AMD - Zen chitchat

[kompukare]

Looks like CCL have jumped the gun a bit by listing four R5s:

https://m.cclonline.com/category/401...Powered-CPUs/#
Judging from the current £300 for the R7 1700, those prices are about £10-£15 too high. All except the 1600X come with cooler (Wraith Spire for the 1600 and 1500X, and Wraith Stealth for the 1400).
Will be interesting if reviewers or early buyers can find out if these are binned similarly to the R7s or not. And for that matter, how much actual power the hex and quad use as the TPUs are 95W for 1600X and 65W for the others.
While the seem good value now that's only because due to GBP<>USD rate and so on, the Intel chips have gone up that much. I paid £160 for my i5-3570K and the equiv is now £240.

[Biscuit]

The 1700 still does look like the sweet spot if those prices are correct. I was expecting a little closer to the £200 mark for the top of the range R5 with the lowest end being around £150.

[watercooled]

It looks like they've just changed the $ for a £ with the prices. Hopefully they'll sell for less than that in most places - the Ryzen 7 CPUs cost less than the dollar price at least, that's probably why these R5's seem a bit expensive.

It's nice that at least the 1500X comes with a cooler, but IMO it would be nice if they offered all of the CPUs with at least the option of a cooler for an extra tenner or something. I've no experience with the Wraith coolers myself, but if they're as good as it seems, then what's wrong with just using that rather than a more expensive aftermarket cooler? At least compatibility is guaranteed and not everyone overclocks just because they're unlocked.

[scaryjim]

The 1700 still does look like the sweet spot if those prices are correct. ...

Only if you're overclocking: and while I accept that a lot of people will be, AMD need to target a broad spread of the market. The 1600X is half the price of the 1800X but almost exactly 3/4 of the chip - that's pretty stonking value. It'll also easily outpace the 1700 in lightly threaded tasks at stock, and should have around 85% of it's multi-threaded performance thanks to its higher base clock. And that's at around 85% of the price. Seems pretty sensibly priced to me.

If anything, you could argue that the 1400 is the overpriced one - it's pretty much exactly half a 1700, but it's significantly more than half the price. OTOH, if it has the same overclocking potential as the other Ryzen chips, a 4C/8T 4GHz chip @ £170 is still pretty damn tasty....

[Xlucine]

'overpriced' is relative. £170 will get you a top end i3 from intel at current exchange rates

[kompukare]

See none of the reviewers were happy about the time they had in their initial reviews.
HT4U.net:

That AMD's new CPU architecture would not be testable and presentable in three days was absolutely clear to us [from the onset]. And that after ten days we would find new insights was also expected. So today we are going to the second round with our Ryzen 7 1700 and 1700X reviews.

https://www.ht4u.net/reviews/2017/am...rneut_im_test/
Google Translated.
As usual they allow readers to access their database with their bench tool where you can pick two CPUs
https://www.ht4u.net/reviews/2017/am...st/index28.php

The also did memory testing but 3200 was the max the managed and only with this G.Skill DDR4-3600 kit.
Despite the reported problems with Ryzen and dual-rank memory, their test suite was actually quicker with dual-rank (by 1%) with the biggest difference (6.7%) being in Doom.

[kompukare]

See PCGH have been busy reviewing core and CCX scaling in games. Basically simulating Ryzen 5. I know they did a bit of that in their initial review but now the tried have more games:

• GTA5
• Ghost Recon:Wildlands
• Battlefield 1
• For Honor
• Rise of the Tomb Raider
• Watch Dogs 2

http://www.pcgameshardware.de/Ryzen-...aming-1224283/
Google Translated

They are also saying that R5 1500X is going to come with the full 16MB cache.
Think their Ryzen list is somewhat different to the ones we saw earlier.


Their benches tested both 2+2 and 4+0 and the 4+0 is consistently a bit quicker. Very close except in BF1 where the difference is a bit more.

[watercooled]

Tom's have done some more power measurements of Ryzen and included results from other CPUs for comparison. Again, it's really impressive especially given where they've come from, even comparing Ryzen's 8 core SoC against Intel's 7700k (even when the latter is underclocked).

http://www.tomshardware.co.uk/amd-ry...w-33843-8.html

[scaryjim]

Hmmm, Tom's have also been given access to a beta build of Ashes: Escalation with Ryzen optimisations:

http://www.tomshardware.co.uk/amd-ry...ews-55229.html

The 1800X gets over 15% boost at 1080p Crazy preset. I guess we can stop blaming the Windows scheduler now

[CAT-THE-FIFTH]

We will soon be distributing AGESA point release 1.0.0.4 to our motherboard partners. We expect BIOSes based on this AGESA to start hitting the public in early April, though specific dates will depend on the schedules and QA practices of your motherboard vendor.

