Amd9 vs i7: AMD Ryzen 9 7900 Vs Core i7-13700K: Which Should You Buy?

The Ryzen 9 7900X fills a particular niche right below the top-of-the-line Threadripper CPUs from AMD. As such, they’re more comparable, at least on paper, to the i9 series of CPUs from Intel. The performance of the Ryzen 7900X, however, is more akin to the i7 range of CPUs, despite the cost being slightly higher than the i7.

A higher base clock of 4.7 GHz means the CPU itself runs a little hotter than average. The manufacturing process is brand new for the Ryzen 9, being a 5 nm die process. The power draw on average is significantly higher than the i7, coming in at an average of 170 watts and a nominal maximum of 230 watts of power under load.

It is slightly more efficient when it comes under heavy load, but you’ll need a more robust power supply to provide power as it sits. Performance is slightly worse than the i7 but is still quite phenomenal considering the price range.

While it might not stack up completely favorably to the i7, it is still a phenomenally fast CPU for the money. If you’re looking for i9 levels of performance, you’ll most assuredly be disappointed. Prospective customers will need to temper their expectations.

Intel i7 13700K vs. AMD Ryzen 9 7900X: 9 Must-Know Facts

Intel i7 13700K

1. Supports overclocking.
2. Compatible with DDR4 and DDR5 RAM.
3. Lower thermal draw.
4. Supports a maximum overclock of 5.3 GHz.

AMD Ryzen 9 7900X

1. Supports overclocking.
2. Compatible with DDR5 RAM.
3. Higher thermal draw at idle, less while under load.
4. Supports a maximum overclock of 5.6 GHz.
5. Has more effective cores.

Intel i7 13700K vs. AMD Ryzen 9 7900X: Which One Is Better?

So, which is the better CPU? In terms of raw performance, the edge goes to the Intel i7, which not only outperforms the AMD Ryzen 9 but does so with fewer cores and a less refined manufacturing process. Without getting into the endless performance wars, the i7 presents a better overall product and the suggested retail price is lower than AMD’s offering.

However, both of these offerings are more than capable CPUs and should provide years of issue-free computing and gaming. Whichever you choose, do keep in mind that both of these CPUs will require accompanying motherboards with a compatible socket set.

If you’re looking to repurpose older DDR4 RAM sticks, the i7 certainly has the means to leverage those until you’re able to get DDR5 sticks. The Ryzen 9 7900X might not be the absolute best choice for gaming, but those seeking to leverage the multiple cores for work in video editing and AI modeling will no doubt benefit from the additional cores.

In short, if you’re gaming, choose the i7. If you are looking to do work requiring sheer core count over anything else, the Ryzen 9 is the better choice. Either way, these are rock-solid CPUs that won’t disappoint in a variety of use cases.

Intel i7 vs. AMD Ryzen: Which CPU Is Right for You? FAQs (Frequently Asked Questions) 

Which factor is the most important for gaming in terms of CPU performance?

Overwhelmingly for most games, you’ll be looking at single-core performance over any other factor. As such, a CPU with strong single-cores can thoroughly outclass a multi-core monster in terms of performance for delivering more frames per second in a game.

If you’re looking solely at getting a stable 60 frames per second in each game, the GPU is going to be more of a consideration than your choice of CPU. Performance is overwhelmingly bottlenecked by the CPU in most cases, especially with modern GPUs and CPUs in the last five or so years.

Why are more cores beneficial for a CPU?

Multiple cores have the benefit of allowing more parallel processing threads for tasks and applications which benefit from it. It is very commonly seen for things like video editing and AI modeling. AI, in particular, relies overwhelmingly on parallel processing threads to allow for more consistent modeling procedures.

More cores may not always benefit professionals; this is especially true for those in the audio production workflow. Audio production generally relies on how much RAM your computer can hold, as well as the single-core performance of a processor.

Would either of these CPUs be suitable for office use?

Well, that would entirely depend on what sort of office you’re running. If you’re doing high-end 3D modeling, AI modeling, or video editing, then these might present a cost-effective CPU.

In lieu of getting into expensive workstations utilizing the Xeon and Threadrippers, these processors certainly have an adequate amount of power.

If you’re in the business of data processing or data entry, then these might be overkill for those purposes. Microsoft Word and Excel don’t require a ton of processing power, so you could effectively cut down on costs for acquisitions by choosing something with less overall power.

Why would you want to overclock a CPU?

