News AMD Ryzen 7 5700X Review: A Price Cut Disguised as a New Chip

Sleepy_Hollowed

Distinguished
Jan 1, 2017
501
195
19,070
Nice review.

Definitively too little too late, I wonder what was the reason for them taking so long, besides waiting for supply issues.

I'd still snatch one of these for light workstations if they're available, or a 5700G if I needed integrated graphics.
 
Oct 29, 2021
2
0
10
Great review, thanks!

Agreed that it's too late to make a meaningful impact, but for latecomers such as me it's perfect. I've got an AM4 system running a Ryzen 2700 and this would extend the lifespan of it by a couple of years.

If anyone has made jump like that, I'd be keen on knowing what's the improvement generally, performance, temps and such.
 

Soaptrail

Distinguished
Jan 12, 2015
301
95
19,420
Clarification question: Are you running the Intel CPUs with the stock coolers?

Most tests I've seen use water cooling setups for the Intel chips, while most users use the bundled inadequate air coolers. Has this been addressed? Thanks.

I know Tom's and others have experimented with different coolers on one specific CPU but it would be nice to see one review show a couple of CPU's one from Intel and one from AMD and how much 2 or 3 CPU coolers affect performance. I have no intention of going AIO so a lot of reviews are already overblowing the results I would ever achieve.
 
  • Like
Reactions: martinch

escksu

Reputable
BANNED
Aug 8, 2019
878
354
5,260
Yes, its definitely very late..... The same goes for 5600 (non-X version)....

I originally wanted to upgrade by 3600 but find 5600x too expensive. Although prices have come down, its now no longer worth upgrading. I rather get 12600K or wait for next gen (raptor lake and Zen4 coming soon). Also a good time to wait for DDR5 to mature and get cheaper.

But, I feel its not the main problem, its late but still ok. The real problem ones are the G series..... Absolute trash. Doesn't even have PCIE 4.0 and only 1/2 the cache.
 

KyaraM

Admirable
Nice review.

Definitively too little too late, I wonder what was the reason for them taking so long, besides waiting for supply issues.

I'd still snatch one of these for light workstations if they're available, or a 5700G if I needed integrated graphics.
They made more money with the lot more expensive 5800X, that's the reason.
 

hannibal

Distinguished
Nice review.

Definitively too little too late, I wonder what was the reason for them taking so long, besides waiting for supply issues.

I'd still snatch one of these for light workstations if they're available, or a 5700G if I needed integrated graphics.

Not enough production capacity to fulfill demand. When everythig did go out, it was better to concentrate to higher margin products. And now it seems that Intel and Apple has bough so much capacity from TSMC that AMD next gen products gets delayed, because they don´t have capacity to produce them...
 
We put AMD's Ryzen 7 5700X through the wringer in our expanse set of benchmarks.

AMD Ryzen 7 5700X Review: A Price Cut Disguised as a New Chip : Read more
Are you ever going to access the fact that Intel ships CPUs with higher TDPs than their bundled coolers can dissipate?

Articles keep saying, "Intel has the performance crown", then show a test of an Intel system with a $120 cooler that most users won't have that beats AMD chips by only 5%. Then in real life, the Intel chips throttle and lose by 5%. I've seen benchmarks where Intels throttle and lose 30%+ of their performance, supposedly in realistic environments.

If I'm wrong, at least do a test that investigates this. That's all I want to see.

Because it looks like Intel is gaming the system. They can put out chips with 250W TDP, knowing that reviewers use open benches and water coolers. Then AMD puts out chips that perform the same inside a user's system and they get penalized by all the media outlets.
 

KyaraM

Admirable
Are you ever going to access the fact that Intel ships CPUs with higher TDPs than their bundled coolers can dissipate?

Articles keep saying, "Intel has the performance crown", then show a test of an Intel system with a $120 cooler that most users won't have that beats AMD chips by only 5%. Then in real life, the Intel chips throttle and lose by 5%. I've seen benchmarks where Intels throttle and lose 30%+ of their performance, supposedly in realistic environments.

If I'm wrong, at least do a test that investigates this. That's all I want to see.

