AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

[DragonQ]

Anandtech also apparently reviewed not using the latest windows build, which means features like fast time from low clocks to idle aren't enabled along with the newer scheduler. As with most releases some fairly lame stuff going on that would never be gotten away with if an Intel chip was launching.

I don't get it, how can professional reviewers not be on the ball with this? If I was doing a review, the first thing I'd check is that EVERYTHING was up to date (unless there was a specific reason not to, e.g. incompatible software, or a specific BIOS I've been told to use) and then list down EVERYTHING I am using: hardware, BIOS settings, software versions, etc. That way at least if I did something dumb/wrong, people could tell me about it immediately.

I suppose it's because they want to do apples-to-apples comparisons without having to rerun older tests, but I'd much prefer reviewers do a smaller subset of actual up-to-date, relevant tests, than a load of dodgy ones!

 

[Gibbo]

It'll do 5.4 with enough juice but I run it at 5 for daily use.

I do a bit of VMware workstation stuff when I want to keep it away from my vSphere lab but not really enough to justify more cores.

If you do not need more cores, pointless, nothing can really beat an 8086k in games.

 

[Yaayuh!]

I don't get it, how can professional reviewers not be on the ball with this? If I was doing a review, the first thing I'd check is that EVERYTHING was up to date (unless there was a specific reason not to, e.g. incompatible software, or a specific BIOS I've been told to use) and then list down EVERYTHING I am using: hardware, BIOS settings, software versions, etc. That way at least if I did something dumb/wrong, people could tell me about it immediately.

I suppose it's because they want to do apples-to-apples comparisons without having to rerun older tests, but I'd much prefer reviewers do a smaller subset of actual up-to-date, relevant tests, than a load of dodgy ones!

They can only do the best with what they have at hand at the time with the products, drivers, BIOSs and info. It's not uncommon for many things to only come to light AFTER the reviews are released.

 

[NexxCat]

@Gibbo I know you're likely rushed off your feet right now, so apologies for asking, but any word on 3700x stock? Is the incoming stock sufficient to cover the pre-orders? I'm torn between placing a pre-order or just giving up on the whole upgrade for a month =(

 

[Oblivion007]

Another price bump on the Taichi at the other retailer. OCUK now cheaper on it.

 

[Orcvader]

I don't get it, how can professional reviewers not be on the ball with this? If I was doing a review, the first thing I'd check is that EVERYTHING was up to date (unless there was a specific reason not to, e.g. incompatible software, or a specific BIOS I've been told to use) and then list down EVERYTHING I am using: hardware, BIOS settings, software versions, etc. That way at least if I did something dumb/wrong, people could tell me about it immediately.

They've been dropping the ball for a couple of years now, they're not as thorough as they used to be. I used to depend only on them back then but now I have to check a couple of other reviews just to compare.

Still, they're redoing the tests now in light of a new BIOS that fixes the boosting issues.

 

[pmc25]

A lot of these reviews are a total mess. From what I can see, most of the reviews which disclose BIOS version, and the only group test of mobos (eTeknix) was, for example, using a BIOS that's 4 versions behind current on the Taichi. It's similar for others.

Most of these reviews are not on optimised BIOS. Not on AGESA 1.0.0.3 (2 instead). Not on new chipset drivers.

They're not even (mostly) on BIOS that was current a week or even 2 weeks ago.

 

[Gibbo]

Might just be MSI related but one things for sure, it's a mild **** show across the board

We have most of them in stock, is there a problem with the MSI boards?

 

[AStaley]

Anandtech also apparently reviewed not using the latest windows build, which means features like fast time from low clocks to idle aren't enabled along with the newer scheduler. As with most releases some fairly lame stuff going on that would never be gotten away with if an Intel chip was launching.

According to this page they were using 1903 for the Ryzen 3000 chips, although they are rechecking their results in light of the revised firmware update.

https://www.anandtech.com/show/14605/the-and-ryzen-3700x-3900x-review-raising-the-bar/5

 

[Armageus]

I don't get it, how can professional reviewers not be on the ball with this?

Because AMD are likely not quite as stringent in their requirements supplied with a review kit. Intel historically supply a very strict, "you must use X Y Z settings, on these firmware/driver versions, and only run X Y Z benchmarks" in a way to show the Intel Product in the most positive light.

 

[Jokester]

A lot of these reviews are a total mess. From what I can see, most of the reviews which disclose BIOS version (and the only group test of mobos - eTeknix) was, for example, using a BIOS that's 4 versions behind current on the Taichi. It's similar for others.

Most of these reviews are not on optimised BIOS. Not on AGESA 1.0.0.3 (2 instead). Not on new chipset drivers.

They're not even (mostly) on BIOS that was current a week ago.

Why are AMD so bad at this, it’s a PR disaster not to have final product being reviewed, especially when performance is affected. From the jaws of victory.

 

[Dirk Diggler]

With 3900x having same TDP as 2700x is it safe to assume that my current cooling solution will perform pretty much same? Also, I'm trying to understand AMD's marketing here: the single core boost of 4.6GHz - is that guaranteed at stock BIOS settings?

