Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I've been looking into the benchmarks, but how significant is the performance difference between the 3570k and 3770k for video rendering/encoding in real world terms?

If I go with the 3770k, I'll probably end up just getting 2 x 8GB of DDR3 1600, but if I go with the 3570k, it sort of gives me room to splurge on 4 x 8GB of DDR3 1600.

What would be the better option for video editing/encoding?

share|improve this question

2 Answers 2

up vote 3 down vote accepted

The similarities between the i5-3570k and the i7-3770k which are relevant to video encoding are:

  • GPU: Both chips have the "same" on-chip GPU (it has the same min/max clock frequency and the same number of execution units; this is very important if you want to use Intel QuickSync Video for hardware-accelerated encoding/decoding)
  • Same instruction set / supported processor instructions
  • Same number of physical cores

The differences between the i5-3570k and the i7-3770k which are relevant to video encoding are:

  • CPU Clock rate: 3770K faster clock rate means on the order of 10% faster at CPU operations
  • Hyperthreading: 3770K has it; 3570K doesn't. Hyperthreading provides a small performance boost for multithreaded applications. Certain video rendering operations can be done in parallel by well-designed software so this may make a slight difference.
  • L3 cache size: 3770K has 8MB, 3570K has 6MB. More cache significantly increases performance on workloads such as this, because more work can be done within the processor itself without even going out to RAM.

You will see an absolutely enormous performance boost from the Core 2 generation to either one of these CPUs. The main difference is that you may be waiting for a rendering job for an hour and 2 minutes instead of an hour if you go with the 3570K.

As for deciding between more RAM and a 3770K, I'd go with more RAM. The disk may end up being the limiting factor if you are working with big files; lots of RAM allows the system's page cache to load most/all of the file into RAM, thus reducing disk grinding during encoding, and eliminating the disk as a potential bottleneck. The disk can easily be the slowest component of a rendering pipeline when you are using hardware-assisted encoding such as Intel QSV, which on my 3770K takes about 20 minutes to encode a 45 minute Full HD video into H.264 using ArcSoft Media Converter. That's with the source media and the destination file on a ramdisk to prevent the HDD from getting in the way.

share|improve this answer
    
I currently have a Crucial M4 256GB SSD as my main drive and will have another Crucial M4 128GB SSD as a secondary drive. Will they still be bottlenecks with the 500MB/s and 150MB/s read/write speeds? With that said, it looks like I may end up going with the 3570k + 16GB DDR3 1600. Thanks for the detailed write-up! –  SCS Aug 13 '12 at 17:41
    
Hard to say if they'll still be a bottleneck with SSDs, but I'd guess that, at the very least, they'll be less of a bottleneck than HDDs. But when you consider that RAM read performance is measured in multiple gigabytes per second and your fastest SSD is 500 MB/s, you have to wonder, if Intel QSV is able to keep up with the SSD, how much that'd slow down your render, versus having the entire file in RAM (either by page cache or an explicit ramdisk)... –  ÃŁŁǫǛȉЖΦΤїҪ Aug 13 '12 at 17:49
    
I should add that if you are doing a lot of processing in addition to encoding, most processing that video editors do is CPU-bound, not disk-bound, so even if you had the entire file in RAM, those certain slow processing operations would still be just as slow. You will see the most benefit from the memory bandwidth if you're simply transcoding, e.g. from an uncompressed format to H.264. –  ÃŁŁǫǛȉЖΦΤїҪ Aug 13 '12 at 17:51
    
In most situations, it'll be from uncompressed to some kind of 720p format for Youtube, so the extra RAM sounds like it'll definitely be a good idea. –  SCS Aug 13 '12 at 17:56

The difference between 3770k and 3570k is in HyperThreading & Clock rate. 3770k can process 8 threads per cycle and runs at slightly higher clock rates, whereas the 3570k will only handle 4 and a slightly lower clock rate.

See the spec sheets here:

Intel 3770K: http://ark.intel.com/products/65523/Intel-Core-i7-3770K-Processor-(8M-Cache-up-to-3_90-GHz)

Intel 3570K: http://ark.intel.com/products/65520/Intel-Core-i5-3570K-Processor-(6M-Cache-up-to-3_80-GHz)

The higher clock rates of the 3770k will likely show a difference under general circumstances, however, the actual performance will vary depending on the application's design, and its ability to utilize multithreading and / or RAM effectively.

Going from a C2D E8400, both of these options will seem INCREDIBLY faster.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.