• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD to unveil Zen 4 CPUs at CES 2022

Status
Not open for further replies.
Don't forget power usage doesn't scale linearly with performance though, take nvidia ampere for example which Jensen claimed had 1.9X better performance per watt than Turing but in reality it's not much better at all.
 
Don't forget power usage doesn't scale linearly with performance though, take nvidia ampere for example which Jensen claimed had 1.9X better performance per watt than Turing but in reality it's not much better at all.

How did you measure "not much better at all"?

Did you use the same wattage Nvidia used in their test and if so, what was it?

I would assume that to conduct such a test would required taking a RTX 2080ti, RTX 3090 and then undervolting both to about 250w and then measure performance in some synthetic test like FP32
 
Last edited:
I think I said £500+ for the X3D and we are now almost there, so no I wasn't waaaaaaaaaaaaaay off.

No you didn't, and yes you were. :cry: :cry: :cry:

It maybe down to the costs also as the 5800X3D could well be a £600 part when it releases as they were comparing it to the £600 12900k

I'd say $550-600 will be in the ballpark as AMD know enough people will pay a large premium for gaming marketed products.

Half the people who buy 5950X/5900X now are just using them as gaming chips.

Also the 5800X3D is available at MSRP, OCUK isn't the only shop on the internet, just in case you didn't know. ;)
 
No you didn't, and yes you were. :cry: :cry: :cry:





Also the 5800X3D is available at MSRP, OCUK isn't the only shop on the internet, just in case you didn't know. ;)
I'd say $550-600 will be in the ballpark as AMD know enough people will pay a large premium for gaming marketed products.

Half the people who buy 5950X/5900X now are just using them as gaming chips.

$550-600 is £451-492 so I'd say I was pretty accurate.
 
$550-600 is £451-492 so I'd say I was pretty accurate.

The MSRP is is $449 and that is what it is available for, stop trying to defend you inaccuracy using a single over priced part at one retailer. You are just making yourself look desperate and daft. Easier to admit you were wrong.

I also note you skipped the part where you said it would be £600. :cry:
 
£429 is the only price I can see where its in stock and thats still above MSRP. The likelihood is the prices will continue to rise as AMD switches focus to Zen 4.

$449 / 1.22 = £368 +20% = £441

That's actually slightly under MSRP.
 
I'm not the only one whose noticed this, MLID is constantly being told by his subscribers that single core is all important, to the point where he pulled them up for it and cited people like Tim from Hardware Unboxed as "should know better".

When you watch him you can see why, in his last video pretty much said he thought Rocket lake will beat Zen 4 in games because Rocket Lake has higher single threaded performance, he should know better so i don't know why he does this, its like he has some sort of cognitive dissonance when it comes to Intel.

The thing is it doesn't matter how much that is disproved they just keep doing it....

I really don't see the issue, I'd be inclined to think the same about Rocket Lake given how well Alder Lake has performed in games this generation. And where does Alder Lake shine? Single threaded performance, except they now also have the number of cores to out do Zen 3 too. As we know there is no single reason why performance is better with one CPU to the next, there is a whole host of reasons, however, outright thread performance is usually quite measurable and also associated with a bump in gaming performance.

It is intresting, the 6 cores Zen 3 has access to the same 32MB cache pool as the 8 core, and yet there is no difference gaming performance, its the same for the 12700K vs 12900K, clock for clock there is no difference despite 25MB vs 30MB of L3, normally its how much cache per core you have.

Perhaps its not quite as simple as just add cache.

Maybe the games you are comparing are satisfied with the 12 threads on offer with the 5600X so won't benefit from a further 4 threads with 5700/5800X? If the 5600X could clock higher than it does then it may perform better than the 8 core siblings, bumping up single threaded performance too, if you get my drift.

It's a little daft to think HUB are biased towards Intel too, It's not that long ago that they wholehearted praised Zen 3 and the 5600X was their go to recommendation for most. It was only when Alder Lake released that they changed their tune, and rightly so. The 5800X3D has retaken their head role again.
 
