Intel Core i9-7900X reviews are going live!

[Fanatic]

Suffice to say I would take what has been written on Toms and others with a massive sack of salt. Based on their article I've been worrying myself silly about shifting heat and it has been exactly the opposite. My XMP based overclock has set a 4.2Ghz base and my typical have this thing running about 20c cooler than my 3930K for a similar overclock. I need to do more benching but this thing just isn't getting hot at all, and the TDP hasn't been anything close to the fabled 240w that they claimed they were seeing. I thought it didn't stack up at the time that I read that, and my actual experience is indicating that my gut instinct was correct. Either that article is a crock, or they were reviewing an engineering sample and perhaps my OEM lower tray priced unit is from a more recent batch - because it's freakin amazing! It's early days and maybe I've just not hit a workload that is really going to push it. I've got some AVX based workloads I can throw at it (which is another reason why I didn't wait for threadripper) and it might be that I'll start seeing something different then.

I think TTL did say under normal usage you won't see those insane power draws and temperatures. To get them you need to make specific changes in the BIOS to the safety features, be running the most demanding version of Prime95, and at 4.8GHz+ on 10 cores.

Personally I've got my 7820X running at 1.0V. Maybe I'll overclock it in the future. I upgraded from a 3770K. Having to buy another 16GB DDR4 kit for quad channel is what took the price higher than I had planned. And then I decided "bugger it" and bought a 500GB PCI-E NVMe SSD as well .

 

[N19h7m4r3]

HAHAHA! Seriously! WCCFT's scale on their graphs!
http://wccftech.com/review/intel-core-x-cpu-asrock-x299-taichi-review/9/

In Cinebench Singlethread the 1800X beats the 7900X despite it's really high boost clocks.

Whelp the CPUs certainly like to sip power and dump heat; even at stock speeds.

 

[Erik13]

Wccftech : 'Lets make 1 fps difference look like 100'

 

[humbug]

HAHAHA! Seriously! WCCFT's scale on their graphs!
http://wccftech.com/review/intel-core-x-cpu-asrock-x299-taichi-review/9/

In Cinebench Singlethread the 1800X beats the 7900X despite it's really high boost clocks.

Whelp the CPUs certainly like to sip power and dump heat; even at stock speeds.

I commented on that, with in minutes it was removed, so i watched it for a bit.... posts commenting on it are going up and are removed minutes later.... pathetic.

 

[Hotwired]

You must be imagining things, the webpage linked has properly zeroed graphs

 

[kitfit1]

Let's do a proper comparison test between all these Intel SKU's and throw in 2 AMD SKU's for good fun. Tell you what though, let's leave the 2 AMD's at stock speed then we can show just how much better the Intel SKU's are.

• 
  
  
  • Intel Core i9-7900X @ 4.5 GHz With 1.283V
  • Intel Core i7-7820X @ 4.7 GHz With 1.261V
  • Intel Core i7-7800X @ 4.8 GHz With 1.237V
  • Intel Core i7-7740X @ 5.0 GHz With 1.270V
  • Intel Core i5-7640X @ 5.1 GHz With 1.278V

 

[DragonQ]

They also clearly don't know what margin of error is.

Let's do a proper comparison test between all these Intel SKU's and throw in 2 AMD SKU's for good fun. Tell you what though, let's leave the 2 AMD's at stock speed then we can show just how much better the Intel SKU's are.

Not sure it makes any difference, an R7 1800X has barely any overclocking headroom anyway.

 

[Panos]

Wccftech : 'Lets make 1 fps difference look like 100'

Actually relative the bar looks 50% for 2 fps, and we do not know the ram speeds on the Ryzen.
Clearly not 3600 as per kit used, because the numbers are close to 2400Mhz ram. Also Tomb Raider isn't with the latest patch, because the numbers are even less than those few months back, on the Ryzen CPUs

Shame I am not a reviewer :/ If I had the gear they have, could have run much better and accurate benchmarks.

