londiste300 and 400 series chipsets are almost identical.

Except for the size of the bios chips.

MakaveliExcept for the size of the bios chips.

This entire AM4 BIOS size thing is not as simple as different BIOS chip size. There are a lot of 400-series motherboards with 128Gb/16GB BIOS chips. Hell, there are 500-series boards with 128Gb/16GB BIOS chips (Gigabyte boards mainly if I remember correctly). At Zen2 launch there were some more technical details revealed that claimed issue is with how older AM4 CPUs are able to read only 16GB of BIOS. Workarounds exist but these are not exactly fun.

There are different BIOSes and some (software/firmware) feature differences but 300 and 400 series chipsets themselves are basically identical.

I have a good enough B450 board that I would really like to get a Ryzen 5000 for but I am not holding my breath and January or after does not sound too encouraging.

This is like reading press releases from white house press office .....


1. After weeks of complaining from AMD fans about launches that are not launchers because most vendors are out of stock, now suddenly a product announce,ents is a "release" with NDAs sill in effect and not a single product sold.

2. When ya read a sentence, all the words count ....

Average of 7% faster in 1080p gaming across ****select**** game titles than the competition

And if we let Intel ***select*** the games, is that going to hold ?

3. Will the argument still be "bang for the buck".... or what's the value of the $799 5950 CPU versus the $180 10400F ?
I'm gonna wait till the game is played (TPU Reviews) before declaring the winner.

birdieAMD fans never fail to disappoint with double standards. Intel and NVIDIA have always been "evil" but once AMD does that, suddenly it's perfectly fine because they just follow suit.

What are you even talking about? It was never about the pricing. Intel was stifling innovation for years and giving us 5% performance increases with the same 4 cores year after year because they knew they could get away with it. How is this the same as what AMD is doing at the moment? We're seeing core counts literally explode with still good single core performance gains. The price was never the issue with Intel, so don't pull that straw man. It's about how they could never justify the price in any way. AMD is doing literally the opposite right now. I'll take my 12 cores instead of 4, thank you. With just Intel, we would still have 4.

I'm waiting for actual independent tests, but considering AMD and Su have been honest about performance, and they literally fixed (if they are to be believed) the latency issue with caches, meaning IPC must increase.

Something I noted is that Boost speed and core count is inverse compared to the competitors approach of higher core count lower speeds.

Caring1Something I noted is that Boost speed and core count is inverse compared to the competitors approach of higher core count lower speeds.

They take a higher bin die and that extends the range to which the cpu can stretch its limits within the unit boundaries. Happened with bulldozer, too. 8370e was better than 9590 since it wasn't leaky and all it took for the user was to keep temperatures in check. Higher bins are higher resisting cpus. What is essentially a good iron press isn't necessarily a good soldering iron and visa versa.

Predictably, they've increased IPC and reduced latency enough to compete with Intel at their strongest point, gaming at 1080p (or older single core apps). There might be some exceptions in terms of specific software, there always will be. Generally, I do trust the benchmark graphs and expect the reviews on launch night to be impressive. The price is a little high, but one can always wait. That also helps with getting a better bin. There won't be much OC headroom but a better bin means lower voltage, lower temps, less fan noise.

I would have preferred they treat the motherboards and BIOS the same way they did Zen 2. There's no real reason why they couldn't have. Like some others have hinted above, the whole thing in the last six months has been contrived. 16MB vs 32MB BIOS chips etc. It's an artificial limitation. There's plenty ways to make it work and a couple X370 boards that are just as good as X470 boards if not better. I guess we should be thankful that they are allowing this at all, because Intel wouldn't be. AMD is still way more on the "pro consumer" side of things, although this is slowly starting to shift now. All CPU's are still unlocked and all motherboards are unlocked. Anyways, a lot of it is in the board makers hands, and they all compete with each other (for the best reputations of support). It's in their best interest, that if one board maker does a good job, they all have to. So let's see what happens. The motherboard side of things has been a bit of a mess in the last 18 months.. chipset fans, B550 released 12 months later (and in some cases being better than X570). Apparently ASUS are releasing a new X570 board (a premium board) without a chipset fan, so there you go. But the X570 chipset silicon itself is inefficient (idle wattage), and it won't be fixed until the switch to AM5 and DDR5. As there will be no X670.

Anyways, the good thing is the IO die is the same. The CPU mostly runs the same, it's only the core chiplets themselves that are new, so in terms of BIOS and overclocking and motherboards and the like.. it should be a pretty smooth transition into a 5000 series CPU. There's further optimizations in terms of controlling or customizing how the cores behave from the BIOS. More expensive prices.. but also the best CPU's that AMD have ever made. Energy effeciency, multi core, single core.. everything. It's possible that the RAM latency (measured in AIDA) will still be higher than Intel, but the way the cache is now structured.. it seems like it won't make much difference. And the CPU's will have enough raw performance to compensate for this. The single core score in Cinebench is super impressive. The fact that these CPU's have this level of performance with 16 cores is super impressive. The energy efficiency and power consumption, also very impressive. The lower clock speeds (vs 5.2 Ghz) are actually an advantage in some ways.. in terms of equal or better performance with less wattage/heat/noise. I was a little worried that the power consumption would go up (like it did from 1700X to 2700X) but apparently, it's fine.

