• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD Offers Enthusiasts More Choice Than Ever Before with New Ryzen 3000XT Processors

The chips in ashes of singularity show average of 8% improvement over x versions. It is most likely a better optimized ECD, PPT, TDC. Something that is internally wired and set most likly.
 
The chips in ashes of singularity show average of 8% improvement over x versions. It is most likely a better optimized ECD, PPT, TDC. Something that is internally wired and set most likly.


Which means a simple firmware or BIOS update should bring just the same improvements to the X chips, too.
 
'new' proseesors?

si they news cpus,loosk just oc'd version,thats it,with new name.

inside not happend nothing...


is it nice to review/test old ryzens VS XT version...

hmm
 
Which means a simple firmware or BIOS update should bring just the same improvements to the X chips, too.

no it most likely won't, because it's probably a new Silicon FIT that can not be done via a bios change, or it's just better quality Silicon FIT. As I said the gains we've seen is only in one game AoS show more than what bios change is valuable of.

I just wonder in that infinity fabric of 2000 MHz is stock, or if it's the ma, overclock. Stock would be better if left some overhead for over clock.
 
The real news here is StoreMi is back? Already?

I thought they dumped it and were going to start over from scratch with a new "product".
 
I honestly see zero reason to be upset or complain. These XT versions ARE NOT replacing the non-Xt versions, AMD is just giving more options and a slight performance boost while leaving the MSRP exactly the SAME. Seriously, what is there to complain about? I also heard that AMD is going to discount the Non-XT versions further as well, so at least to me this seems good all around...or am I missing something here?
Try telling that to the people that bashed intel for "refreshing" SKUs from the angry AMD hate train when intel also made process improvements from 14nm to 14nm+ to 14nm++ with higher drive current and less leakage and yet AMD puts out this XT series with "more mhz" when the main bottleneck of Zen2 is the infinity fabric clockspeed and yet it's perfectly excusable to charge this much apparently according to the cult of AMD fanboys. Not to even mention the 3900X is dirt cheap these days for example.......
 
Try telling that to the people that bashed intel for "refreshing" SKUs from the angry AMD hate train when intel also made process improvements from 14nm to 14nm+ to 14nm++ with higher drive current and less leakage and yet AMD puts out this XT series with "more mhz" when the main bottleneck of Zen2 is the infinity fabric clockspeed and yet it's perfectly excusable to charge this much apparently according to the cult of AMD fanboys. Not to even mention the 3900X is dirt cheap these days for example.......


It's not 2015. It's 2020 and twelve cores must cost $200-$250. Not double of that.
 
I wonder why some Intel fans here declare a 4% increase in ST performance small whereas Intel's leadership in gaming performance in 1440P is 5% and they still hold the flag of gaming leadership very proudly? And I don't critisize the non-Intel fans for saying the same for the small XT's increase in clocks but I will wait for a full review because some binning would help with better IF-memory performance and higher all-core boost clocks that could reach Intel's gaming perfomance enough to call it a real draw (1-2%) and not an almost draw.

If I had any Ryzen 3000 CPU for now I wouldn't bother upgrading at all. Zen3 will make a much bigger difference me thinks.
 
It's not 2015. It's 2020 and twelve cores must cost $200-$250. Not double of that.
;) and they sure are making a good job of doing that - sales are having the 3900X around the $415 mark rather regularly now but no, gotta put out a product that's basically the same CPU with 100mhz more for ONLY $84 for 4% performance gains! Utterly fantastic, or as the AMDs say "price2performance", honestly they were better off just cutting pricing of Zen2 on the higher end and calling it a day.
 
;) and they sure are making a good job of doing that - sales are having the 3900X around the $415 mark rather regularly now but no, gotta put out a product that's basically the same CPU with 100mhz more for ONLY $84 for 4% performance gains! Utterly fantastic, or as the AMDs say "price2performance", honestly they were better off just cutting pricing of Zen2 on the higher end and calling it a day.
Why not ask your savior Intel to do that?
 
Why not ask your savior Intel to do that?
Yes, let's scapegoat to blame intel when we ignore the fact people mocked the 7XXX/8XXX/9XXX series for being "refreshes" despite the fact they had process improvements - particularly 7XXX that pushed overclocking headroom substantially and didn't charge $84 for a mere 4% ST gain in a synthetic benchmark which also ignores the fact the Zen2 arch is still severely IF bottlenecked and a 10600K or similar can quite easily outperform the 3900X in gaming.
 
I don't see any difference from these XT versions compared to the 9900KS or 8086K. It seems like they're just small bumps as the binning has allowed. It is a bit odd to release new chips right before Zen 3, but it makes some sense as Intel has release their 10 series and AMD doesn't want to be uncompetitive for even a couple months. I do hope this doesn't push Zen 3 further down the line. There are plenty of people that will be buying a computer or computer parts now even though they should wait for Ampere, Big Navi, Rocket Lake-S, and Zen 3. And these will likely be the fastest processors for people with X370 or B350 motherboards, and maybe some X470 or B450 boards that miss the update to enable Zen 3. So even if it doesn't make sense for everyone, these could be the last and best processors for that group.

