Try as you might but - new Intel 8-core still behind AMD

[Trik'Stari]

9 hours ago, Bombastinator said:

I haven’t seen a laptop in general in years that didn’t run hot.  They want small so bad (still don’t get that one) and the result is hot machines

Better thermal paste couldn't hurt. Might raise the price $00.50 per unit, but it couldn't hurt.

[leadeater]

41 minutes ago, porina said:

I get what you're saying, but I guess my point is that, the spec defines operation at one (or more) points. Unless operating at those points are mandatory, operating at different points is not in itself out of spec.

 

To pick a car analogy, typically you'll be given a fuel economy at a given speed. If you then travel at a different speed, you don't consider that out of spec do you?

 

1, run at defined TDP settings, get expected performance - in spec

2, run at defined TDP settings, don't get expected performance - out of spec

3, run other allowable settings: not out of spec

 

That doesn't really fit the situation well, like I know these can be more considered default values but they are considered the specification of the CPUs to basically every OEM computer manufacturer out there, it's the minimum expectation set by Intel. So if you are HP or Dell designing a business class desktop computer that is what they actually design around PL1, PL2 and Tau all being the exact values Intel put in the documentation, they can configure it to something else if they choose.

 

By Intel allowing these values to be configured I simply consider that the transference of responsibility to the party that changed them and it is their responsibility to document these values. An Intel CPU in a Asus motherboard is operating at Asus defined spec not Intel's etc etc.

 

I consider any Intel CPU where the PL1 is not the TDP value on the Intel ARK page, PL2 not 1.25 times PL1 (for desktop parts) and Tau not 8 seconds as not being Intel spec. It's just some other value that Intel allows you to change with their blessing, CPU still isn't Intel strict spec. I think you are confusing running to defined spec as being in or outside of Intel allowed parameters and would void warrant or something.

 

All I'm pointing to is that Intel has these values documented and the HP system is following those documented values and gaming motherboards do not. So that means this HP system is not an exact indicator of the CPU performance us gamers building our own systems will get.

 

My original comment was around the CPU operating conditions, not what Intel allows or disallows you to do with the CPU.

[porina]

3 minutes ago, leadeater said:

I think you are confusing running to defined spec as being in or outside of Intel allowed parameters and would void warrant or something.

"out of spec" requires for there to be an actual violation of specification. If the TDP defined operating behaviour is used or not doesn't matter (unless it is made mandatory by Intel).

 

3 minutes ago, leadeater said:

My original comment was around the CPU operating conditions, not what Intel allows or disallows you to do with the CPU.

Agreed, however we got here we're on a bit of a tangent now.

[leadeater]

8 minutes ago, porina said:

"out of spec" requires for there to be an actual violation of specification. If the TDP defined operating behaviour is used or not doesn't matter (unless it is made mandatory by Intel).

True but to be very clear I'm using as to spec as the most literal exact value in the documentation. Not the context you are using, if I open technical documentation and it has X value that is the defined value in the specification. It's not a matter at all what is allowed or not, not here for what I am saying to "running as to spec". If it's not the value in the documentation it's not the value in the documentation. 

[SpaceGhostC2C]

 

Now that Apple is not involved, am I allowed to say Geekbench doesn't mean much to me again?

 

 

 

Now

17 hours ago, Random_Person1234 said:

Between this and Intel only selling 5,000 10th/9th gen CPUs compared to 35,000 for AMD Zen 3 and Zen 2 on MindFactory (German retailer), this doesn't look good for Intel.

17 hours ago, Techstorm970 said:

Holy shit!  Didn't know it was that big a disparity lol!

 

You need to consider that the CPUs sold separately to end consumers for them to build their own computers or replace pre-installed ones is a small fraction of the market. While it is a reversal from pre-Zen days, it's not new (I think Mindfactory already was selling more AMD CPUs with the Ryzen 2000 series), and it's not going it to make it or break it for either company.

[leadeater]

3 minutes ago, SpaceGhostC2C said:

Now that Apple is not involved, am I allowed to say Geekbench doesn't mean much to me again?

No

 

 

Geekbench is the best benchmark ever and represents very closely application performance and user experience of your computer.

[AlTech]

18 hours ago, GDRRiley said:

did people really think they could match the 5800x?

 

It's a little bit sad that we've got to this point but it explains the 5800X's quite terrible price especially in relation to 5900X for only $100 more with 50% more cores.

 

16 hours ago, BuckGup said:

This is a terrible comparison. A prebuilt will cut corners everywhere to save money so the motherboard power delivery is probably shit, shit PSU, shit cooling, and a shit OC profile. 

Yes but Intel can't go beyond 8 cores in Rocket Lake so there's that. A 5GHz 8 core is still an 8 core.

 

I'd personally be interested to see if they skip i9 for Rocket Lake and make i5 8C/8T and i7 8C/16T. i3 could then be 6C/6T and Pentium could then be 4C/4T.