The prices are a little high for my liking, especially internationally, when you add GST on top of it.. but this is AMD making a statement. "We are premium now.. we are Intel, we are NVIDIA" etc. Budget versions like 5700X or 5600 (non X) will be available later on. And the future is bright as well, the 2nd or 3rd iteration of CPU on the new AM5 platform, once DDR5 has had a chance to be optimized and is a bit cheaper (with high performance speeds).. is going to be super fast. If the software can catch up.. being coded to take advantage of super fast nvme, plus lots of cores.. and RAM that has VRAM speeds, it's almost like a new world of computing is starting to open up.

John NaylorThis is like reading press releases from white house press office .....


1. After weeks of complaining from AMD fans about launches that are not launchers because most vendors are out of stock, now suddenly a product announce,ents is a "release" with NDAs sill in effect and not a single product sold.

2. When ya read a sentence, all the words count ....

Average of 7% faster in 1080p gaming across ****select**** game titles than the competition

And if we let Intel ***select*** the games, is that going to hold ?

3. Will the argument still be "bang for the buck".... or what's the value of the $799 5950 CPU versus the $180 10400F ?
I'm gonna wait till the game is played (TPU Reviews) before declaring the winner.

I believe ZEN3 event was a product announcement and not a launch. Nov5 is the launch.
If I misunderstood your statement and you were trying to say something else I’m sorry.

As for the “selected” games a lot of people saying that AMD deliberately picked games (some of them) that “traditionally” was not doing well as opposed to Intel. They even show a loosing-to-Intel one.

The value of these 4 CPUs can be a subjective matter. Don’t forget that these are only the high binned X SKUs. It could be the case that yields are so good now on the mature 7nm node that most of the chiplets are higher binned than previous gen, and so the don’t have enough low binned chips to launch the nonX along with the others. And they still have available ZEN2. Users should wait and not buy them if the don’t see value on them.

On the other hand some others may see value on them.
For instance the 6core/12threaded 300$ 5600X according to IPC claims, speed and performance/watt improvements, will probably be faster in ST/gaming from the higher ZEN2 and faster or equal in all-core loads from the 8core/16threaded 3700X. If that turns out to be the case, the 5600X has more value than the 3700X ever had. And it’s 30$ less MSRP from 3700X with gaming perf higher than a 700$ ZEN2. Is that something or what?

PowerPCWhat are you even talking about? It was never about the pricing. Intel was stifling innovation for years and giving us 5% performance increases with the same 4 cores year after year because they knew they could get away with it. How is this the same as what AMD is doing at the moment? We're seeing core counts literally explode with still good single core performance gains. The price was never the issue with Intel, so don't pull that straw man. It's about how they could never justify the price in any way. AMD is doing literally the opposite right now. I'll take my 12 cores instead of 4, thank you. With just Intel, we would still have 4.

The difference is that AMD had no choice. They were struggling to be competitive. AMD is to blame as much as Intel. If your goal is to be the best and you become the best do think you'd just keep getting better and better without any competition whatsoever? I think that's what Intel's thought process was . Their problem was that they got too comfortable and now they're in a bad spot. The story of the tortoise and the hair is a perfect example of what has happened here between Intel and AMD.

Intel got themselves into a situation where they continued to milk the cow of their previous success and dominance.. but the problem is.. it sort of set them up for a dead end. They got stuck.

They continued on the 14nm process.. iteration after iteration, and they refined it REALLY well, but the problem is.. if they go down to a lower nm process 10nm or lower.. the clock speed of the new CPU will be less. The performance in gaming at 1080p would be less. You can't release a new CPU if it has LESS performance than the previous CPU. Right? They focussed on a priority that would continue to give them an advantage in the short term, but hold them back from moving forward.

If your main advantage is gaming at 1080p? And your new CPU has LESS performance. No one would want to buy it. Especially if you charge high prices. They had refined the 14nm process so much to the extent that it can clock over 5Ghz. Any new process (10nm) would clock lower than that on it's first release. It might take a year or two until you can refine the new node to produce higher clock speeds to compete with the refined node of the previous generation. But clock speeds are their biggest advantage? See the problem?

There needed to be better future planning and forethought. Innovating, rather than complacency. They left themselves vulnerable, and created a situation where they are better off staying where they are.. in the short term, but in the long term it's a dead end and creates future pain. There needs to be some short term pain and some refining until they can recover. They can recover eventually, but there needs to be a change to their approach. A more thought out approach, thinking less about the past and more about the future.

The competititon has caught them out and exposed them, where as in GPU terms.. NVIDIA is staying ahead of the game and aren't being caught out in the same way. AMD might catch up to NVIDIA, but they aren't letting them get ahead.. because they are continuing to push forwards.