Overclocking a CPU increases the base clock speed, instead of relying on the turbo boost to effectively raise it under extreme load. Overclocking leaves the performance at a constant instead, leaving it at its maximum for all purposes.

Where there can be some complications is in it presenting stability issues on your PC. Unstable overclocks can make a PC unusable, negating the performance benefits. Messing with overclocking can be a very sensitive process, so do some research before deciding to apply overclocking to your machine.

Would these CPUs be good for gaming?

Absolutely, these excel during gaming. During testing, they performed admirably, delivering stable frames to a variety of modern PC titles like Modern Warfare 2, Doom Eternal, and Forza Horizon 5 without missing a beat. If you’re looking for the heart of your setup, you can’t go wrong with either of these CPUs.

About the Author

Liam Frady

Liam is a freelance writer with a passion for professional audio, cybersecurity, and information technology. Aside from writing, he can be found in his home studio moonlighting as a mixing and mastering engineer. Outside of work, Liam can be found spending time with his family, cooking up fun recipes he found online, or making music in his spare time.

More from History-Computer

• Intel Core i7-13700k Processor | Intel Available here: https://www. intel.com/content/www/us/en/products/sku/230500/intel-core-i713700k-processor-30m-cache-up-to-5-40-ghz/specifications.html
• AMD Ryzen 9 7900X Desktop Processors | AMD Available here: https://www.amd.com/en/products/cpu/amd-ryzen-9-7900x
• Passmark — Intel Core i7-13700K — Price performance comparison Available here: https://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-13700K&id=5060
• PassMark — AMD Ryzen 9 7900X — Price performance comparison Available here: https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+9+7900X&id=5027
• Intel Core i7 13700K vs AMD Ryzen 9 7900X: performance comparison | Nanoreview Available here: https://nanoreview.net/en/cpu-compare/intel-core-i7-13700k-vs-amd-ryzen-9-7900x

AMD or still Intel | Render.ru

42RUS

Site user

Oct 17, 2014

#1

Oct 17, 2014

#1

Hello everyone, the question arose of what is better to take to work with 3d rendering, mainly 3d max + vray, there will be water cooling. Processor Intel Core i7-4790K 4000MHz 8Mb TDP-88W S1150 tray (Haswell Refresh) or Processor AMD X8 FX-9590 4700MHz (TurboCore 5.0GHz) 16Mb TDP-220W SocketAM3+ tray and which motherboards are desirable for them, overclocking will be, what power supply is needed who came across what is better for my tasks, I will be very grateful. Budget 30.t.r

Last edit by moderator: Jan 9, 2015

— — (Toer)

Expert

Oct 17, 2014

#2

Oct 17, 2014

#2

in general, iron from AMD in terms of price-quality ratio is higher than that of Intel, but it is better to take vidyuhi dzhiforsovskie as they are more optimized for most programs

Ole-Ze

Active member

Oct 17, 2014

#3

Oct 17, 2014

#3

here you also need to look at which processor socket is more promising — so that after a while you can put a more powerful percentage into it and the socket does not have time to become obsolete . .. I once bought an i7 920 for socket 1366 — and Intel released a couple of processors for it and now it makes no sense to put the older model of the processor under this socket — since there are already more powerful ones, but it has not fallen in price much. I don’t know how AMD has about this — how long can you upgrade a computer on the same socket. In my opinion, Intel has been doing something new for my socket for two years — and that’s it.
When rendering, pure mathematical performance is important — you can look here:

http://www.maxwellrender.com/benchwell
— just pay attention to the number of cores of the test system

or see comparative tests on thg.ru, overclockers.ru

HunterWolf

Expert

Oct 18, 2014

#4

Oct 18, 2014

#4

Well, if you take AMD, you immediately understand that a lot of things in modeling itself and working with polys will be many times slower and slower if you put it purely on the render, then yes, the price-performance ratio of AMD is higher, but in the coffin I saw it work when half of the processes are on one core and half of the working software is optimized for Internet technologies. I have 8350 and 4770k I don’t even try to work on AMD anymore, I have it purely for rendering and it’s still good to unpack the back archives on it, otherwise when you don’t have 5 MHz in your work, but God forbid I work at 2.5 MHz, then this is not very pleasant especially on high polys are easy to move. We must remember that in many ways your 5 MHz AMD gives 2.5 MHz

42RUS

Site user

Oct 18, 2014

#5

Oct 18, 2014

#5

Thank you very much for the advice, well, here on the Internet people advise switching to GPU rendering, they say it will be many times faster and not bother with the processors, who can advise on this case, on a budget of 30t. r, also thankful for earlier.

