[Various] AMD Ryzen 7600x/7700x/7900x/7950x Reviews

[Papusan]

From Techpowerup sum up. Overall this is great news for Zen 4 buyers and owners: Just stop worrying about the cooler, buy anything reasonably decent, it doesn't have to be an AIO, and you'll be fine. Because 95°C is the new 65°C. And if you're spending $200 on an AIO, then you're going from 95°C to 95°C, LOOL. Exactly same max temp as if you go with the cheaper sub 40$ air coolers.  Can't beat that... Or can you. At least you can save some money on your new desktop build if you go with cheap (or re-use old cooler) cooling. You can't do that with Intel. So the new 95°C target have it's advantage🙂

 

I can just imagine where Dell will cut costs for their new Ryzen Zen4 gaming desktop

AMD Ryzen 9 7950X Cooling Requirements & Thermal Throttling

[jaybee83]

8 hours ago, Papusan said:

From Techpowerup sum up. Overall this is great news for Zen 4 buyers and owners: Just stop worrying about the cooler, buy anything reasonably decent, it doesn't have to be an AIO, and you'll be fine. Because 95°C is the new 65°C. And if you're spending $200 on an AIO, then you're going from 95°C to 95°C, LOOL. Exactly same max temp as if you go with the cheaper sub 40$ air coolers.  Can't beat that... Or can you. At least you can save some money on your new desktop build if you go with cheap (or re-use old cooler) cooling. You can't do that with Intel. So the new 95°C target have it's advantage🙂

 

I can just imagine where Dell will cut costs for their new Ryzen Zen4 gaming desktop

AMD Ryzen 9 7950X Cooling Requirements & Thermal Throttling

 

and now lets include the most important factor here: frequency and thus performance 😉 95C is just the target to which the cpu will boost to. improve the cooling, temp target stays the same buuuut... clocks go higher! thats the whole idea bud 😄 at this point 95C is just a number, doesnt really tell u anything. gotta look at the resulting frequency. its not comparable to intel cpus where hitting TJMax would result in thermal throttling. the clocks just "stop climbing", they dont regress.

[Papusan]

3 hours ago, jaybee83 said:

 

and now lets include the most important factor here: frequency and thus performance 😉 95C is just the target to which the cpu will boost to. improve the cooling, temp target stays the same buuuut... clocks go higher! thats the whole idea bud 😄 at this point 95C is just a number, doesnt really tell u anything. gotta look at the resulting frequency. its not comparable to intel cpus where hitting TJMax would result in thermal throttling. the clocks just "stop climbing", they dont regress.

Yep. You gain wopping 69MHz clock speed jump for all cores going from an air cooler to the biggest AIO. A fantastic gain... Isn't it?🙂 But the biggest gain you get going with the huge AIO is lower noise. With the air cooler you need to run the fans at max speed to stop further clock drop in all cores load. And the cheapo AMD Wraith air coooler is a no go.

 

[Mr. Fox]

I think the correct way to look at it is 95°C is the threshold at which performance degrades. It can't really be viewed as a "target" unless the end goal is experiencing diminished performance. Time will tell what, if any, effect is has on the lifespan of the silicon. An ultimate outcome can only result from keeping temperatures as cool as possible. That much remains constant. I think accepting such ludicrous temperatures as the new normal is not very intelligent. It wasn't for laptop and should not be for desktop. Unless you just don't give a rat's tail about performance and get off on the idea of slow-cooking your parts.

[Papusan]

18 minutes ago, Mr. Fox said:

I think the correct way to look at it is 95°C is the threshold at which performance degrades. It can't really be viewed as a "target" unless the end goal is experiencing diminished performance. Time will tell what, if any, effect is has on the lifespan of the silicon. An ultimate outcome can only result from keeping temperatures as cool as possible. That much remains constant. I think accepting such ludicrous temperatures as the new normal is not very intelligent. It wasn't for laptop and should not be for desktop. Unless you just don't give a rat's tail about performance and get off on the idea of slow-cooking your parts.