CubanBIntel got themselves into a situation where they continued to milk the cow of their previous success and dominance.. but the problem is.. it sort of set them up for a dead end. They got stuck.

They continued on the 14nm process.. iteration after iteration, and they refined it REALLY well, but the problem is.. if they go down to a lower nm process 10nm or lower.. the clock speed of the new CPU will be less. The performance in gaming at 1080p would be less. You can't release a new CPU if it has LESS performance than the previous CPU. Right? They focussed on a priority that would continue to give them an advantage in the short term, but hold them back from moving forward.

If your main advantage is gaming at 1080p? And your new CPU has LESS performance. No one would want to buy it. Especially if you charge high prices. They had refined the 14nm process so much to the extent that it can clock over 5Ghz. Any new process (10nm) would clock lower than that on it's first release. It might take a year or two until you can refine the new node to produce higher clock speeds to compete with the refined node of the previous generation. But clock speeds are their biggest advantage? See the problem?

There needed to be better future planning and forethought. Innovating, rather than complacency. They left themselves vulnerable, and created a situation where they are better off staying where they are.. in the short term, but in the long term it's a dead end and creates future pain. There needs to be some short term pain and some refining until they can recover. They can recover eventually, but there needs to be a change to their approach. A more thought out approach, thinking less about the past and more about the future.

The competititon has caught them out and exposed them, where as in GPU terms.. NVIDIA is staying ahead of the game and aren't being caught out in the same way. AMD might catch up to NVIDIA, but they aren't letting them get ahead.. because they are continuing to push forwards.

I can agree with the general idea or the post, about Intel and nVidia.
One thing though. Yes shrinking nodes does prevent high clocks for starters. But, engineers can do wonders. If they come up with a really nice architecture with a nice uplift of IPC and performance/watt, loosing 200~400MHz of clock would be less significant. Sure, the more they delay the more difficult will be to catch up, and to be honest so far they're not too far behind in performance. In performance/watt its a very different story and thats the node to blame.
They want to think ahead but they cant when stuck on the same node for ages. Probably they were preparing architecture(s) for 10nm but the node gone all wrong and when they realize it, time has passed and they needed to reschedule for 14nm... again. They were preparing 10nm fabs and then they were forced to turn them back to 14nm. Its a mess.
Seems like they will not get out of this any time soon... before 2022-23, and AMD has build up momentum (see 5nm and ZEN4/5).

To be honest, that is not good for us as users. Not good at all!

And so, reality slowly intrudes:

wccftech.com/asus-not-supporting-amd-ryzen-5000-cpus-on-x470-motherboards-suggests-x570-b550-upgrade/

RandallFlaggAnd so, reality slowly intrudes:
wccftech.com/asus-not-supporting-amd-ryzen-5000-cpus-on-x470-motherboards-suggests-x570-b550-upgrade/

This is probably either ASUS Support guy reading something wrong, wanting to get rid of annoying question or just making things up.

Th3pwn3rThe difference is that AMD had no choice. They were struggling to be competitive. AMD is to blame as much as Intel. If your goal is to be the best and you become the best do think you'd just keep getting better and better without any competition whatsoever? I think that's what Intel's thought process was . Their problem was that they got too comfortable and now they're in a bad spot. The story of the tortoise and the hair is a perfect example of what has happened here between Intel and AMD.

Oh the story, what happened between Intel and AMD, reach far into the history my friend than just recent events. If you are a dominant company you want it to stay that way. If you don't innovate and you refurbish old tech you get loads of money. If you think that Intel bring something spectacular to the table then you are mistaken. After AMD hit it hard with the core and close enough performance, Intel started to do something but that doesn't mean you don't need to innovate.
I think you are wrong. The innovation is not about competition but the attitude and/or stand the company has. That is why a lot of people hate Intel for this. Instead of innovating they were selling same refurbished processors over and over just to get money.

ratirtOh the story, what happened between Intel and AMD, reach far into the history my friend than just recent events. If you are a dominant company you want it to stay that way. If you don't innovate and you refurbish old tech you get loads of money. If you think that Intel bring something spectacular to the table then you are mistaken. After AMD hit it hard with the core and close enough performance, Intel started to do something but that doesn't mean you don't need to innovate.
I think you are wrong. The innovation is not about competition but the attitude and/or stand the company has. That is why a lot of people hate Intel for this. Instead of innovating they were selling same refurbished processors over and over just to get money.

I agree with the general idea.
Intel kept refurbishing CPUs but that mindset has changed after ZEN era. What happened to Intel is that they failed to refine 10nm node. I don’t know what exactly happened... they were too late to pick up? ...they have internal management issues? or what else...

The bottom line is that they now stuck at 14nm node... The upcoming 11th gen with (at last) new architecture was meant to be on 10nm and they forced to import it back to 14nm. They even had prepared 10nm fabs and then also forced to convert them back to 14nm.

It’s a mess...