HunterWolf

Expert

Oct 18, 2014

#6

Oct 18, 2014

#6

Why advise something here, do you work with them, or do you think the only difference in them is that they are the only ones on the process of others on video, get yourself a working render and then think about the hardware

LMA

Caretaker

Oct 18, 2014

#7

Oct 18, 2014

#7

here you need to look at which processor socket is more promising

Click to expand. ..

This is the wrong approach. In addition to the socket itself, there is also such a thing on the motherboard as a chipset, which, like processors, is constantly updated, memory and all sorts of protocols are also updated.
Fundamentally more productive processors of newer generations, for which it really makes sense to update the system, already probably have all sorts of new gadgets that only support new chipsets on the motherboard. According to the format, the processor may and will be included in the old socket, but the old chipset of the board will either not support them at all, or will do it, but not in full.
In short, with the replacement of the processor, in the vast majority of cases it is necessary to change the motherboard, especially since the processor costs much more than the motherboard, in relation to it, the motherboard is like a consumable.

the question arose of what is better to take to work with 3d rendering, mainly 3d max + vray, there will be water cooling.

Click to expand…

If you only render, I won’t say which is better, but if you also model or do something else, then in my opinion you need to proceed from the performance per core, since operations are often encountered , which do not support multithreading, or support it to a limited extent. In these cases, a processor with higher performance per core will be noticeably faster at doing work. And Intel, it seems, has a much higher performance per core than AMD.

HunterWolf

Expert

Oct 18, 2014

#8

Oct 18, 2014

#8

I wrote about this that when modeling, they update most processes on one core and in most there is not even support for castrated AMD, so as a result, we get exactly half from AMD, namely 2. 5 instead of 5 on 8 castrated cores. And I don’t care that there are eight of them or 16 and 120 cores, one stone works, which honest AMD is 2.5 MHz

Ole-Ze

Active Member

Oct 18, 2014

#9

Oct 18, 2014

#9

LMA said:

This is the wrong approach. In addition to the socket itself, there is also such a thing on the motherboard as a chipset, which, like processors, is constantly updated, memory and all sorts of protocols are also updated..

Click to expand…

In practice, I have never seen such a thing: a manufacturer changes the north bridge for the same socket — for example, the entire Gigabyte line for my 1366 was made for several years (all revisions of all motherboards) only on Intel® X58. I do not presume to say that this is a global pattern — but of all the cases that I have met — it was just that. After all, the logic here is slightly different — first, the motherboard manufacturer receives the specifications of the chipset and processor — and the motherboard is developed on the basis of this — and not vice versa. That is, the chipset is the basis on which the motherboard is developed, and not vice versa — the chipset is produced for the motherboard.
As for changing protocols — you probably meant new instruction sets for the processor — has this happened? — so that the new percentage is not supported by the old board for the same socket? or this board did not allow the processor to execute certain sets of instructions (such as SSE2, SSE4, ACPI) — did such problems really occur? Throw off a link to a description of such a problem — it’s interesting to read.
Well, about the memory, specify, can there be any incompatibility? — DDR3 will work in any motherboard designed for DDR3 (with the exception of some cases of hardware incompatibility between some models from different manufacturers), the difference will be only in the maximum frequency and the total amount of memory that can be installed. But in the motherboards of the last few years, they are already sufficient. And DDR4 is still only in the plans.
So if the processor is planned to be replaced for about two years, then it will be possible to get by with replacing only the processor, and not the entire system. This is, of course, if a new motherboard model is taken, and of course, it will be necessary to take into account the prices for specific processor models — because sometimes it happens that top old models are more expensive than average new ones — and the performance is the same.

K D (Rebate)

Expert

Oct 18, 2014

#10

Oct 18, 2014

#10

Numbers. The price of Intel is approximately 14,600 rubles today, the price of AMD is 9,400 rubles. 4790K is 50% more expensive. According to visa 4790K in stock is about 15% faster. FX-9590 is an oven with 220 W, yes, water will not interfere with it; 4790K — 88 W, enough air.