Yep, node shrinks can have their diminishing returns. To be able to compete then you'll need to take out all the performance from it and hence you get 95C in return. And no AIO or custom water in the world can change that facts (if you don't delidde, undervolt and cap the temp threshold below AMDs temp target). Otherwise 95C it will be. Gamers shouldn't be so worried and won't see 95C and hence less the need for an AC running home due AMDs new Cpu temp target. 

 

As a side note... Have seen people talk about degraded Ryzen silicon due overclocking, higher voltage paired with high temps on later gen Threadripper. iUnlock was one of them seeing the silicon degraded. But this will apply both for Intel and AMDs chips.  

[Reciever]

I dont think people that can afford a 7950x are going to sweat buying a good cooler (pun intended) 

[Mr. Fox]

1 minute ago, Reciever said:

I dont think people that can afford a 7950x are going to sweat buying a good cooler (pun intended) 

Not if their priorities are straight. That sometimes proves to be a pretty big "if" for those that don't know any better.

[Papusan]

13 minutes ago, Reciever said:

I dont think people that can afford a 7950x are going to sweat buying a good cooler (pun intended) 

Probably not, but the slower and cheaper Zen 4 chips run equal hot and have same temp target. But most likely saved due they use their chips for gaming 🙂 Its for a reason many buy the cheaper Zen 4 chips. Cost savings. And this will apply for the cooling as well. And many don't like have water in their boxes. I just hope they will try avoid re-use the AMD Wright air cooler. Because that cooler isn't good enough for Zen 4 chips.

[Mr. Fox]

24 minutes ago, Papusan said:

Probably not, but the slower and cheaper Zen 4 chips run equal hot and have same temp target. But most likely saved due they use their chips for gaming 🙂 Its for a reason many buy the cheaper Zen 4 chips. Cost savings. And this will apply for the cooling as well. And many don't like have water in their boxes. I just hope they will try avoid re-use the AMD Wright air cooler. Because that cooler isn't good enough for Zen 4 chips.

The factory cooler (if and when one is included) should be a minimum of 240mm AIO as an entry point. People that want something better can buy it, or if they want to try to manage it with air because they have a predisposed objection to using liquid cooling, they can give it their best shot with something special. If OEMs are going to continue down the path of trying to sell the idea that 95°C is the new normal, they need to either step up their efforts with something better or not provide anything and leave it up to the end user how to deal with it. Providing something not up to the task would be really lame. I think it is messed up to call it the new normal. Maybe the new unmitigated exception. It shouldn't be "OK" or normal.

[Ashtrix]

The way AMD's new CPU works is very radical.

 

The fact that you can use any cooler the chip will always boost to 95C, like yeah it's fine on AIO or even midrange Air cooler or high performance Noctua, the fact that the cooler will run at high temps 24.7 that is a point of discussion which I want to hear, not about how TPU said the CPU is going to melt down and throttle, we already knew from AMD presentation/review/slides that Zen 4 will always hit 95C no matter what.

 

Also the point @Mr. Fox mentioned is very important observation that how the parts will live up when they are running this hot always, a.k.a slow cooking means we have the HSF areas always running at full speed or be it fans or the entire apparatus of Vapor chamber or Heat pipes doing their thing over at max.

 

A silicon is a silicon, unless there's some significant change in chemical composition how TSMC 5N silicon is  and it's Zen 4 lineup being 100% fine for 24/7 running at 95C always. I can recall one thing, remember Nvidia.. They said GDDR6X was rated at 100C+ and Max temp is 113C or something as per Micron datasheets so they said the VRAM on 3090 at back heating is all ok.

 

Well we know that how BAE Systems makes RAD products (Radiation Hardened), well they are geared for Space. High heat, High radiation, Extreme cold. Not all are certified for that class. If you see the space grade tech is not even near the latest lithography node. They are on 200nm to 150nm technology process. RAD5500 is their latest chip, it's on PowerPC 64-Bit uArch. Just food for thought..