Because it looks like Intel is gaming the system. They can put out chips with 250W TDP, knowing that reviewers use open benches and water coolers. Then AMD puts out chips that perform the same inside a user's system and they get penalized by all the media outlets.
You don't need that 120 bucks cooler, though. Most Intels can be comfortably cooled by a cooler you get for less than 40. For gaming alone, that goes even for the 12700K (same for the high-end AMDs, btw, so your point is utter bs), which doesn't even come with a stock cooler, and the new Intel cooler keep the smaller Alder Lakes in check really well for being a free cooler. Hell, a good air cooler also costs less than 100 here and does the job very well. And it's not like the situation is really different for AMD... I wouldn't use either stock cooler. They are all garbage.
 
I think the issue of cooling is probably a little more nuanced. Intel's approach seems to be predicated upon the CPUs being used in a "bursty" workload (e.g. editing photos one after the other has frequent idle periods for user input), which seems to be a reasonable assumption for the majority of office/home office users. However, if you use a motherboard which allows "infinite turbo" (or sets it by default), the 65W bundled cooler is inadequate for long-running tasks - from personal experience, an i5-11400 will hit 90C quite quickly in this case.

On the other hand, AMD's approach seems to be more based upon something closer to peak power draw, so doesn't (from memory) suffer from this issue as much. Having said that, the Ryzen 7's don't come with a cooler, and most after-market coolers (even the cheaper ones) should provide better cooling performance and acoustics than the AMD and Intel stock coolers, so are probably very much worth considering (in my opinion :)).
 
  • Like
Reactions: KyaraM

KyaraM

Admirable
I think the issue of cooling is probably a little more nuanced. Intel's approach seems to be predicated upon the CPUs being used in a "bursty" workload (e.g. editing photos one after the other has frequent idle periods for user input), which seems to be a reasonable assumption for the majority of office/home office users. However, if you use a motherboard which allows "infinite turbo" (or sets it by default), the 65W bundled cooler is inadequate for long-running tasks - from personal experience, an i5-11400 will hit 90C quite quickly in this case.

On the other hand, AMD's approach seems to be more based upon something closer to peak power draw, so doesn't (from memory) suffer from this issue as much. Having said that, the Ryzen 7's don't come with a cooler, and most after-market coolers (even the cheaper ones) should provide better cooling performance and acoustics than the AMD and Intel stock coolers, so are probably very much worth considering (in my opinion :)).
I feel the important thing to take away is that proprietary coolers are better than stock for both, yes. You can get something like a BeQuiet! Pure Rock 2 for about 30-40 bucks. That's pretty reasonable and there are other, cheaper options available as well that will keep those CPUs in check, too. That's why I feel naming the cooler as a reason for performance differences in end-user systems is highly misleading. It takes two seconds to Google it. The issue existed with AMD processors as well. Claiming that using an overpowered cooler changes anything when a cheap, affordable cooler is enough already is what actually misleads people. Especially when considering that the test here even shows performance with stock cooler for the 12400...
 
  • Like
Reactions: martinch
the 65W bundled cooler is inadequate for long-running tasks - from personal experience, an i5-11400 will hit 90C quite quickly in this case.
Turbo is, by design capable and supposed to, push the system up to 100 degrees, if you get lower temps than that you are either not using all of the performance or are cooling "too hard" / loosing value by spending more than you have to on cooling.
If it helps your psyche to have a cooler CPU then sure, no problem, but it's not an issue for the CPU.

https://www.intel.com/content/www/u...intel-technologies-boost-cpu-performance.html
Intel® Adaptive Boost Technology kicks in when the CPU is below the ICCMax limit (maximum current) and a temperature limit of 100°C. That means it can remain active even at temperatures above 70°C, the threshold for Thermal Velocity Boost.
 
  • Like
Reactions: KyaraM
Turbo is, by design capable and supposed to, push the system up to 100 degrees...If it helps your psyche to have a cooler CPU then sure, no problem, but it's not an issue for the CPU.
Yes, I was aware how turbo boost works - my old work laptop (i7-4600U) didn't have great cooling and so spent a lot of time running at 90C+, which resulted in performance issues due to the CPU running ~500MHz below maximum turbo whilst at 100% load due to the lack of thermal headroom.

