Intel i7-13700k vs AMD Ryzen 7900x ?

Peat Moss

Gawd
Joined
Oct 6, 2009
Messages
543
These are the two CPUs I'm considering. I realize the Ryzen 7700x is more of a direct rival to the 3700K but price-wise right now they are very similar, and still not too far apart in specs.

I'm coming from a very old system, so whether Intel or AMD I'll have to buy a new mobo as well. I do not game, thus my decision will be based on things like multi-tasking, video editing, some 3D modelling.

Intel pros:
- slightly cheaper
- faster boot time than Ryzen (is this correct?)
- plays nicer with video codecs than Ryzen (is this correct?)
- draws less power than Ryzen when idling

AMD pros:
- 7900x 200 MHz faster than 13700K
- longer upgradability path
- more cores
- eco mode helps reduce thermals

Anything I've missed?
 
I think you got the main ones. Avx512 is present on Ryzen. More “full fat” cores on Ryzen too.

Ryzen will probably perform better for 3d modelling and video work than the intel equivalent, due to AVX512, more cores, and the architecture
 
For some video work intel iGPU will give you more, at least I imagine why you see it sometime with 50-70% better scores:
https://www.pugetsystems.com/labs/a...o-13th-gen-intel-core-vs-amd-ryzen-7000-2381/

I feel some tester goes in CPU only mode to equal the playing field and you loose that element:

At first glance, things are looking pretty darn great for Intel! In terms of the Extended Overall Score from our Premiere Pro benchmark, the Intel Core 13th Gen CPUs are around 50% faster than the equivalent AMD Ryzen 7000 CPU, What you don't see is that our overall score is a bit skewed due to just how much faster Intel CPUs with Quick Sync are when running the 4K H.264 and 8K HEVC sets of tests. you get a clear look at just what is throwing off the overall score. For longGOP codecs like H.264 and HEVC, the Intel 13th Gen CPUs are anywhere from 60% to 2x faster than AMD.

But who that buy cpus for that workload would disable quicksync ?

I would validate for the exact program and things you want to do if possible, but it could be a non brainer Raptor Lake winner.

7900x vs 13700k comparison in maybe relevant application:


There is often a big gap in power consumption in favour of the 13700K has well
 
I am not sure how much better having those feature on the threadripper but impossible to use if you one day want to use it has an Epyc cpu instead would be here, it seem to make a lot of sense to me.

How much wasted working but disabled silicon is there in the field currently ?
 
I am not sure how much better having those feature on the threadripper but impossible to use if you one day want to use it has an Epyc cpu instead would be here, it seem to make a lot of sense to me.

How much wasted working but disabled silicon is there in the field currently ?

It’s more that I think you should get what you pay for. Eg if I buy an intel KF model processor, I expect that I won’t be able to use the graphics module on the processor, but will be able to overclock it.

I expect a discount cause I can’t use the graphics.

If I buy a processor that has graphics/encoding, but I need to pay a subscription for its use- I will not be happy about that idea..

On the other hand if it is “pay to unlock” then it is permanently unlocked, then I am less unhappy about that. In fact if there was an 8 core processor that I could later unlock to 10, 12, 14 and 16 core if I needed with a one off or incremental payments, I would actually probably go for that
 
Last edited:
It seems that things like quick sync and some programs that are more Intel friendly could make a 13th gen build a no brainer. Classic case of use case here. What programs and workloads are you going to run and which tool is best for the job. What’s amazing right now for us consumers is that Raptor lake and AM5 are both excellent and it’s hard to go wrong.
 
It seems that things like quick sync and some programs that are more Intel friendly could make a 13th gen build a no brainer. Classic case of use case here. What programs and workloads are you going to run and which tool is best for the job. What’s amazing right now for us consumers is that Raptor lake and AM5 are both excellent and it’s hard to go wrong.