[Reciever]

5 minutes ago, Mr. Fox said:

Not if their priorities are straight. That sometimes proves to be a pretty big "if" for those that don't know any better.

Always exceptions to the rule, though I would wager (or at least, have the expectation) that if they used the wraith cooler, it would be as an interim cooler until budget allows for the heatsink they desire.

 

That time duration shouldnt be concerning, but only if we dont see failures over the next 6 months or so.

1 minute ago, Papusan said:

Probably not, but the slower and cheaper Zen 4 chips run equal hot and have same temp target. But most likely saved due they use their chips for gaming 🙂

 

Yeah most titles dont saturate all cores, thats not to say that no games do either though. Prospective owners will need to be responsible enough to anticipate to some degree what their respective use cases are, which is a policy I generally prefer in all aspects of life. That of course only applies in the event that the information we have is true and accurate.

[Papusan]

2 hours ago, Mr. Fox said:

The factory cooler (if and when one is included) should be a minimum of 240mm AIO as an entry point. People that want something better can buy it, or if they want to try to manage it with air because they have a predisposed objection to using liquid cooling, they can give it their best shot with something special. If OEMs are going to continue down the path of trying to sell the idea that 95°C is the new normal, they need to either step up their efforts with something better or not provide anything and leave it up to the end user how to deal with it. Providing something not up to the task would be really lame. I think it is messed up to call it the new normal. Maybe the new unmitigated exception. It shouldn't be "OK" or normal.

Look from 1:50 where DELLs head thermal engineer Travis North in short say you don't need proper cooling. Run your HW at max temps is just fine. Why overspend money on better cooling if the HW run within specs? And if you bought an OEM pc, be prepared for proprietary solutions so forget spend more money on cooling that may not fits🙃

 

 

[jaybee83]

7 hours ago, Papusan said:

Yep. You gain wopping 69MHz clock speed jump for all cores going from an air cooler to the biggest AIO. A fantastic gain... Isn't it?🙂 But the biggest gain you get going with the huge AIO is lower noise. With the air cooler you need to run the fans at max speed to stop further clock drop in all cores load. And the cheapo AMD Wraith air coooler is a no go.

 

 

welp....ure comparing running Noctua fans balls to the wall vs. an AiO. no matter what cpu u compare at such baselines, ull likely get very similar results in terms of clocks gained or temps lowered. Id wager aside from the likes of brother @Mr. Fox the regular user would not use his cooling fans full tilt all the time.

That said, if you look at the whole cooling range, clock difference DO indeed become very significant, but all of them at the same target temp, namely 95C. 

As mentioned before, compare this with how a gpu behaves at stock. Strap an AiO on a 3090 Ti and it'll only clock max. 100 Mhz higher, if at all compared to a good air cooler. No difference here, CPUs, especially AMD ones, are becoming more and more like GPUs in their boosting behaviour. And just like GPUs have a temp target, so do CPUs now.

 

6 hours ago, Mr. Fox said:

I think the correct way to look at it is 95°C is the threshold at which performance degrades. It can't really be viewed as a "target" unless the end goal is experiencing diminished performance. Time will tell what, if any, effect is has on the lifespan of the silicon. An ultimate outcome can only result from keeping temperatures as cool as possible. That much remains constant. I think accepting such ludicrous temperatures as the new normal is not very intelligent. It wasn't for laptop and should not be for desktop. Unless you just don't give a rat's tail about performance and get off on the idea of slow-cooking your parts.

 

I have to disagree here. A threshold at which performance degrades would be more like how Intel CPUs behave. They have a set max. multicore boost, i.e. 5.2 Ghz all core, which holds steady until TjMax / thermal throttling point is reached, upon which they start to downclock. Zen 4 does not behave like that, it rather "looks" at the given thermal headroom / cooling implemented and boosts as high as it can go until reaching 95C, at which point it plateaus. Thus, the temperature is now less meaningful than the actually reached clocks.