Cache is just a resource. It's about the workloads. If you extend cache by 5mb, the workloads that were on the border and can now fit within that extra 5mb will benefit and show up in benches, etc. However, that's going to limit your pool as 5mb likely isn't enough to cover a large array of workloads.

When you greatly expand the cache say from 32mb to 96mb (like x3d) you open up a much larger window for various workloads to fit into that expanded pool and thus you see performance gains. It just so happens that games (specifically certain game engines) tend to benefit greatly from this. Thus we call the x3d a gaming cpu.

If a workload already was maxed out on cache usage, you won't see any benefit, Cinebench being a great example. CS:GO being another example. In both of these instances the higher core frequency provides greater performance than having more cache which goes unused.
 
I really don't see the issue, I'd be inclined to think the same about Rocket Lake given how well Alder Lake has performed in games this generation. And where does Alder Lake shine? Single threaded performance, except they now also have the number of cores to out do Zen 3 too. As we know there is no single reason why performance is better with one CPU to the next, there is a whole host of reasons, however, outright thread performance is usually quite measurable and also associated with a bump in gaming performance.

This is fine for desktop where power consumption doesn't matter so much, but Alder Lake is slower in MT and still uses more power in power constrained formats like mobile, this despite AMD's architecture being a generation behind and on a process node 2 generations behind.

Besides that the point your making has nothing to do with the point i was making, ST performance is not the be all and end all of gaming performance, arguably, i would argue AMD have the fastest gaming CPU right now, with ST performance very much lower than Alder Lake.
 
This is fine for desktop where power consumption doesn't matter so much, but Alder Lake is slower in MT and still uses more power in power constrained formats like mobile, this despite AMD's architecture being a generation behind and on a process node 2 generations behind.

Besides that the point your making has nothing to do with the point i was making, ST performance is not the be all and end all of gaming performance, arguably, i would argue AMD have the fastest gaming CPU right now, with ST performance very much lower than Alder Lake.
I don't disagree with you about the long-standing performance (and per-watt) of the aging Zen3 architecture, but I wasn't comparing that to begin with. But whilst we're on the point of process nodes, Intel have had a poor time with their process nodes for a while. AMD Zen3 is on TSMC 7 nm which Intel have only just managed to get competitive with (Intel 7), which in turn has offered them the similar perf-per-watt but with them pumping more juice to get ahead.

How do you know that ST performance of the 5800X3D is in fact lower than that of Alder Lake in gaming? I reckon it is actually faster, at least in the games that have responded well to the extra cache. I don't know of any specific testing that has been done to explore this.
 
I don't disagree with you about the long-standing performance (and per-watt) of the aging Zen3 architecture, but I wasn't comparing that to begin with. But whilst we're on the point of process nodes, Intel have had a poor time with their process nodes for a while. AMD Zen3 is on TSMC 7 nm which Intel have only just managed to get competitive with (Intel 7), which in turn has offered them the similar perf-per-watt but with them pumping more juice to get ahead.

How do you know that ST performance of the 5800X3D is in fact lower than that of Alder Lake in gaming? I reckon it is actually faster, at least in the games that have responded well to the extra cache. I don't know of any specific testing that has been done to explore this.

It would be more correct to say that CB R23 ST performance is not the be all end all of gaming performance. This is very clearly shown with the 5800X3D which is about on par with a 5700X in CB R23 ST but is on par with a 12900K + DDR5 in games across a AAA biased suite and if you are into the sim or strategy stuff the 5800X3D is the fastest part on the planet for those kinds of games.

Anybody who is trying to extrapolate how Rocket Lake and Zen 4 will perform in games based on CB R23 ST numbers is going to be wide of the mark because the switch to DDR5 and doubling of L2 cache will help gaming performance but it does not impact CB r23 ST performance.
 
Status
Not open for further replies.
Back
Top Bottom