BIOSes based on this new code will have four important improvements for you
We have reduced DRAM latency by approximately 6ns. This can result in higher performance for latency-sensitive applications.
We resolved a condition where an unusual FMA3 code sequence could cause a system hang.
We resolved the “overclock sleep bug” where an incorrect CPU frequency could be reported after resuming from S3 sleep.
AMD Ryzen™ Master no longer requires the High-Precision Event Timer (HPET).
We will continue to update you on future AGESA releases when they’re complete, and we’re already working hard to bring you a May release that focuses on overclocked DDR4 memory.

Seems some more updates will be arriving soon.

[DanceswithUnix]

I think some of the older kit is starting to sell out. Ebuyer had a third off the Athlon X4 845 a couple of days ago, no listing at all today. Was considering an AM1 board for a low powered ITX box, but the 5350 is hard too find and 50% more expensive than it used to be.

[watercooled]

Not sure if this has already been posted, but AdoredTV has another video where he investigates what's going on with Ryzen's performance in Tomb Raider:


It's really quite interesting and picks up on something that has perhaps been missed by the press due to AMD's lack of single-GPU competition to Nvidia's high-end - that Nvidia's DX12 drivers are currently lacking in some areas when it comes to CPU scheduling. Now, we've known for a while that Nvidia have had some room for improvement in their post-DX11 performance, but unfortunately Ryzen has almost exclusively been tested with Nvidia's GPUs so we're perhaps not seeing the whole picture.

I'm definitely looking forward to see how a Ryzen+Vega system performs, given AMD obviously have the access and motivation necessary to develop GPU drivers that are well-optimised for their CPUs.

[CAT-THE-FIFTH]

Not sure if this has already been posted, but AdoredTV has another video where he investigates what's going on with Ryzen's performance in Tomb Raider[/video]

It's really quite interesting and picks up on something that has perhaps been missed by the press due to AMD's lack of single-GPU competition to Nvidia's high-end - that Nvidia's DX12 drivers are currently lacking in some areas when it comes to CPU scheduling. Now, we've known for a while that Nvidia have had some room for improvement in their post-DX11 performance, but unfortunately Ryzen has almost exclusively been tested with Nvidia's GPUs so we're perhaps not seeing the whole picture.

I'm definitely looking forward to see how a Ryzen+Vega system performs, given AMD obviously have the access and motivation necessary to develop GPU drivers that are well-optimised for their CPUs.

Here is another test seeing the same thing:



But its not only Ryzen you see weirdness in:

http://forums.hexus.net/reader-revie...ml#post3781996

That is my review of the GTX1080 on a IB Core i7. Look at what I saw in the same area under DX12!!

Look at when I then tried an RX470 4GB under DX12:

http://forums.hexus.net/reader-revie...ml#post3783502

[watercooled]

I mean, I do understand that when you're reviewing for a website, you have probably hundreds of benchmarks to process, along with hardware changes, and having that much information in front of you means you're not as likely to spot weirdness which might seem like nothing at first glance. Independent reviewers and even users often pick up on things like this, and it would be nice if it got some more attention in the media, to better explain the results. I respect sites more for doing things like this than publishing a ton of robotic test results, not least because it demonstrates the reviewers have a better understanding of what they're actually seeing

I know a few like Hexus do this when the need arises, but it would be nice if more sites could take this sort of new information and run their own investigation into why some results are the way they are.

We've seen something similar happen with GPUs too, where in some cases we get benchmarks with release drivers, and that's it. 6 months down the line when you're looking to upgrade, those numbers could be completely invalid because of both software and driver updates, and few places re-run benchmarks (again, I do understand it's a time-consuming process so it's probably not practical to do it frequently). In a way, I suppose this is the sort of reason we get rebrand launches from the likes of AMD as it, in a way, forces sites to re-test the products with the latest drivers, and can show performance improvements even if nothing has changed on the hardware side.

[Terbinator]

Apparently MS are revealing some of the hardware details on XB1X this coming Thursday.

Accepted wisdom would be that Zen isn't in given what Lisa has had to say on custom APUs. However, some of the feedback from 'insiders' over on NeoGAF says hardware is impressive or better than was initially expected (up-clocked Jag/Puma).

[watercooled]

I've been wondering what the Xbox will use. Did Lisa Su rule out the possibility of Zen or was it more of a dismissal of the question? Either I've forgotten or I didn't watch it.