A good indicator for this difference in behaviour is the fact that ALL four Zen 4 CPUs show this behaviour and shoot up to exactly 95C no matter what kind of cooling is applied. If it was just a fact of the cpus "running hot", then surely a 6 core 7600X would show way lower peak temps and a 7950X, no? But instead, they all target the same set 95C and just differ in the respective clocks.

 

Granted, as GN recently detailed in a video, there actually IS also a power cap in addition to the target temp, so throwing LN2 on the chip does not mean itll still boost all the way up to 95C, that would entail reaching 1000W+ lulz 😄

Point here being, that the power cap is way way higher than ambient cooling could conceivable provide thermal headroom for to keep below 95C.

 

6 hours ago, Papusan said:

Probably not, but the slower and cheaper Zen 4 chips run equal hot and have same temp target. But most likely saved due they use their chips for gaming 🙂 Its for a reason many buy the cheaper Zen 4 chips. Cost savings. And this will apply for the cooling as well. And many don't like have water in their boxes. I just hope they will try avoid re-use the AMD Wright air cooler. Because that cooler isn't good enough for Zen 4 chips.

 

Thanks for bringing up this point, as I already mentioned above in reply to brother Fox! One thing to consider here is also that this 95C temp target is only for heavy all core loads. Gaming and regular desktop usage will not push the cpus far enough for them to be pegged at 95C 24/7.

So as opposed to Intel chips, where crappy vendors such as Dell or Apple just let them hit TJMax at 100C and throttle to kingdom come, this 95C set point did not come from a vendor or from insufficient cooling, but rather directly from AMD to instruct Zen 4 cpus to boost as long as 95C is not reached and the much higher wattage cap is not breached either.

 

5 hours ago, Ashtrix said:

The way AMD's new CPU works is very radical.

 

The fact that you can use any cooler the chip will always boost to 95C, like yeah it's fine on AIO or even midrange Air cooler or high performance Noctua, the fact that the cooler will run at high temps 24.7 that is a point of discussion which I want to hear, not about how TPU said the CPU is going to melt down and throttle, we already knew from AMD presentation/review/slides that Zen 4 will always hit 95C no matter what.

 

Also the point @Mr. Fox mentioned is very important observation that how the parts will live up when they are running this hot always, a.k.a slow cooking means we have the HSF areas always running at full speed or be it fans or the entire apparatus of Vapor chamber or Heat pipes doing their thing over at max.

 

A silicon is a silicon, unless there's some significant change in chemical composition how TSMC 5N silicon is  and it's Zen 4 lineup being 100% fine for 24/7 running at 95C always. I can recall one thing, remember Nvidia.. They said GDDR6X was rated at 100C+ and Max temp is 113C or something as per Micron datasheets so they said the VRAM on 3090 at back heating is all ok.

 

Well we know that how BAE Systems makes RAD products (Radiation Hardened), well they are geared for Space. High heat, High radiation, Extreme cold. Not all are certified for that class. If you see the space grade tech is not even near the latest lithography node. They are on 200nm to 150nm technology process. RAD5500 is their latest chip, it's on PowerPC 64-Bit uArch. Just food for thought..

 

Exactly! I think its pretty safe to assume that not thermal throttling or melt-downs will happen if AMD specifically set this target temp for clock boosting at 95C. 

 

Again, to clarify here, since a LOT of people are so used to Intel cpu boosting behaviour and seem to draw the wrong conslusions:

- the 95C temp target is not comparable to the Intel cpu TjMax. the former is a set point, until which clocks are boosted, the latter is a threshhold, upon reaching starts to decrease clocks due to thermal throttling. the end results are vastly different from each other.

- Since all Zen 4 cpus have that thermal set point at 95C, its not just about "HAHA AMD cpu runs hot and goes brrrrrr" it just rather means than in order to judge a cooling implementation, clocks are now more important than temps.

