AMD Ryzen 7 5800X

[W1zzard]

Right, but how do you know the price wouldn't have been higher by 20 bucks had they included the cooler?

Fair point, maybe I'm just looking at it from the wrong side

 

[Charcharo]

Lol where I live we sell the Wraith Prisms for 25-30 EUR. This is like 2-3 newish games on a discount or many old games.

Its added value. People on lower-end chips buy them up since the Wraith series coolers are downdraft and help cool the motherboard's VRMs very well too. Obviously, even better such coolers exist but I am personally not happy that the Prism is no longer included in the CPUs.

 

[shmuck]

Lol where I live we sell the Wraith Prisms for 25-30 EUR. This is like 2-3 newish games on a discount or many old games.

Its added value. People on lower-end chips buy them up since the Wraith series coolers are downdraft and help cool the motherboard's VRMs very well too. Obviously, even better such coolers exist but I am personally not happy that the Prism is no longer included in the CPUs.

I understand that for some people it might be well worth it, but I can't be arsed to spend the time to package a cooler and go to the post office for a one-time gain for 25 euros.

>Its added value
No, we don't know that, not if they increased the price of 3700X because of it. If you buy a pre-built PC for 1k, you aren't buying a case for 1k with the added value of extra components .

>I am personally not happy that the Prism is no longer included in the CPUs
Buy it separately, no biggie. Now you can't counter with not wanting to spend extra money, because then what you're not happy about in actuality is the price of the CPU itself (if it was cheap in the first place, this wouldn't matter).

 

[Charcharo]

I understand that for some people it might be well worth it, but I can't be arsed to spend the time to package a cooler and go to the post office for a one-time gain for 25 euros.

>Its added value
No, we don't know that, not if they increased the price of 3700X because of it. If you buy a pre-built PC for 1k, you aren't buying a case for 1k with the added value of extra components .

>I am personally not happy that the Prism is no longer included in the CPUs
Buy it separately, no biggie. Now you can't counter with not wanting to spend extra money, because then what you're not happy about in actuality is the price of the CPU itself (if it was cheap in the first place, this wouldn't matter).

It is different for me since my second "job" (hobby with benefits) in a sense is reselling stuff and fixing small devices / goods. I package things every single day anyway so the Prism's 1-2 minute packaging timeis very quick for me. But I did give my last one for my girlfriend's 3600-based PC. Its good on a Ryzen 3600.

For me it is added value for sure.

As for the 3700X - you know what is interesting? The 3700X is basically the 3700 non-X, the successor to the 2700 and 1700. AMD bumped its price by 30 USD and gave it the Prism just to increase their margin without gamers protesting. The 3800X is actually the 2700X successor, weirdly enough.

 

[Hossein Almet]

Something is wrong with TechPowerUP, every other websites I come across like Tom's Hardware, Guru3D, Kitguru, the 5900X and the 5950X occupy top spot in games benchmarks, here they are 'nowhere to be seen'.

 

[Jism]

@Wizzard,

Anything on the max 24/7 voltage for these CPU's? 1.4 is'nt really recommended as degradation is real on these things.

 

[bug]

@Wizzard,

Anything on the max 24/7 voltage for these CPU's? 1.4 is'nt really recommended as degradation is real on these things.

What do you expect when these are only 2 days on the market?

 

[Jism]

What do you expect when these are only 2 days on the market?

Well i suspect just as the 2x and 3x00 series its proberly around 1.34 ~ 1.38V all-core 24/7. We've seen various CPU's degrade once running on 24/7 1.4v and above. No joke.

So it should always be said, in any review that OC'ing beyond the 1.4V is not a bright thing todo if you want to keep your CPU alive.

 

[bug]

Well i suspect just as the 2x and 3x00 series its proberly around 1.34 ~ 1.38V all-core 24/7. We've seen various CPU's degrade once running on 24/7 1.4v and above. No joke.

So it should always be said, in any review that OC'ing beyond the 1.4V is not a bright thing todo if you want to keep your CPU alive.

Ok, but how do you determine whether it's safe or not after just two days? Or a week, or however long @W1zzard had these?
What you're asking for needs both time and a sizable sample.

 

[jayjr1105]

Something is wrong with TechPowerUP, every other websites I come across like Tom's Hardware, Guru3D, Kitguru, the 5900X and the 5950X occupy top spot in games benchmarks, here they are 'nowhere to be seen'.