LMA

Caretaker

18 Oct 2014

#11

Oct 18, 2014

#11

Well, about the memory, specify, can there be some kind of incompatibility? — DDR3 will work in any motherboard designed for DDR3

Click to expand…

I’m not talking about incompatibility, I’m talking about new features appearing in new generations, such as higher memory frequency. By the time a new processor appears, a new, faster memory will also appear, i.e. together with the processor, you will change the memory, and there the new video cards will arrive in time, i.e. also take a new video card, and they are more powerful, and therefore a new power supply is needed. If we are to increase productivity, then one processor is indispensable. And what is left of your old system besides the motherboard? And where are you going to put everything old, but working? Insert a new processor, memory, video card and power supply onto the old motherboard, and everything else is old in the trash or what? After all, without a motherboard, they are of no use.
I believe that if you have so much extra money to change the awl for soap every two years, then you can simply buy a whole new computer without any problems, and either put the old working one in second, or just give it to a neighbor boy. And an awl for soap — because processors for one socket are just a flea for soap, well, only if, for example, you didn’t move from i3 to i7, but among different generations of i7 there is no fundamental difference on the same socket, such the difference is to change one for another, and at the same time throw the old worker into the trash. And when a fundamental difference (this is at least a one and a half to two times increase in performance) appears, then these are already different processors and different sockets.

HunterWolf

Expert

Oct 18, 2014

#12

Oct 18, 2014

#12

Ole-Ze said:

And DDR4 is still in the plans.
So if the processor is planned to be changed for about two years

Click to expand… by spring it becomes socket 2011-3 and new processors. But for now, both the assortment and prices somehow do not allow you to think about it, but taking into account the dynamics from spring to spring of the next one (that is, somewhere in half a year), you can be puzzled by this socket and the processes for it

3DKing

Master

Oct 24, 2014

#13

Oct 24, 2014

#13

has been using AMD for a long time (from 1999 to 2014 only AMD) and everything seemed to suit everyone: the price is not high, heat dissipation is low, a good heatsink with the quietest and slowest cooler was cooled without problems, the speed seemed to be normal too. But when it took a lot to render and produce images as quickly as possible, then I ran into performance. The last stone from AMD was my maximum hair dryer x6 overclocked to 4 GHz. Later I sold it and started using the laptop at full capacity: a MacBook Pro with an i7 on board, and soon I took 2x Intel Xeon e5-2630v2 for rendering. Now I use it for work, and a laptop for visiting clients. AMD in rendering certainly merges .. and when speed and stability are required, it is better to take intel and preferably XEON.
If you choose hardware for everything and your CG is not your main income — don’t waste your money on INTEL, it won’t be justified.

Aidar

Site user

4 Dec 2014

#14

Dec 4, 2014

#14

3DKing said:

took 2x Intel Xeon e5-2630v2

Click to expand. ..

Well, the price tag is completely different, you would also compare the F1 car with the products of the domestic auto industry))). For the price of just one stone, you can assemble a system unit on AMD. Of course, when the money spent quickly pays off — no question. But as I understand the author’s budget is very limited, especially now.

3DKing

Master

Dec 25, 2014

#15

Dec 25, 2014

#15

30 tyr. As I understand it, the author already has water cooling. And it seems like we are talking only about the processor, motherboard, memory and power supply. Well, then take i7 once such a specific focus on graphics and rendering.

Aidar

Site user

25 Dec 2014

#16

Dec 25, 2014

#16

3DKing said:

30 tyr. As I understand it, the author already has water cooling. And it seems like we are talking only about the processor, motherboard, memory and power supply.

Click to expand…

Only one percent for 30k, 16GB memory (this is the minimum) is already 13-15tyr, better 32, total 60k. To this, another 15-20k for the mother and 5k BP, as a result, we get almost 100k for the mother + percent + memory + BP . .. in this case, it’s better to take Haswell then, although you will overpay for memory and mother 5k each, but it will be full-fledged 8 cores, which will be faster than 6 Aivik cores, i.e. with an overpayment of 10%, the increase in productivity will be more than 30%.
AMD with half the performance of 8-core Haswell can be obtained for 10+5+30+5=50k(perc+mother+memory+PSU), while taking 6-core Ivic the lag will be only 50%, i.e. with a twofold difference in price, it is better to take either 8x Haswell or AMD 8350.
You can also consider 4770K / 4790K, for 25 + 5 + 30 + 5 = 65k we get a performance comparable to AMD with a superiority of 10-15% towards intel. It depends on personal preferences, the required performance and the budget framework.