In the case of the 11400 I mentioned, when using the stock cooler with the power limits switched off (the default state for the motherboard used), the ~150W of heat being generated by the CPU when under load could not be adequately dissipated by the cooler, causing the temperature to increase, causing the CPU to dial down the frequencies. However, this shouldn't be an issue for the majority of users, because they won't be running the CPU at sufficient load levels for long enough for it to be an issue. Apologies if the inference I was making wasn't obvious.
 
In the case of the 11400 I mentioned, when using the stock cooler with the power limits switched off (the default state for the motherboard used)
Ah yes but see, that's not what intel is advertising, that's not what the stock cooler is meant for.
Of course going above the rated CPU/cooler TDP will cause issues.
And honestly getting "only" to about 90 with basically OC settings and the stock cooler isn't even bad.
 
  • Like
Reactions: KyaraM
that's not what intel is advertising, that's not what the stock cooler is meant for...Of course going above the rated CPU/cooler TDP will cause issues.
I'd agree, were it not for the fact that this is default behaviour for many basic motherboards. If Intel didn't intend that this be the case, then the spec should preclude defaulting to this behaviour (i.e. require the user to say, "yes, please").
 
I'd agree, were it not for the fact that this is default behaviour for many basic motherboards. If Intel didn't intend that this be the case, then the spec should preclude defaulting to this behaviour (i.e. require the user to say, "yes, please").
But a CPU needs the mobo to boot up, how is intel possibly going to be able to circumvent the mobo settings until you agree when the mobo starts first and then boots up the CPU?
The only alternative would be for intel to forbid it outright which would kick off a huge crap storm.

Yes mobo makers should be forced to make it very clear what kind of settings they use but I don't think that it's intel's responsibility, it should be handled by the gov like any other consumer right, like the 14 day cool-off period, that's not left to each manufacturer separately either, it's a common law.
 
  • Like
Reactions: KyaraM

KyaraM

Admirable
I'd agree, were it not for the fact that this is default behaviour for many basic motherboards. If Intel didn't intend that this be the case, then the spec should preclude defaulting to this behaviour (i.e. require the user to say, "yes, please").
I mean. If a vendor keeps to Intel's specs or does provide more to have a leg up over the competition, be it power draw defaults or connectivity options (as long as they at least provide the minimum specs from Intel), is up to them.

Edit: however, when I resetted my mainboard back to optimized defaults because for some reason XMP didn't work anymore after installing Windows, my maindoard actually did ask me what cooling solutionI have, or claim to have. Had I said stock air, it would have set my power limits to Intel defaults. So it really is upto the vendor. Hardware Unboxed also did some VRM testing and iirc found that some mainboards do keep to the limits or only run without with a K-series SKU.
 
Last edited:

TJ Hooker

Titan
Ambassador
But a CPU needs the mobo to boot up, how is intel possibly going to be able to circumvent the mobo settings until you agree when the mobo starts first and then boots up the CPU?
The only alternative would be for intel to forbid it outright which would kick off a huge crap storm.

Yes mobo makers should be forced to make it very clear what kind of settings they use but I don't think that it's intel's responsibility, it should be handled by the gov like any other consumer right, like the 14 day cool-off period, that's not left to each manufacturer separately either, it's a common law.
I mean, Intel doesn't seem to have an issue telling mobo makers what to do when it comes to things like preventing non-K overclocking, alder lake AVX512, etc. I'm sure if they cared, they would enforce it. But it mostly benefits them (by improving performance results), so they don't.
 
I mean, Intel doesn't seem to have an issue telling mobo makers what to do when it comes to things like preventing non-K overclocking, alder lake AVX512, etc. I'm sure if they cared, they would enforce it. But it mostly benefits them (by improving performance results), so they don't.
Oh I see, non-k and avx wouldn't improve performance for them...

Also how is getting a pile of reviews with "omg 300W+ " good for them?!
Intel enforcing PL would show their CPUs in a much better light on power usage and overclocking numbers for them would still show the same performance as they do now.
 
  • Like
Reactions: KyaraM