- the 95C set point is not set in stone. If you dont like it, just change it in bios! "problem" solved 😄 

[Papusan]

9 minutes ago, jaybee83 said:

welp....ure comparing running Noctua fans balls to the wall vs. an AiO. no matter what cpu u compare at such baselines, ull likely get very similar results in terms of clocks gained or temps lowered. Id wager aside from the likes of brother @Mr. Fox the regular user would not use his cooling fans full tilt all the time.

You'll only lose 24MHz if you drop the fan speed from 100% down to 60%. And at that point I think the average Joe dare to push their fans 🙂 And you'll still get above 5100MHz if you drop it further down to 50%. 

[jaybee83]

4 minutes ago, Papusan said:

You'll only lose 24MHz if you drop the fan speed from 100% down to 60%. And at that point I think the average Joe dare to push their fans 🙂 And you'll still get above 5100MHz if you drop it further down to 50%. 

i doubt average joe would care about 24 Mhz in exchange for the added noise 😄 

[Reciever]

54 minutes ago, jaybee83 said:

i doubt average joe would care about 24 Mhz in exchange for the added noise 😄 

Wait, does that mean I am no longer average? 

 

Damn. 4 window fans, standing fan, 10 fans for my uhh "server". I have a box of about dozen+ fans waiting for me to stop being lazy. 

 

Yeah, I guess I need to toss in my average membership card 

[jaybee83]

5 minutes ago, Reciever said:

Wait, does that mean I am no longer average? 

 

Damn. 4 window fans, standing fan, 10 fans for my uhh "server". I have a box of about dozen+ fans waiting for me to stop being lazy. 

 

Yeah, I guess I need to toss in my average membership card 

im surprised they didnt kick you out of the average joe club already 😄 

[Reciever]

1 minute ago, jaybee83 said:

im surprised they didnt kick you out of the average joe club already 😄 

It gets tricky when you pay for the year up front, because then I would issue a charge back on the card 

[Ashtrix]

On 10/5/2022 at 2:36 PM, Papusan said:

Look from 1:50 where DELLs head thermal engineer Travis North in short say you don't need proper cooling. Run your HW at max temps is just fine. Why overspend money on better cooling if the HW run within specs? And if you bought an OEM pc, be prepared for proprietary solutions so forget spend more money on cooling that may not fits🙃

 

 

 

ULTIMATE BULLS**T LOOOOL

 

Alienware fools are quoting TjMax Intel Ark specification of 100C for Alder Lake and using that as a base and saying here we are running into the temperate by using some fancy garbage formula and pushing the machine to max ofc they do not mention the absolute trash cooling. Also that guy says, we are making sure that "we do not waste cooling" WTF... Like since when more cooling is bad.

 

Honestly their trashbox computers are the prime example of human stupidity and gullible nature of the normies who do not do basic research. Yeah I get it many do not have time BUT we are talking about Alienware tag products which are expensive to begin with. so you have to do some damn research.

 

Horrible trash. Now with AMD Ryzen 7000 they will say the CPU is operating at 95C but will not mention the clocks LOL. It will be probably worse than Wraith in a few days thanks to the propreitary garbage mobo and dumpsterfire logic they use.

[Papusan]

1 hour ago, Ashtrix said:

 

ULTIMATE BULLS**T LOOOOL

 

Alienware fools are quoting TjMax Intel Ark specification of 100C for Alder Lake and using that as a base and saying here we are running into the temperate by using some fancy garbage formula and pushing the machine to max ofc they do not mention the absolute trash cooling. Also that guy says, we are making sure that "we do not waste cooling" WTF... Like since when more cooling is bad.

 

Honestly their trashbox computers are the prime example of human stupidity and gullible nature of the normies who do not do basic research. Yeah I get it many do not have time BUT we are talking about Alienware tag products which are expensive to begin with. so you have to do some damn research.

 

Horrible trash. Now with AMD Ryzen 7000 they will say the CPU is operating at 95C but will not mention the clocks LOL. It will be probably worse than Wraith in a few days thanks to the propreitary garbage mobo and dumpsterfire logic they use.

