OcUK Comet Lake S review thread

[ljt]

Up to 30fps difference, at 1080p, between the i7/i9 and 3900X in some games.

*if you game at 1080p with a 2080Ti

 

[Grim5]

I see. It certainly looks attractive to gamers though, I reckon these will sell well in that demographic. Up to 30fps difference, at 1080p, between the i7/i9 and 3900X in some games. Of course that difference vanishes at 1440p or above.

Better to use % than "30fps"

For instance 30fps extra at counter strike when you already get 500fps means absolutely nothing

 

[RavenXXX2]

Let's check out the average of the seven games. Given what we just saw it won’t surprise you to learn that on average the 10900K is 7% faster than the 3900X. That’s not a big margin at 1080p when using an RTX 2080 Ti, but it does mean Intel retains the gaming performance crown.

Dunno about you but I would save the £100 and reduced power bill and no need for exotic cooling and go with the 3900X.


https://www.techspot.com/review/2028-intel-core-i9-10900k/

 

[LeMson]

Dunno about you but I would save the £100 and reduced power bill and no need for exotic cooling and go with the 3900X.


https://www.techspot.com/review/2028-intel-core-i9-10900k/

Isn't it way more than 100 though, you need a cooler too don't forget, so it's more like 200? and the 3900x will run happy on a 100 quid board, the 10900k needs a board maybe twice that? so that's a total of 250-300?

 

[RavenXXX2]

Absolutely, and for 2 less cores, madness.

 

[Panos]

*if you game at 1080p with a 2080Ti

And play Farcry 5 in DX11 with no Gsync/freesync 240hz monitor.........

Things are narrowing to a tiny segment of the market.
And lets not forget. All motherboards except the one tested by HU, use MCE by default. So the 10900K is overclocked on all benchmarks we see.

Yet nobody goes to activate CPPC on the 3000 Ryzens just to force the system use the fastest cores in games!!!!! (not even overclock etc).

Which comes to down, why the motherboard manufacturers have CPPC off to start with.

 

[silvagti]

*if you game at 1080p with a 2080Ti

In 2020 if you have £500 to spend on a processor it's *massively* unlikely you'll be gaming at 1080p though! If you're willing to spend that much on a CPU alone you'll almost certainly have a 1440p/4K monitor as well.

I understand why they run the benchmarks for games at 1080p to highlight the performance difference but it's increasingly meaningless especially for those running high end gaming systems. It's a tough sell even to the "money no object; I'll pay top dollar for the absolute best part that's a few percent faster crowd"

 

[LeMson]

Absolutely, and for 2 less cores, madness.

I'm just looking at these reviews comparing 10900k to 3900x wondering why when cost difference is massive O_O

Or am I wrong?

Thing this who in 2020 if you have £500 to spend on a processor it's *massively* unlikely you'll be gaming at 1080p though! If you're willing to spend that much on a CPU alone you'll almost certainly have a 1440p/4K monitor as well.

You are also using gsync or freesync and the difference disappears into smoothness on both

 

[benjii]

Thing this who in 2020 if you have £500 to spend on a processor it's *massively* unlikely you'll be gaming at 1080p though! If you're willing to spend that much on a CPU alone you'll almost certainly have a 1440p/4K monitor as well.

I understand why they run the benchmarks for games at 1080p to highlight the performance difference but it's increasingly meaningless especially for those running high end gaming systems.

That is very true. However, collectively people are stupid. Many people take these benchmarks at face value without realising that the bottleneck at 1440/4k removes all but a few percent of the advantage in gaming.

 

[CAT-THE-FIFTH]

It will be interested to see how the Core i5 10400F performs TBF. It seems to be halfway between a Ryzen 5 3600 and Ryzen 5 3600X in price.

 

[Panos]

You are also using gsync or freesync and the difference disappears into smoothness on both

At last another voice of common sense and logic in this forum.

 

[LeMson]

At last another voice of common sense and logic in this forum.

ow my

So having a looksie at tom's hardware review. In the space of 2 paragraphs, we have the 10900k competing with the 3900x on price -> Instead, the 10900K competes with the 12-core 24-thread Ryzen 9 3900X in terms of both performance and price, but Intel's chip has the highest power consumption we've seen recently on the mainstream desktop.

then to -> The 10900K's high power consumption even overwhelmed our 280mm watercooler during some tasks.

How is it competing on price if it's pwning a 280mm AIO when pushed. It's a new socket so the mobo price has to factor into these reviews, as does the cooler if one is needed.

Kitguru review, cons: no mention of needing a cooler O_o

I can prolly pick on others, but meh

 

[Minstadave]

ow my

So having a looksie at tom's hardware review. In the space of 2 paragraphs, we have the 10900k competing with the 3900x on price -> Instead, the 10900K competes with the 12-core 24-thread Ryzen 9 3900X in terms of both performance and price, but Intel's chip has the highest power consumption we've seen recently on the mainstream desktop.

then to -> The 10900K's high power consumption even overwhelmed our 280mm watercooler during some tasks.

How is it competing on price if it's pwning a 280mm AIO when pushed. It's a new socket so the mobo price has to factor into these reviews, as does the cooler if one is needed.

Kitguru review, cons: no mention of needing a cooler O_o

I can prolly pick on others, but meh

If they're overly critical perhaps those review samples won't keep turning up.

 

[Pannalicour]

I find it strange that the 9900k in gaming benchmarks is performing worse then before. In hardware umboxed it was sometimes slower than the 9700k. Is this some kind of trick to make the 10900k look better?

 

[FireTwo]

3900x + x570 unify cancelled, hello 10900k z490 unify!

 

[Panos]

I find it strange that the 9900k in gaming benchmarks is performing worse then before. In hardware umboxed it was sometimes slower than the 9700k. Is this some kind of trick to make the 10900k look better?

No. HU always patches the BIOS to the latest when trying to do reviews. It doesn't use 18 month numbers from a review back then (like Anandtech does)
Which means all hardware security mitigation are applied, which for Intel means reduced performance because of the security flaws on the HyperThreading.

Last September batch of mitigations found itself into Linux, reduced the Intel 9900K perf by 5%-7%. However Windows wasn't patched until December, AFTER the 9900KS reviews were out....

Same we will see now, as Intel still has 172 mitigations for the whole line up (incl 10 series) on the Windows 10 -2004 (due out by end of the month). So don't be surprised if comes November and RKL-S is out (11 series with Z590) you see 10900K having lost performance.

 

[Panos]

3900x + x570 unify cancelled, hello 10900k z490 unify!

Ahh someone who took the blue pill.
Enjoy.

 

[Toasty]

It’s cool for an overclocking platform to play with some serious cooling, but kind of mad to buy one for £100 more than 3900x for regular use the benefits are so niche now.

The marketing/shill reviews works though a lot of people seem to think it’s automatically best for gaming by a large margin no matter res they play at or GPU they have.

Can see people who buy will have buyers remorse in few months when replaced with Rocketlake if it’s 15% IPC or Zen3 it’s just not right time to buy now.

 

[FireTwo]

Ahh someone who took the blue pill.
Enjoy.

not sure I wanted to be forcibly ejected from the liquid-filled chamber unconscious

 

[seanyc5]

Am I still right in saying then at 4K a 6700k @ 4.5ghz with a 2080ti is still ok, I play with v-sync on at 60fps on a 4K TV.

I'm safe for another generation still?

I won't see any benefit gaming-wise while at 60fps. Hopefully, we get some nice benchmarks when Cyberpunks comes out with 4K in it. I no the difference disappears but its nice to see to actually see some numbers haha