 

[Dirtyganker]

My AVX workloads I did at the weekend simply got crushed and I don't have any meaningful data for that on my Core i9-7900X @4.2Ghz. I did some Prime 95 runs for about 3 hours and couldn't get anything above 55c. Cores 1 and 2 @4.5Ghz, the rest scaling down toward 4Ghz. I've also removed my 240mm rad and now only running with a single 360mm rad, so nothing too fruity there either and it's made no difference to the temps.

It's rock solid and as a full time overclocked I'm more than happy.

 

[N19h7m4r3]

My AVX workloads I did at the weekend simply got crushed and I don't have any meaningful data for that on my Core i9-7900X @4.2Ghz. I did some Prime 95 runs for about 3 hours and couldn't get anything above 55c. Cores 1 and 2 @4.5Ghz, the rest scaling down toward 4Ghz. I've also removed my 240mm rad and now only running with a single 360mm rad, so nothing too fruity there either and it's made no difference to the temps.

It's rock solid and as a full time overclocked I'm more than happy.

Ah, you're essentially running it at it's boost settings the entire time. Shame you have to get a 360mm rad to tame the system. Although yes, if you have lots of AVX you really do need it all.

I find it rather annoying when reviewers and overclocks/binner sell these CPUs are 100% stable at 4.5-4.7Ghz, but then in small print state it's for non AVX loads; and even then hit 90+ Degrees on 240/280mm rads.

I saw on Tech Yes City that he mentioned his 2 7900Xs are very different for voltages and temps. So much so, his delidded one topped at 80 degrees, where as his other un-delidded one topped at 82 degrees.

So the silicone and thermal paste lottery has a large delta.

 

[Dirtyganker]

Ah, you're essentially running it at it's boost settings the entire time. Shame you have to get a 360mm rad to tame the system. Although yes, if you have lots of AVX you really do need it all.

I dare say you could get away with something smaller, although I suspect an all-in-one cooler might be a challenge as I don't supposed they can manage anything close to the flow rate that a custom loop with a D5 pump running at full tilt can achieve.

I saw on Tech Yes City that he mentioned his 2 7900Xs are very different for voltages and temps. So much so, his delidded one topped at 80 degrees, where as his other un-delidded one topped at 82 degrees.

So the silicone and thermal paste lottery has a large delta.

I suspect that the early review samples are definitely a lottery, but we've already had a price drop and OEM parts are now available too so I suspect that the current newer low-price stock will be a different batch. I was led to believe that I was going to have a challenge cooling it on stock settings, but that has been proven to be utter BS. My 3930K was way hotter in the order of 75c at a similar overclock on Prime95.

 

[Erik13]

'So much so, his delidded one topped at 80 degrees, where as his other un-delidded one topped at 82 degrees.'

I swear users who delid this cpus have balls. I would have gone nuts if i have broken one during the process lol

 

[N19h7m4r3]

'So much so, his delidded one topped at 80 degrees, where as his other un-delidded one topped at 82 degrees.'

I swear users who delid this cpus have balls. I would have gone nuts if i have broken one during the process lol

I know right? Two $1000 CPUs, delids one, voids warranty and chance of killing it.

I wonder if the 12-18 cores will be soldered, or if people will delid them as well to replace TIM.

 

[AllBodies]

Not sure it makes any difference, an R7 1800X has barely any overclocking headroom anyway.

Not really true for multi-core loads, only single-core.

The 1800X will boost to 4.0-4.1 GHz for 1-2 core loads, but only 3.7 GHz for all-core loads.

So it has ~10% overclocking headroom for all-core. That's enough to be significant.

Also lets not forget RAM speed. Does that test show what RAM speeds they used?

 

[humbug]

They also clearly don't know what margin of error is.


Not sure it makes any difference, an R7 1800X has barely any overclocking headroom anyway.

What these days does? look at the Intel side, 7700K 4.2Ghz base clock to 4.8Ghz is less than 15%.