[porina]

10 minutes ago, leadeater said:

Geekbench is the best benchmark ever and represents very closely application performance and user experience of your computer.

I thought that was Cinebench  

 

3 minutes ago, AluminiumTech said:

Yes but Intel can't go beyond 8 cores in Rocket Lake so there's that. A 5GHz 8 core is still an 8 core.

And still plenty for the vast majority of consumer uses. Those really needing more cores may gravitate more towards AMD, but it isn't a necessity at 8 or below.

 

3 minutes ago, AluminiumTech said:

I'd personally be interested to see if they skip i9 for Rocket Lake and make i5 8C/8T and i7 8C/16T. i3 could then be 6C/6T and Pentium could then be 4C/4T.

Intel have certainly left themselves in an awkward marketing position with a numerical downgrade in max cores available, and the average IPC increase expected by itself wont be enough to cover those missing cores. I doubt they're going to pull more clock out either. Still, my gut feeling is that selling non-HT CPUs now is not really a realistic thing, especially not on the higher end, so I'd think the logical offerings are 8c16t, 6c12t, 4c8t. How they stack it up the chain is another matter...

[OscarMike]

wait are you comparing a laptop part with a desktop part in this benchmark?

[AlTech]

7 minutes ago, porina said:

I thought that was Cinebench  

 

And still plenty for the vast majority of consumer uses. Those really needing more cores may gravitate more towards AMD, but it isn't a necessity at 8 or below.

We'll see how good or bad Intel's Rocket Lake is gonna be but for now AMD is taking full advantage and milking customers for every penny with Zen 3.

 

[AlTech]

8 minutes ago, OscarMike said:

wait are you comparing a laptop part with a desktop part in this benchmark?

No? It's a desktop to desktop comparison.

 

The Rocket Lake CPU is in a pre-built desktop though.

[AlTech]

29 minutes ago, leadeater said:

snip

You forgot to put a /s at the end.

[OscarMike]

1 minute ago, AluminiumTech said:

No? It's a desktop to desktop comparison.

 

The Rocket Lake CPU is in a pre-built desktop though.

ah right i see

[leadeater]

3 minutes ago, AluminiumTech said:

You forgot to put a /s at the end.

But that is the best part, people can choose to take that however they like 

[MageTank]

3 hours ago, porina said:

It's not an AMD product, and therefore it must be rubbish.

 

More seriously, I'm not really familiar with geekbench in general, but I just poked around the website a bit. The result posted in OP is generally "a bit" behind other 5800X results, but I also had a look at 3800XT results, which get beaten.

 

So, Rocket lake will be somewhere between Zen 2 and Zen 3 performance overall, putting aside the lack of offerings above 8 cores. I think this is not an unexpected place for the CPU to land, and there is still some scope to move around depending on final clocks. We already know Zen 2 is, on average, ahead of Skylake architecture. Rocket Lake/Cypress Cove, drawing heavily on Ice Lake/Sunny Cove configuration, is a major update from Skylake, and it looks like clocks aren't suffering to implement it although we have to see where they end up in practice.

 

When Ice Lake/Sunny Cove was originally released, Intel claimed an 18% IPC uplift relative to Skylake based on a mixed workload average. AMD's Zen 3 vs Zen 2 was claimed to be 19%. The test conditions may not be the same between them, but slotting in between Zen 2 and Zen 3 in IPC is still a good catch up move.

Worry not. I can familiarize you with Geekbench...

 

This post is a few years old, but sums my feelings up about this benchmark. Admittedly, I have not used it since, so it's entirely possible they've cleaned up their ways, but I belong to a circle of overclockers that still gripe about it to this day, and my place of employment refused to contract with them for benchmarking for a reason.

 

You'll find my test runs and methodology in that post, along with proof of my various clock speeds relative to Geekbenches reporting. It was such an easy test to trick and the results were all over the place. I could easily convince you I had a god tier system by overclocking my RAM and the test would be none the wiser. For those that believe I faked it, remember to use the time stamps for Geekbenches submission, CPU Z's submission and the forum post date as cross-references.

 

I am down to test everything all over again given I have far more hardware resources as my disposal now, but I'd be lying if I told you I had even an ounce of faith in any Geekbench result.

[porina]

1 minute ago, MageTank said:

This post is a few years old, but sums my feelings up about this benchmark. Admittedly, I have not used it since, so it's entirely possible they've cleaned up their ways, but I belong to a circle of overclockers that still gripe about it to this day, and my place of employment refused to contract with them for benchmarking for a reason.

Thanks, I'll have to look through that in more detail another time. I did dabble a bit with hwbot in past and that's about the limit of my experience of running geekbench of any version. Overclocking isn't a major thing for me to play with any more, although I might still help out the Country Cup if called upon. By chance, I see it is running now...