He only used 3200MHz RAM. Check out GN review. They crown the 5600X best overall gaming CPU period, matches or beats the 10900K is just about everything and uses less than half the power, and costs $250 less.
https://www.reddit.com/r/Amd/comments/jp2uab
Also, keep an eye out for 1usmus' CTR tool. Currently it only works with Zen2 CPU's. When it's ready for Zen3, things are going to get interesting.

 

[Makaveli]

The LTT review also uses DDR4 3600 CL 14 memory and shows the 5600X and up also leading.

 

[jayjr1105]

See Robert's point #2. If 4000 RAM can run 1:1 with infinity fabric, the gaming lead is going to skyrocket

https://twitter.com/i/web/status/1324561815104356352

 

[Makaveli]

See Robert's point #2. If 4000 RAM can run 1:1 with infinity fabric, the gaming lead is going to skyrocket

https://twitter.com/i/web/status/1324561815104356352

Nice find.

i've seen a report that the current AGESA based on 1.1.0.0 is having some difficulty hitting IF at 2000 and a upcoming version will fix that.

 

[jayjr1105]

Nice find.

i've seen a report that the current AGESA based on 1.1.0.0 is having some difficulty hitting IF at 2000 and a upcoming version will fix that.

Even hitting somewhere between 1800 and 2000 consistently will be a win. I doubt everyone's will hit 2000 stable but we'll see.

 

[bug]

Even hitting somewhere between 1800 and 2000 consistently will be a win. I doubt everyone's will hit 2000 stable but we'll see.

How cheap/expensive is good DDR4-4000?
(And yes, I totally expected this, I asked about it above.)

 

[Makaveli]

Even hitting somewhere between 1800 and 2000 consistently will be a win. I doubt everyone's will hit 2000 stable but we'll see.

Agreed it will be the same as 3800 on Zen 2.

You will need the combination of good cpu, motherboard and ram with abit of luck mixed in.

 

[Fasola]

He only used 3200MHz RAM. Check out GN review. They crown the 5600X best overall gaming CPU period, matches or beats the 10900K is just about everything and uses less than half the power, and costs $250 less.
https://www.reddit.com/r/Amd/comments/jp2uab
Also, keep an eye out for 1usmus' CTR tool. Currently it only works with Zen2 CPU's. When it's ready for Zen3, things are going to get interesting.

GN also used 3200 CL14 memory although it was 4 DIMMs. I'm really curious what W1zzard's investigation will turn up.

 

[jayjr1105]

GN also used 3200 CL14 memory although it was 4 DIMMs. I'm really curious what W1zzard's investigation will turn up.

View attachment 174686

RDR2 was the exception to all the AMD beatdowns on the GN 5600X review.
Nvm, Yeah wth

 

[Ripcord]

ill stick with my 1600

 

[tfdsaf]

Every review I saw elsewhere had the 5000 series beating Intel counterparts by a big margin in both games and productivity applications!

Well games was a tie essentially in many games, but Zen 3 won significantly in several games like SOTTR, HZD, Death Stranding, Hitman 2, etc...

 

[Mtom]

Something is very wrong with these tests.

All other reviews show 15-35% uplift compared to Zen2 CPUs in 1080p and your review shows 1-2%

 

[gravel]

This is test are wrong, techpowerup loses all credibility for me . !!!

 

[Whitestar]

This is test are wrong, techpowerup loses all credibility for me . !!!

The tester says he's looking into it. You're not willing to give him the benefit of doubt?

 

[Mussels]

its w1zz, if he finds a flaw he'll fix it.

The difference will be in one of three places:

1. the tests themselves (in game benches vs how w1zz does it)
2. the math he uses to average out the differences (does he ignore outliers? does he use more 'fair' titles?)
3. the sheer number of tested games (the bigger the number, the more average the results)

The reason for the low RAM speed vs other reviewers is consistency so it can be compared to previous systems - wizz may need to go to some extra effort for high speed DDR4 testing on ryzen, as well as timing testing that he mentioned doing somewhere.

 

[bug]

This is test are wrong, techpowerup loses all credibility for me . !!!

The whole world isn't quite right, does it also lose credibility in your eyes?