Before the community browbeat AMD to support X470 and B450 with the Zen 3 chips, these chips likely made a lot of sense to upgrade to for those who didn't want to buy a whole new motherboard.
 
Yes, let's scapegoat to blame intel when we ignore the fact people mocked the 7XXX/8XXX/9XXX series for being "refreshes" despite the fact they had process improvements - particularly 7XXX that pushed overclocking headroom substantially and didn't charge $84 for a mere 4% ST gain in a synthetic benchmark which also ignores the fact the Zen2 arch is still severely IF bottlenecked and a 10600K or similar can quite easily outperform the 3900X in gaming.
Are you telling that those are not resfresh?? 10600K may have more performence in gaming, but what is the performence in multitasking?? And people only palys games and dont do anything on their computer??
Does Intel fans have nothing but games to play on their computers??
 
Try telling that to the people that bashed intel for "refreshing" SKUs from the angry AMD hate train when intel also made process improvements from 14nm to 14nm+ to 14nm++ with higher drive current and less leakage and yet AMD puts out this XT series with "more mhz" when the main bottleneck of Zen2 is the infinity fabric clockspeed and yet it's perfectly excusable to charge this much apparently according to the cult of AMD fanboys. Not to even mention the 3900X is dirt cheap these days for example.......
What AMD is doing now seems closer to Intel's devil canyon than the 6700k becoming a 7700k. The 10 nm and the cove arch delay was (is) just frustating for some people. Iirc, devil canyon reception was mostly good. Benchmarks will tell the whole story, but those appears to just be filler for the people who where mad about that 5% lower fps in games with vanilla zen2.
If you have a zen 2 cpu, right now, or don't care about those fps, you have zero incentive to get those xt zen2, and can just wait for zen 3. The 7700k was the new gen, those xt are not.
 
Are you telling that those are not resfresh?? 10600K may have more performence in gaming, but what is the performence in multitasking?? And people only palys games and dont do anything on their computer??
You do realize the arch is being refreshed but the node it's on isn't a refresh? 14nm - 14nm+ alone was a 12% increase in drive current on the node alone whilst running 52% less power. The 10600K is still a better buy than a 3900X for pure gaming given the fact it can overclock without the issues of degrading (and actually gain performance from doing so)
Here's a nice video on that:
And in some cases, it can outperform a 3950X - whilst costing a ton of money less (yes even the cooler and board that people seem to scapegoat to when in desperation searching for something to point their finger at).
In addition to this, only the 3900X ever makes sense for productivity @ $415 as it offers reasonable pricing on the AMD side, whilst the 10900K costs quite an amount more, for gaming though they're just not competitive due to their higher latency design and the infinity fabric and more communication steps making them less competitive in that aspect
 
They need to milk it fast. History has shown us 2 years they have advantage next 20 years they are well down to competitor. So they need to milk you consumers asap.
 
They need to milk it fast. History has shown us 2 years they have advantage next 20 years they are well down to competitor. So they need to milk you consumers asap.
Pretty much this, one day people will realize a business is only there to make money and it's down to the consumer to vote with their wallet rather than making poor purchase decisions (like this lineup here given pricing is appalling for what you're getting against last gen) not to even mention Zen3 is due soon.........
 
You do realize the arch is being refreshed but the node it's on isn't a refresh? 14nm - 14nm+ alone was a 12% increase in drive current on the node alone whilst running 52% less power. The 10600K is still a better buy than a 3900X for pure gaming given the fact it can overclock without the issues of degrading (and actually gain performance from doing so)
Here's a nice video on that:
If those 14nm is so new why it consuming so much power??
What is the storage performence of those processor?? Windows Defender performence??
My system becomes unsable if I ran Windos defender on my C drive. And my ssd is older sata with less I/O, not newer PCI-e NVME with higher I/O.
Do you Intel fans have nothing but games to play on your computer??
And in some cases, it can outperform a 3950X - whilst costing a ton of money less (yes even the cooler and board that people seem to scapegoat to when in desperation searching for something to point their finger at).
A page from Nvidia's fanboy argument : Intel processor may cheaper than a 3950X, but 10600K will be more expensive on long run with those excesive heat and electric bill.

Pretty much this, one day people will realize a business is only there to make money and it's down to the consumer to vote with their wallet rather than making poor purchase decisions (like this lineup here given pricing is appalling for what you're getting against last gen) not to even mention Zen3 is due soon.........
Sadly Intel fanboy forget that everytime.
 
I am waiting for the reviews. I wouldn’t be surprised if these chips actually hold higher sustained base clock on auto like compared to the old models did. I am sure they have made some improvements with process to benefit these a little.
 