I just love your honnst comments. But be you sure.... Dell will milk the 95C normal for AMD ecual as they say 100C is ok for Intel. This way they can cut costs as usual. Aka nothing have changes as long Dell rely on Travis North.

[jaybee83]

2 hours ago, Ashtrix said:

Horrible trash. Now with AMD Ryzen 7000 they will say the CPU is operating at 95C but will not mention the clocks LOL. It will be probably worse than Wraith in a few days thanks to the propreitary garbage mobo and dumpsterfire logic they use.

 

omg im worried ull very likely be right.... might as well run well below base clock and still be inside "Dellienware spec" since it runs at 95C, riiiiiiite? geez 😅

[1610ftw]

On 10/3/2022 at 8:46 AM, Papusan said:

If they run colder be you sure the OEM will either shrink the chassis or save money on the cooling. I saw for a few days ago a video with Travis North (Dell’s thermal engineer) and Intel’s engineering team, said over capacity coolers is wasted money if you can run HW at specs with less powerful cooling. He said etc processors is meant to run at tj-max. 24/7. Because of TAU the boost won’t last forever. Boost clock is for burst loads. Long term load is base clock speed (lower TDP).

 

The lack of ambition even in top of the line products is rather sickening.

[1610ftw]

15 hours ago, Ashtrix said:

 

ULTIMATE BULLS**T LOOOOL

 

Alienware fools are quoting TjMax Intel Ark specification of 100C for Alder Lake and using that as a base and saying here we are running into the temperate by using some fancy garbage formula and pushing the machine to max ofc they do not mention the absolute trash cooling. Also that guy says, we are making sure that "we do not waste cooling" WTF... Like since when more cooling is bad. 

 

More cooling is indeed very bad as your laptop may get a little bit heavier and most of all a bit more expensive to produce.

 

Never mind that it would also:

 

- run cooler

- produce less noise

- last longer

- clock higher

 

And we all know that nobody would want that...

 

[Mr. Fox]

So, temperatures do matter and taking active steps to avoid the 95°C "target" is something worth pursuing to achieve the best performance even without consideration for sub-zero extreme cooling. It is going to be beneficial to improve cooling even for daily driver performance.

 

The silicon shroud that Roman made is something very useful that every Ryzen 7XXX user should think about. I think it is a great idea and will probably solve some issues the notched and unsealed IHS and multitude of exposed SMDs can present that may not have been contemplated. An absolute must have for liquid metal, but would be good to have to minimize tedious mess cleanup with ordinary thermal compound. The nifty stuff this guy can create on the fly is always impressive. I wish I had all of the the tools and fabrication resource he has at his disposal.

 

This video makes me want to buy a DICE pot. 

 

 

 

[jaybee83]

6 hours ago, 1610ftw said:

 

More cooling is indeed very bad as your laptop may get a little bit heavier and most of all a bit more expensive to produce.

 

Never mind that it would also:

 

- run cooler

- produce less noise

- last longer

- clock higher

 

And we all now that nobody would want that...

 

 

u said it brother! form over function FTW! glad were ALL on the same page for this one /s 🤣

 

5 hours ago, Mr. Fox said:

So, temperatures do matter and taking active steps to avoid the 95°C "target" is something worth pursuing to achieve the best performance even without consideration for sub-zero extreme cooling. It is going to be beneficial to improve cooling even for daily driver performance.

 

The silicon shroud that Roman made is something very useful that every Ryzen 7XXX user should think about. I think it is a great idea and will probably solve some issues the notched and unsealed IHS and multitude of exposed SMDs can present that may not have been contemplated. An absolute must have for liquid metal, but would be good to have to minimize tedious mess cleanup with ordinary thermal compound. The nifty stuff this guy can create on the fly is always impressive. I wish I had all of the the tools and fabrication resource he has at his disposal.

 

This video makes me want to buy a DICE pot. 

 

 

 

 

German Engineered Perfection 😄