7900X reviewed!

[GER_T4IGA]

7 minutes ago, PCGuy_5960 said:

But that 4.6-4.7Ghz is with ONLY 1.25V, if you delid it and use 1.35V, 4.9-5GHz should be possible

.1V for 300MHz? I'd probably go as far as to say you should be able to go beyond five if you can still keep it cool at that voltage. Still only speculation.

[Brooksie359]

2 minutes ago, PCGuy_5960 said:

Nope, Naples/EPYC is just 4x 1800Xs

I mean if it's priced right and the server user doesn't need to use avx then it will sell. It just won't be a viable option for those who need to use avx.

[PCGuy_5960]

Just now, GER_T4IGA said:

.1V for 300MHz? I'd probably go as far as to say you should be able to go beyond five if you can still keep it cool at that voltage. Still only speculation.

Yeah, they will probably OC like Kabylake, so 4.9-5.1GHz at 1.35V

[cj09beira]

1 minute ago, GER_T4IGA said:

.1V for 300MHz? I'd probably go as far as to say you should be able to go beyond five if you can still keep it cool at that voltage. Still only speculation.

you probably cant go much further than 4.8-4.9 without deliding it

[PCGuy_5960]

Just now, cj09beira said:

you probably cant go much further than 4.8-4.9 without deliding it

Yep, they said that if they used 1.3V the CPU would hit 100C, so delidding is a must for higher overclocks

[done12many2]

 

 

I doubt this guys voltage was reporting right in CPUz as there's no way he could cool 10 cores at 1.5v using an AIO.    Either way, another 10 core running high clocks.  

[cj09beira]

Just now, PCGuy_5960 said:

Yep, they said that if they used 1.3V the CPU would hit 100C, so delidding is a must for higher overclocks

thats a bad thing bro, then the cpu can't hit 5ghz, deliding a 1000 bucks cpu is risky

[PCGuy_5960]

Just now, cj09beira said:

thats a bad thing bro, then the cpu can't hit 5ghz, deliding a 1000 bucks cpu is risky

I would be more than happy with 4.7GHz at 1.25V TBH, so not a very bad thing

[cj09beira]

Just now, done12many2 said:

 

 

I doubt this guys voltage was reporting right in CPUz as there's no way he could cool 10 cores at 1.5v using an AIO.    Either way, another 10 core running high clocks.  

you can run high voltages if the benchmark doesn't produce too much heat

[cj09beira]

Just now, PCGuy_5960 said:

I would be more than happy with 4.7GHz at 1.25V TBH, so not a very bad thing

sure

[PCGuy_5960]

1 minute ago, done12many2 said:

I doubt this guys voltage was reporting right in CPUz as there's no way he could cool 10 cores at 1.5v using an AIO.    Either way, another 10 core running high clocks.  

Yep, so der8auer was 100% right when he said that Skylake-X will be an overclocking beast

[Lord Nicoll]

Well the performance isn't surprising, It's what Kabylake should have been (slightly better IPC as well as clock boost) but the platform is still an utter mess, I don't think Intel really thought about X299 is or was going to be, so they threw a hell of a lot of extraneous crap onto it in a blind attempt to counter what ever Thread Ripper will be, and it really hurt the platform, if Intel had just stepped back and focused and made the best platform they could from the beginning AMD's rise would have been stopped before it started. I'ma be sitting back to see how this one plays out, should have a few good keks. 

[mynameisjuan]

I am glad to see the shit storm behind not soldering the heat sink really followed through to everyone complaining. Oh hey would you look at that, its not that bad.

[knightslugger]

10 minutes ago, done12many2 said:

 

Nice vehicle man.  I leased a 2000 Grand Cherokee many many years ago and only changed the oil in it one time during the 2-year lease right before I turned it in.  lol.  That sucker ran perfectly on 2-year old (40k mile) oil.  

bought into the euro-diesel limited. just tack on another $4500 to that loan! still though, wow what a "truck". Find new stuff in it every week! found a button on the paddle shifter that does some weirdness with the radio. a button on a button. "...so we heard you like buttons..." --FCA. makes about as much sense as the non-sense of this x299 platform these days. though, if this is the sort of performance we can reasonably expect... I mean i like my R7. but my pants got tight at that CB15 score. esp. when OCd. what a beast. OH BTW, i think the 1700x is borked. wont stay stable at stock speed anymore. and my aura RGB stopped working some time ago. sent the whole lot in for RMA so i'll hopefully have a new working system to off load. would put a sizable dent in a x299 system with this little guy in it.

[PCGuy_5960]

2 minutes ago, cj09beira said:

you can run high voltages if the benchmark doesn't produce too much heat

CInebench produces quite a lot of heat.....