If those 14nm is so new why it consuming so much power??
It's consuming less than a 3950X whilst outperforming it in games, is that not enough for you? Not to mention it's only $280 against $750.
1592330454256.png

10600K pulls less than a 3900X stock to stock and outperforms it significantly and you mention "power"?
What is the storage performence of those processor?? Windows Defender performence??
My system becomes unsable if I ran Windos defender on my C drive. And my ssd is older sata with less I/O, not newer PCI-e NVME with higher I/O.
Do you Intel fans have nothing but games to play on your computer??
PCIe 4.0 controllers are useless given Q1T1 performance is appalling, latency isn't great for the drive either and it's just nothing but "MOAR SEQUENTIAL" which can be achieved for similar pricing using raid. So what "I/O"? :roll:
Future proof memers are a joke, take the 3770K for example for being "PCIe 3", why aren't you using that after all it's an excellent platform to pair a 2080ti with!!!!! :roll: can literally guarantee people would've upgraded off that platform by the time it's actually relevant or used to its complete potential, and it ends up being useless by the time it's utilised......
A page from Nvidia's fanboy argument : Intel processor may cheaper than a 3950X, but 10600K will be more expensive on long run with those excesive heat and electric bill.
Really? The 3950X pulls more power if anything... whilst providing less gaming performance.
Sadly Intel fanboy forget that everytime.
Quite ironic, isn't it?
 

Attachments

  • 1592330405782.png
    1592330405782.png
    121.5 KB · Views: 113
It's consuming less than a 3950X whilst outperforming it in games, is that not enough for you? Not to mention it's only $280 against $750.
View attachment 159245
10600K pulls less than a 3900X stock to stock and outperforms it significantly and you mention "power"?
You only saw the gaming performence, but what about other performence? And security holes?? Performence degradation due to security patch??
Looks like you intel fanboys have nothing but games to talk about.

PCIe 4.0 controllers are useless given Q1T1 performance is appalling, latency isn't great for the drive either and it's just nothing but "MOAR SEQUENTIAL" which can be achieved for similar pricing using raid. So what "I/O"? :roll:
Future proof memers are a joke, take the 3770K for example for being "PCIe 3", why aren't you using that after all it's an excellent platform to pair a 2080ti with!!!!! :roll: can literally guarantee people would've upgraded off that platform by the time it's actually relevant or used to its complete potential, and it ends up being useless by the time it's utilised......
3770K supports PCI-e 4.0?? Didn't knew that. o_O
And if 3700K dont bottleneck 2080Ti then why buy 10600K, when board for that cpu is cheaper.
 
You only saw the gaming performence, but what about other performence?
This CPU is designed for gaming and light threaded workloads, not really blender and multitasking, so buy AMD instead there.
And security holes?? Performence degradation due to security patch??
This is why something called hardware-level mitigations exist.
They minimize the impact of patching said vulnerabilities.
3770K supports PCI-e 4.0?? Didn't knew that. o_O
No, but it clearly supports PCI-e 3.0, and how many people have kept their 3770K platform today? Same applies to PCI-e 4.0 ; )
Looks like you intel fanboys have nothing but games to talk about.
Looks like fanboys like you do enjoy avoiding answering my main points, what happened to "muh intel pulls more power in games" and somehow racks up $470 in power to reach the cost of a 3950X and exceed that... whilst pulling 30W or so less? Or are you nothing but a low quality troll?
 
Performance are expected, but I'm quite surprise with the price. In the end I'll pick regular 3600 :D
 
This CPU is designed for gaming and light threaded workloads, not really blender and multitasking, so buy AMD instead there.
Blender/Adobe is a benchmark, but games not. :confused:

This is why something called hardware-level mitigations exist.
They minimize the impact of patching said vulnerabilities.

[/QUOTE]
Minimal empact. https://www.phoronix.com/scan.php?page=news_item&px=RdRand-3-Percent
No, but it clearly supports PCI-e 3.0, and how many people have kept their 3770K platform today? Same applies to PCI-e 4.0 ; )

Looks like fanboys like you do enjoy avoiding answering my main points, what happened to "muh intel pulls more power in games" and somehow racks up $470 in power to reach the cost of a 3950X and exceed that... whilst pulling 30W or so less? Or are you nothing but a low quality troll?
[/QUOTE]
o_O.
 
Do you even read? Or do we need to send you a video on how to do so?
Hardware mitigation minimize the impact of discovered patches - can't patch something that was only discovered recently after the processor was made unless you make newer steppings...........
Is this all you have to say? So clearly we've been reading power consumption wrong, less power apparently means a higher power bill then. :roll: Clearly logical, just like your points.
In case you need the definitions of "minimize" and "mitigations", here you go! Free of charge:
1592333565634.png


1592333495040.png

Examples of mitigation here:
1592333507822.png

If any of these words are too big for you to understand, please do ask for further assistance I can google for you if you'd like, since you're seemingly incapable of doing so yourself.
 
Back
Top