[Lord Szechenyi]

20 hours ago, williamcll said:

A New Benchmark for a HP Omen PC has appeared on Geekbench equipped with a new intel CPU, with performance numbers close but still behind the Ryzen 7 5800X

Source: https://videocardz.com/newz/intel-8-core-16-thread-rocket-lake-s-cpu-offers-similar-performance-to-ryzen-7-5800x
Thoughts: If Intel can somehow price their products low now, it might actually beat AMD in price over performance and maybe win back some potential customers. However I also suspect that the new chip can actually go faster but HP's usual poor thermal/power management is bringing it down, as seen with Linus' recent video. One thing I would like to ask, has the IPC improvement as good as Intel claimed back on LTT's Zen 3 review video?

3 things

- No price (as you mentioned)

- No number (there will be many 8/16 11th gen cpus)

- HP machines are meh

[LAwLz]

5 hours ago, TempestCatto said:

I'm sorry, but I just don't see this as near enough of an "improvement" to be competitive with AMD right now. Not unless the price speaks for the performance.

Why did you put improvement in quotation marks? 

What would Intel have to do to be competitive with AMD in your eyes? I think being tied in terms of performance makes them competitive. Pricing seems to be fairly similar too, at least current gen stuff, now that AMD has raised their prices. 

[porina]

https://videocardz.com/newz/msi-mistakenly-reveals-8-core-16-thread-intel-rocket-lake-s-cpu-z-specificaitons

 

MSI apparently leaked some Rocket Lake specs through a CPU-Z screenshot, showing 125W TDP, 3.4 GHz base, running at the time 4.3 GHz, and a max multiplier of 50 suggesting 5 GHz turbo.

 

Rocket Lake leak: 8c16t, 3.4 GHz base, 5.0 GHz turbo, 125W TDP

i9-10900k: 10c20t, 3.7 GHz base, 5.1 GHz regular turbo, 5.3 GHz max, 125W TDP

17-10700k: 8c16t, 3.8 GHz base, 5.0 GHz regular turbo, 5.1 GHz max, 125W TDP
 

Regular turbo clock can be reached by any core, but not necessarily at the same time. The higher turbo modes are on selected cores if conditions are right, more like AMD style. The possible drop in base clock is interesting as it implies worse power efficiency per clock, but this should be more than offset by IPC improvements leading to an overall higher perf/clock and perf/watt.

 

Some noted the presence of AVX512F, although if existing client CPU implementations are to go by, it'll be single unit so FP64 performance wont see the massive uplift that Skylake-X and related saw.

[StDragon]

21 hours ago, comander said:

...why use that hot mess when AMD has overall better parts (perf/watt), cheaper motherboards and what not...

X570 Chipset fans are loud! It's basically a Matisse IO die at 14nm consuming anywhere from 11w 15w of power. All others are about in the 5w range which is why it can be passively cooled.

 

Unless this is a one-off solution from AMD, this doesn't bode well going into the rumored X670 chipset. Some have speculated it might run around 8w. We shall see.

[DildorTheDecent]

13 minutes ago, StDragon said:

X570 Chipset fans are loud!

>what is fan tuning and how can it help reduce noise.

 

Just turn the fan down lmao.

[RejZoR]

X570 fans were loud in the beginning. Boards sold these days have the fan curves tuned to a point they are basically inaudible. Especially on higher end boards.

[StDragon]

1 hour ago, RejZoR said:

X570 fans were loud in the beginning. Boards sold these days have the fan curves tuned to a point they are basically inaudible. Especially on higher end boards.

I have an x570 Taichi with the latest BIOS v3.61. So I know all about that. I can still hear it with the tower 3ft way from me. But you can really hear it once you start gaming.

 

I've already ensured good SB die contact on the heat spreader plate. The only solution is to put a waterblock on it; which BTW I'm already in the stages to do so. 11w of heat is 11w of heat. There's no getting around it.

 

 

 

 

[RejZoR]

43 minutes ago, StDragon said:

I have an x570 Taichi with the latest BIOS v3.61. So I know all about that. I can still hear it with the tower 3ft way from me. But you can really hear it once you start gaming.

 

I've already ensured good SB die contact on the heat spreader plate. The only solution is to put a waterblock on it; which BTW I'm already in the stages to do so. 11w of heat is 11w of heat. There's no getting around it.

 

 

 

 

I don't know about Taichi, but X570-E from ASUS has different fan profiles as option even for chipset.

[Blademaster91]

44 minutes ago, StDragon said:

I have an x570 Taichi with the latest BIOS v3.61. So I know all about that. I can still hear it with the tower 3ft way from me. But you can really hear it once you start gaming.

 

I've already ensured good SB die contact on the heat spreader plate. The only solution is to put a waterblock on it; which BTW I'm already in the stages to do so. 11w of heat is 11w of heat. There's no getting around it.

 

 

 

 

It could be something with the X570 taichi, not sure, but most boards allow you to set chipset fan so its off unless the chipset hits a certain temp.