[done12many2]

Just now, PCGuy_5960 said:

Yep, so der8auer was 100% right when he said that Skylake-X will be an overclocking beast

 

My favorite parts is when all the overclocking "experts" here on LTT had their explanations of how it wouldn't be so.  Der8bauer just likes to make shit up, ya know?

[cj09beira]

1 minute ago, PCGuy_5960 said:

CInebench produces quite a lot of heat.....

but its short in time, they would probably have the cpu slowing down on a second run

[VagabondWraith]

4 minutes ago, PCGuy_5960 said:

I would be more than happy with 4.7GHz at 1.25V TBH, so not a very bad thing

Me too. Even 8 cores at 4.5-4.7 and I'd be happy.  Shit, that's like a 7700K with 8 cores instead of 4.

[Spectrez]

Nice performance boost thats for sure. But damn that power consumption.

 

I wonder if the temps are still under NDA

[PCGuy_5960]

3 minutes ago, done12many2 said:

My favorite parts is when all the overclocking "experts" here on LTT had their explanations of how it wouldn't be so.  Der8bauer just likes to make shit up, ya know?

Not only on LTT but on OCN too

[knightslugger]

1 minute ago, cj09beira said:

but its short in time, they would probably have the cpu slowing down on a second run

When on liquid, you HAVE to let the liquid saturate to get highest temps. Otherwise you're just pissing on lava.

[PCGuy_5960]

2 minutes ago, cj09beira said:

but its short in time, they would probably have the cpu slowing down on a second run

Even in this short time, if the voltage was at 1.5V, the system would have overheated

[MageTank]

@PCGuy_5960 Friend, I think you've reached the "how many times one can say AVX in a single thread" quota. Dial it back some. AVX, as good as it is, is not widely adopted, and likely won't be any time soon. Even with the extremely efficient AVX-512 on these upcoming chips, I can't imagine many people will adopt it due to the amount of heat it will add when used. Consumers are afraid of hot things, and anything that shakes their "pseudostable" overclocks stability. As I stated before, you have to be clear which AVX you speak of when comparing Ryzen and Intel*. Their AVX2 performance (after tweaked tertiary timings, specifically tRDWR and tWRRD) should be relatively close. AVX1 and AVX3 (512) should absolutely hammer Ryzen, but as I said before, the use cases outside of niche applications are few and far in between.

 

I am interested to see why the 10c Skylake X chip is smashing the 10c Broadwell-E chip, but we will find out once it's released. My money is on the cache improvements that Intel announced, but it's hard to be certain at the moment.

 

7 minutes ago, mynameisjuan said:

I am glad to see the shit storm behind not soldering the heat sink really followed through to everyone complaining. Oh hey would you look at that, its not that bad.

Who said it was going to be bad? The stock paste was NEVER the issue. We've repeated this all over the forums so much, that it's beyond beating a dead horse by now. The gluing process was always the issue. Intel went with a multi-layered substrate this time around, and it's almost certainly changed their gluing process. It's not out of the question to assume they are using the exact same Dow Corning paste (or at least a similar formula) and the results when delidded (as shown by Der8aur) are still in line with what we expect from a LM vs silicone paste comparison. Not the dramatic 20C-30C difference from the terrible aforementioned application process, but an improvement nonetheless.

 

As for the others that are mentioning it's overclocking, be very weary. It's hard to tell whether or not these review samples are binned. Der8aur himself stated he received a heavily binned sample for his test, and it wouldn't be out of the question for these reviewers to receive a similar sample. While it won't matter to me (10c at 4.5ghz is still extremely impressive in my eyes) it may not be as simple to achieve the 4.7-5ghz that these people are getting. Not without some heavy binning and extremely exotic cooling solutions. 

 

Either way, this is shaping up to be a massive fight at the HEDT/HPC level. Hopefully this results in some better pricing to trickle down to the consumer hardware. I think AMD is going to try to win the core-war, and push at Intel's Xeon market, while Intel is going to try to maintain it's performance crown in the enthusiast market. Both seem to be doubling down on their own bets, so it's going to be interesting to see how it all plays out. 

[TacoSenpai]

Pretty impressive stuff hopefully this pushes things forward for CPUs.  Maybe in 2-3 years time I'll build one.

[PCGuy_5960]

7 minutes ago, MageTank said:

 

@PCGuy_5960 Friend, I think you've reached the "how many times one can say AVX in a single thread" quota. Dial it back some. AVX, as good as it is, is not widely adopted, and likely won't be any time soon. Even with the extremely efficient AVX-512 on these upcoming chips, I can't imagine many people will adopt it due to the amount of heat it will add when used.

LOL, yeah  But almost everyone is quick to say "Threadripper will be better" without knowing that there are a lot of instances in which this is simply not true...

BTW, in case you are interested,the article has memory benchmarks