Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AMD Ryzen 3000 announced

anandtech.com

Post reply on HN
Print view View original

71–80 of 347 posts

Re: AMD Ryzen 3000 announced

#71

Post by gratilup »

danbolt wrote: post #57
gratilup wrote: post #43

Now imagine a Threadripper with the Zen2 cores, higher IPC and frequency would be certainly welcome. Have the 32 core 2990WX and it's an incredible CPU for compiling large C++ programs, running big test suites and never having to worry about running too many tasks at the same time.

A lot of AAA game studios with larger C++ codebases use Incredibuild a lot. I can imagine having something with this level of parallelism would be incredibly useful.

Don't even need Incredibuild, even MSBuild or just plain /MP option in VC++ can take advantage of it. A build from VS of the Unreal Engine 4 client takes around 2 min, for example.

Re: AMD Ryzen 3000 announced

#72

Post by snvzz »

techntoke wrote: post #63

Nothing mentioned about their "G" series.

They already launched a new generation of that recently, based on Zen+ with Vega. The previous one was pre-zen+.

Their gpu+cpu chips are always based on the more mature tech.

Re: AMD Ryzen 3000 announced

#73

Post by sq_ »

snvzz wrote: post #58
sq_ wrote: post #35

Earlier quoted context omitted.

Isn't that what AMD GPUs are for? They'd be eating their own market share for PCs that double as frying pans/ovens!

If you're basing that comment on TDP, think again. NVIDIA redefined TDP to their convenience, to mean something more like "averages". So their numbers can't be compared directly. Always look at third party measurements. I'm looking forward to Navi's, while on the topic, as they've announced large improvements in power efficiency.

In response to what you said about Navi, I'm also excited about that. I'd really love to see some really strong competition to Nvidia's offerings that might be able to give them a good fight.

Their dominance and tendency to push for proprietary features seems quite bad for the industry as a whole.

Re: AMD Ryzen 3000 announced

#74

Post by eropple »

shmerl wrote: post #50
DiabloD3 wrote: post #27

I watched the keynote live. I'm sold, my desktop is most likely going to be a Ryzen (although not the 8/16 monster, come on, it's a desktop, if I need high core count, I have stuff at work for that).

I don't mind 12 cores / 24 threads on my desktop. Makes compiling Mesa, Wine and Linux kernel a lot faster.

I do a bunch of live video stuff (which, for the uninitiated, means "hi, I'm ffmpeg, I want all your processors") and I learned, the hard way, that I couldn't put a 16C/32T Threadripper in the box--it's a 4U case in a Gator transport case--because 180W would blow out my heat budget. 12C/24T at 105W should be doable. I'm excited; right now I'm constrained because stuff like my CG system is built on Chromium and so it gets a little chewy when I run multiple independent overlays. Doubling my cores--I have a Ryzen 1600 in there right now--and probably increasing single-thread performance by 40% should make my life pretty interesting.

Re: AMD Ryzen 3000 announced

#75

Post by wtallis »

DuskStar wrote: post #45
klingonopera wrote: post #42

Earlier quoted context omitted.

A modular approach! Kudos to AMD for making it work nicely (anyone remember the first Pentium Ds, that were just two Pentium 4s clobbered together to make a dual-core chip?). Apart from that, I think you meant two -and-a-half, not one?

Nope, one-and-a-half! Two of the CPU dies, one interconnect die - that's going to be somewhere between one and two! And their interconnect tech seems like it's going to be of huge importance in the server space - I can only imagine the yields on 8x4 core modules will be far higher than on a monolithic 32 core chip.

> I can only imagine the yields on 8x4 core modules will be far higher than on a monolithic 32 core chip.

The next EPYC's going to be up to 8 chiplets x 8 cores, competing against Intel's current 28-core monolithic die (or their dual-die non-socketed 56-core). How many cores remain enabled on a hypothetical next-generation Threadripper is an open question, but they would probably go beyond 32 cores total. And a 32-core Threadripper would probably not have 8 chiplets but rather four fully-enabled active chiplets and four mechanical spacers.

Re: AMD Ryzen 3000 announced

#76

Post by m0zg »

coder543 wrote: post #70
m0zg wrote: post #40

Earlier quoted context omitted.

Could be that it'll still continue to have stability issues. For some reason Ryzens are extremely picky when it comes to memory. There's really no guarantee that 3-rd gen will resolve this issue. It's to the point where e.g. Corsair makes memory specifically designed to work with AMD CPUs. This memory typically contains Samsung B-die chips which work fairly well.

No, the stability issues they're talking about are on a release day 1800X. It's not memory issues, and I've never had memory issues with either of my Ryzen processors. This is the issue they're certainly talking about: https://www.phoronix.com/scan.php?page=news_item&px=Ryzen-Se... And it is assuredly not a problem with any Ryzen processors manufactured after the very first few months.

So I must be imagining my problems with Threadripper memory kits as of early this year as well as my coworkers problems with his 1800X. Hmm, OK then. HN knows best.

Re: AMD Ryzen 3000 announced

#77

Post by gigatexal »

Yup. Just waiting on the benchmarks from independent reviewers and to see how XFR in this generation works but I’ll be getting a Ryzen 9 if everything checks out. 24 threads will be amazing for local development (microk8s, and others) when I’m not gaming and save me from having to build a separate box.

Re: AMD Ryzen 3000 announced

#78

Post by kalessin »

m0zg wrote: post #40
lettergram wrote: post #28

Finally! I’ve been using a Ryzen 1800x since it’s release. Unfortunately, it has some stability issues and I’ve been waiting to upgrade on the Ryzen 3000, 7nm line. This is going to be a solid 75%+ boost to performance, given I regularly max out my machines threads. Pretty amazing improvement in 2 years.

Could be that it'll still continue to have stability issues. For some reason Ryzens are extremely picky when it comes to memory. There's really no guarantee that 3-rd gen will resolve this issue. It's to the point where e.g. Corsair makes memory specifically designed to work with AMD CPUs. This memory typically contains Samsung B-die chips which work fairly well.

I had troubles with my memory (4*16GB with a 2950X) until I realized it was configured in 1T command rate by default. Things appear to be rock stable in 2T. Anyway, yeah, memory is more tricky than it should on Ryzen.

Re: AMD Ryzen 3000 announced

#79

Post by phonon »

sadris wrote: post #20
p1mrx wrote: post #19

It will be interesting to see whether they can match Intel in single-threaded performance across the board, and not just some carefully-selected benchmarks. This would be the first time since the Core2/Athlon64 days.

I'm begging for a good single threaded CPU that isn't $600

Try the i3-9350K or i5-9600

https://ark.intel.com/content/www/us/en/ark/products/186606/...

https://ark.intel.com/content/www/us/en/ark/products/134900/...

Re: AMD Ryzen 3000 announced

#80

Post by chaosbutters314 »

empyrical wrote: post #68
gratilup wrote: post #43

Now imagine a Threadripper with the Zen2 cores, higher IPC and frequency would be certainly welcome. Have the 32 core 2990WX and it's an incredible CPU for compiling large C++ programs, running big test suites and never having to worry about running too many tasks at the same time.

Conjecture on my part, but I wouldn't be surprised if we also see a 64 core Threadripper - there's going to be a 64 core Epyc: https://wccftech.com/amd-7nm-epyc-64-core-32-core-cpus-specs... Although they may save it for a "Zen2+" or something similar, like they did with 32 core Threadripper

I'm looking forward to dual socket 64 core epyc for at work (128 cores)
Post reply on HN