Last edit: Dec 25, 2014

3DKing

Master

Dec 25, 2014

#17

Dec 25, 2014

#17

so

Aidar said:

Only one percent for 30k, 16GB memory (this is the minimum) is already 13-15tyr, better 32, total 60k. To this, another 15-20k for the mother and 5k BP, as a result, we get almost 100k for the mother + percent + memory + BP … in this case, it’s better to take Haswell then, although you will overpay for memory and mother 5k each, but it will be full-fledged 8 cores, which will be faster than 6 Aivik cores, i.e. with an overpayment of 10%, the increase in productivity will be more than 30%.
AMD with half the performance of 8-core Haswell can be obtained for 10+5+30+5=50k(perc+mother+memory+PSU), while taking 6-core Ivic the lag will be only 50%, i.e. with a twofold difference in price, it is better to take either 8x Haswell or AMD 8350.
You can also consider 4770K / 4790K, for 25 + 5 + 30 + 5 = 65k we get a performance comparable to AMD with a superiority of 10-15% towards intel. It depends on personal preferences, the required performance and the budget framework.

Click to expand…

Well said!

Alexander Vorontsov 153126

Site user

Jan 9, 2015

#18

9 Jan 2015

#18

Ole-Ze said:

here you need to look at which processor socket is more promising — so that after a while you can put a more powerful processor in it and the socket does not have time to become obsolete.

Click to expand…

Never understood this logic.
Here you buy a processor. The new one will be released in 1.5-2 years.
Performance increase will be 10%
Will you immediately change to a new processor? For 10%?
The real expediency of changing the processor will appear in 3-4 years.
And by this time you will already have to change mothers. either a new memory standard will come out, or USB, or who knows what else.

And on the topic — if you have money, buy Intel.
No money — AMD.

GISmen

Active member

17 Apr 2015

#19

17 Apr 2015

#19

Alexander Vorontsov 153126 said:

And on the subject — there is money, buy Intel.
No money — AMD.

Click to expand…

That’s right.
Therefore, having a limited budget, I assembled a farm in my office on AMD 8370-E processors (8 cores, TDP max 95W). 4 months — normal flight! Let’s see what the summer will show, you may have to install air conditioning.

3DKing

Master

17 Apr 2015

#20

17 Apr 2015

#20

GISmen said:

That’s right.
Therefore, having a limited budget, I assembled a farm in my office on AMD 8370-E processors (8 cores, TDP max 95W). 4 months — normal flight! Let’s see what the summer will show, you may have to install air conditioning.

Click to expand…

Norm. AMD is actually not bad processors and is not badly chased.

comparing performance on HTTP server implementations / Habr

ogregor
Jan 2, 2019 at 10:28 pm

Time to read
2 min

Views 23K

Website development *JavaScript *Programming *

From the sandbox

Good day everyone!

This year, with the release of Flutter, a framework for cross-platform application development, there has been a rise in hype for the Dart language. Like any perfectionist procrastinating from boredom lazy I thought about comparing the performance of the server implementation of the Dart virtual machine with its potential antagonist Node.js. I will say right away that I had a glimmer of hope that Dart would win, and I would gain the holy grail that would give me superiority over potential competitors for the next couple of three five-year plans, but the reality turned out to be a little different. ..

Toolkit

• Test machine: Core I7, SSD, 12GB RAM (courtesy of my former employer)
• Load testing: k6.io (by the way, a very interesting framework in its architecture)

Application code organization

Sources

Here I decided not to bother too much and decided to follow the recommendations that I read in my time on Habré. In particular:

• Added payload as work to generate random data (random to avoid potential caching of results)

 class Human {
        constructor (id, name, surname, age, gender) {
            this.id = id
            this.name = name
            this.surname = surname
            this.age = age
            this.gender = gender
        }
    } 

• Both Dart and Node.js used synchronous and asynchronous request processing options.
• Applied native solutions and solutions on industry frameworks (aqueduct for dart and express for node. js)
• Since the study was able to get a significant speedup of Dart when using aqueduct, which runs isolates on each core, I used the cluster module for node.js for balancing purposes

Test methodology

• ran load tests with a given number of requests per second (500, 750) and a limit on the number of test iterations (number of completed requests)
• both the application and the testing framework were run on the same machine, so it should be understood that all results are relative and can only be compared with each other

Results

Native Dart

500 rps

750 rps

• http_reqs: 309.10154/s

Aqueduct framework for Dart

500 rps

750 rps

Node Express with Cluster

500 rps

750 rps

Conclusions

• Of course a lot depends on how you implemented the application logic, I’m not really sure that my code is optimal both in the case of dart and node.