MS Office, DaVinci Resolve, Navisworks, some document management software, Adobe Acrobat Pro, Abbyy Fine Reader, EndNote, etc. My impression is that Intel works more closely with Microsoft than AMD does, but maybe that's just a misperception.
 
MS Office, DaVinci Resolve, Navisworks, some document management software, Adobe Acrobat Pro, Abbyy Fine Reader, EndNote, etc. My impression is that Intel works more closely with Microsoft than AMD does, but maybe that's just a misperception.

Which graphics card are you planning on using?
 
Change the thread title to "Intel i7-13700K vs AMD Ryzen 7900X". The 3700K part is misleading (likely a typo) as that refers to a completely nonexistent 3rd-Gen Ivy Bridge quad-core CPU from 2012.
 
Probably RTX 3060 ti
Depending on the size of your Navisworks project one could consider the non TI version, more VRAM:

https://zworkstations.com/apps/autodesk-navisworks-workstations/

On ebay, A4000 are sometime even something possible to buy now (at least was a the crypto dumb).

it has been a while that program that supported hardware decoding only via quicksync does also for a Dgpu which could make the choice of CPU quite moot and equivalent, but one could make sure that Davinci support dGPU for the wanted codec:

https://www.pugetsystems.com/labs/a...-is-supported-in-davinci-resolve-studio-2122/
Has for that up to date enough to have RTX 4000 graph, a lot of hardware H.264, H.265 decoding in Resolve are still quicksync only
 
I keep going back and forth in my mind between the two CPUs. Can't decide.

Any other factors I should consider?
 
I keep going back and forth in my mind between the two CPUs. Can't decide.

Any other factors I should consider?
There could be a dead platform on raptor lake versus an upgradable one on AM5, but how relevant it would be is always a bit of a guess work and considering you are coming from a very old system maybe it does not matter to you.

second gen DDR-5 memory controller on Intel seem a bit better but going with higher DDR-5 would remove all notion of being a cheaper option fast.
High end motherboard PCI-5 support is higher on AMD side but could never be something relevant for the pratical lifetime of the machine.
 
While Intel did have a brief gen to gen upgrade without a total replace, historically, AMD has done better. So, it's possible (very) that by going AMD, you may have longer upgrade choices before having to replace your motherboard.

If you always buy all components new all the time, then, you can pretty much pick and choose.
 
It says so in the specs.

Yes, for the CPU. But you have to pair that with a chipset that also supports ECC for such DIMMS to actually be supported. Intel's current consumer-level desktop chipsets (e.g., Z790, B660, etc.) do not support ECC. This is clearly noted on their product pages (linked).

If you need ECC, then you need a different chipset such as the aforementioned W680. For example, on the SuperMicro X13SAE.
 
Yes consumer CPU on the server chipset (W680) can work, but apparently those motherboard are not that easy to find at good price.

Yes, for the CPU. But you have to pair that with a chipset that also supports ECC for such DIMMS to actually be supported. Intel's current consumer-level desktop chipsets (e.g., Z790, B660, etc.) do not support ECC. This is clearly noted on their product pages (linked).

If you need ECC, then you need a different chipset such as the aforementioned W680. For example, on the SuperMicro X13SAE.

Ah, thanks. This helps in my final decision.
 
You do know that ddr5 has essentially got ecc baked in yeah?

https://en.m.wikipedia.org/wiki/DDR5_SDRAM

It's not the same thing. Did you read that link at all?

Wikipedia said:
Unlike DDR4, all DDR5 chips have on-die ECC, where errors are detected and corrected before sending data to the CPU. This, however, is not the same as true ECC memory with an extra data correction chip on the memory module. DDR5's on-die error correction is to improve reliability and to allow denser RAM chips which lowers the per-chip defect rate. There still exist non-ECC and ECC DDR5 DIMM variants; the ECC variants have extra data lines to the CPU to send error-detection data, letting the CPU detect and correct errors that occurred in transit.[
(Emphasis mine.)
 
Back
Top