 

[drunkenmaster]

They've been dropping the ball for a couple of years now, they're not as thorough as they used to be. I used to depend only on them back then but now I have to check a couple of other reviews just to compare.

Anandtech aren't the same old Anandtech any more, people need to get that. Anandtech, as in the guy behind it, sold and it's now run by the people who own Tom's hardware. Media conglomerates buying up smaller news outlets then pushing the same branded news under new supposedly independent names has been going on for donkeys years in all arenas of information. Anandtech as they are run now are still better than Toms, but there has been a huge drop off in quality in recent years.

Though I think part of the reason here is Anandtech's move to their "bench" thing and doing the kinda lazy testing of, lets update every old chip slowly throughout the year but also test new products with the same settings even if that means missing out on key updates and changes.

The issue is here AMD said very plainly at the announcement what, a month or more ago, how useful the windows update would be. They've had a month to update to the new windows and test the 9900k and everything else relevant they want to compare then test Zen 2 when it comes in with the same updates enabled, instead they've gone the lazy way and tested new stuff with old windows builds that misses out on important features of new chips.

But like I said, if this were an Intel launch Anandtech would have done it the other way around, spent the last month retesting everything else up to new settings then benchmarked the new chips when they came in.

 

[opethdisciple]

From the techspot review concerning the x570 platform....

"For the most part we feel X570 boards -- particularly the more premium models -- should only be paired with the 3900X or the upcoming 3950X. For those grabbing the 3700X or the more affordable the Ryzen 5 3600, we’d recommend existing X470 or B450 boards. We’ve heard the concerns regarding X570 board's pricing, but know you can fully unleash Zen 2 processors on B450 and probably even B350 boards, which we’ll set about testing shortly."

 

[pmc25]

Why are AMD so bad at this, it’s a PR disaster not to have final product being reviewed, especially when performance is affected. From the jaws of victory.

AMD? They can't force reviewers to use newer BIOS. For the Taichi there are 3 newer public versions (all published online) over a period of 3 weeks. This isn't about AMD or the mobo maker not furnishing reviewers with BIOS in time.

I doubt either AsRock or AMD are thrilled about this. Some of the Gigabyte BIOSes appear to be even older.

Why they are doing this, I don't know. But I could guess.

AGESA 1.0.0.3 was out for most X570 boards 3-6 days ago. They're typically using versions 1-2 prior to that. It's ridiculous.

 

[ChrisD.]

If you do not need more cores, pointless, nothing can really beat an 8086k in games.

Cheers that's my own conclusion. I'd guess by the time which 8 cores and above pay dividends in games CPU architecture will have moved beyond what Zen 2 is capable of anyway. Having been GPU bound my GPU itch is somewhat calmed since my 2080 Ti purchase.

 

[CAT-THE-FIFTH]

So Anandtech just posted a twitter message saying they are redoing their CPU tests because of a new bios showing very different boost behavior.

Have you a link to the post please?

 

[Zernath]

Why they are doing this, I don't know. But I could guess.

Brown envelope syndrome most likely.

 

[pmc25]

Anandtech aren't the same old Anandtech any more, people need to get that. Anandtech, as in the guy behind it, sold and it's now run by the people who own Tom's hardware. Media conglomerates buying up smaller news outlets then pushing the same branded news under new supposedly independent names has been going on for donkeys years in all arenas of information. Anandtech as they are run now are still better than Toms, but there has been a huge drop off in quality in recent years.

Though I think part of the reason here is Anandtech's move to their "bench" thing and doing the kinda lazy testing of, lets update every old chip slowly throughout the year but also test new products with the same settings even if that means missing out on key updates and changes.

The issue is here AMD said very plainly at the announcement what, a month or more ago, how useful the windows update would be. They've had a month to update to the new windows and test the 9900k and everything else relevant they want to compare then test Zen 2 when it comes in with the same updates enabled, instead they've gone the lazy way and tested new stuff with old windows builds that misses out on important features of new chips.

But like I said, if this were an Intel launch Anandtech would have done it the other way around, spent the last month retesting everything else up to new settings then benchmarked the new chips when they came in.

It's worse than that. Ian Cuttress, who is decent, but does seem to kowtow to whatever the ownership's agenda is, took time off for this. A couple of minions did that Ryzen testing ... and it's the worst, least detailed, most opaque testing on Anandtech for a major product launch ever ...

... and this is the biggest product launch in 10 years.

 

[Jokester]

AMD? They can't force reviewers to use newer BIOS. For the Taichi there are 3 newer public versions (all published online) over a period of 3 weeks. This isn't about AMD or the mobo maker not furnishing reviewers with BIOS in time.

I doubt either AsRock or AMD are thrilled about this. Some of the Gigabyte BIOSes appear to be even older.

Why they are doing this, I don't know. But I could guess.

AGESA 1.0.0.3 was out for most X570 boards 3-6 days ago. They're typically using versions 1-2 prior to that. It's ridiculous.

AMD need to be putting finished hardware in reviewers hands, we already know there’s issues even with current BIOS’.

It has the same rushed out the door feel of the original Zen launch.