have more cores, so you are faster, so whats your point ?

The cheapest I can find on ebay is $1650 EACH, and you have 2 of them, and a $500-600 motherboard vs TR2 is $1800. So $3800 (Xeon) vs $2120 (TR2) and yours is faster. Gee, no surprise here. Money quite often wins.

You assume price changes linearly and in only one direction with no chance for more than one reversal. You know you can't take the price "then" and the price "now" and draw a line between them and call that the price line any more than you can take the price now and write up a quote and present that as my total cost. Completely disingenuous line of reasoning.

I am making an important point (while simultaneously puffing my chest). Here it is. Are you ready?

I. Was. Right.

I called the present state of the market years ago and purchased used system components at deep discount and put together a build that continues to wipe the floor of everything, including the brand new TR2.

Where's that leaderboard now?

 

[Hitman928]

You assume price changes linearly and in only one direction with no chance for more than one reversal. You know you can't take the price "then" and the price "now" and draw a line between them and call that the price line any more than you can take the price now and write up a quote and present that as my total cost. Completely disingenuous line of reasoning.

I am making an important point (while simultaneously puffing my chest). Here it is. Are you ready?

I. Was. Right.

I called the present state of the market years ago and purchased used system components at deep discount and put together a build that continues to wipe the floor of everything, including the brand new TR2.

Where's that leaderboard now?

You discovered you can buy used server class hardware for deep discounts. Congratulations. That isn't exactly a secret and has been practiced and suggested by many on these forums over the years. Most people, however, don't have a use for this type of hardware and some don't want to buy used, even at steep discounts.

You also have to specify your use case to say, "it wipes the floor of everything." Especially with the new spectre mitigations, I wouldn't be surprised if there are some use cases where the latest TR would outperform your system and significantly so. There's also power use concerns as well as possible maintenance costs associated with your system that people might not want to deal with. It sounds like you built your system for a specific purpose and are very happy with it, and that's great, but to come into a thread and say based on this 1 test my system rules the world and I am a genius system builder is a little much.

 

[UsandThem]

You discovered you can buy used server class hardware for deep discounts. Congratulations. That isn't exactly a secret and has been practiced and suggested by many on these forums over the years. Most people, however, don't have a use for this type of hardware and some don't want to buy used, even at steep discounts.

You also have to specify your use case to say, "it wipes the floor of everything." Especially with the new spectre mitigations, I wouldn't be surprised if there are some use cases where the latest TR would outperform your system and significantly so. There's also power use concerns as well as possible maintenance costs associated with your system that people might not want to deal with. It sounds like you built your system for a specific purpose and are very happy with it, and that's great, but to come into a thread and say based on this 1 test my system rules the world and I am a genius system builder is a little much.

Dude came in here to peacock around about his PC, and at the same he's "simultaneously puffing my chest", and you have go and pee on his self-declared "victory" party.

Hope you feel better, and good day to you sir!