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

131–140 of 347 posts

Re: AMD Ryzen 3000 announced

#131

Post by vetinari »

ubercow13 wrote: post #111
WC3w6pXxgGd wrote: post #100

Earlier quoted context omitted.

What will you be doing that will actually use 24 threads?

make -j24

Be sure to have enough RAM, so you will not run out of it.

(For some reason, when building projects like LLVM with -j 16, 32 GB w/o swap may be not enough. With -j 2, it is enough, but takes an eternity).

Re: AMD Ryzen 3000 announced

#133

Post by tmd83 »

MrGilbert wrote: post #107
tmd83 wrote: post #105

I'm really looking forward to this. But one issue with current AMD cpu is that you have to buy a desktop GPU even when you are not doing any kind of gaming and such. I know intel iGPU haven't been terribly good but for work they are good enough and is one less part and cheaper to boot. For same performance a Ryzen 3rd gen + gpu might still be cheaper but the price advantage gets reduced. I haven't really seen that me…

You could pick a Ryzen with "G", like the Ryzen 5 2400G, though... They have an integrated GPU.

I know of the G series and I know why they are usually a GEN behind specially given the quality of discrete graphics AMD has and they need to sale. But for work, for just productivity a lot of the savings washes up with the GPU. It used to be that you get super lowend graphics card for those workload but these days almost the lowest price GPU is ~90.

Re: AMD Ryzen 3000 announced

#134

Post by PixyMisa »

fencepost wrote: post #59

My question is how is their architecture doing regarding the assorted speculative execution baked-in issues and what kind of impact is there on the AMD processors compared to comparable Intel CPUs?

Phoronix is the number one site for that sort of article.

https://www.phoronix.com/scan.php?page=article&item=intel-md...

Re: AMD Ryzen 3000 announced

#135

Post by pja »

xvector wrote: post #127
fencepost wrote: post #59

My question is how is their architecture doing regarding the assorted speculative execution baked-in issues and what kind of impact is there on the AMD processors compared to comparable Intel CPUs?

Frightening lack of press around this. Basically all of Intel's and AMD's lineups are centered around high thread count right now. Completely useless if I can't reasonably enable SMT.

So far, AMD has only been exposed to the Spectre (branch cache) information leaks haven’t they? They aren’t vulnerable to Meltdown or any of the recent MDS information leaks because they don’t speculatively execute if the thread doesn’t have access rights. Intel chose to speculate regardless and fixup the register contents on instruction retirement, which is why they have had so many problems.

Re: AMD Ryzen 3000 announced

#137

Post by nikic »

vetinari wrote: post #131
ubercow13 wrote: post #111

Earlier quoted context omitted.

make -j24

Be sure to have enough RAM, so you will not run out of it. (For some reason, when building projects like LLVM with -j 16, 32 GB w/o swap may be not enough. With -j 2, it is enough, but takes an eternity).

If you build LLVM often, use the shared library build (BUILD_SHARED_LIBS=true). Most of the memory usage (and a large part of the time in incremental builds) comes from linking final artifacts if you do not use shared libraries.

Re: AMD Ryzen 3000 announced

#138

Post by wtallis »

jaytaylor wrote: post #119

It's not clear to me from TFA: Do any existing AM4 mobos / chipsets have support for full PCIe 4.0 bandwith (64Gbps)? Or will the existing mobos be limited to PCIe 3.0 (~5-6Gbps)? All of the five processors will be PCIe 4.0 enabled, and while they are being accompanied by the new X570 chipset launch, they still use the same AM4 socket, meaning some AMD 300 and 400- series motherboards can still be used. I was just re…

Existing 300 and 400 series boards may be able to operate at PCIe 4 speed for the CPU-provided lanes (as opposed to the ones routed through the chipset that you can't upgrade), however signal integrity issues may limit this to just the slot closest to the CPU. So far, I haven't heard about any particular boards that have been validated for a specific number of slots working at gen4 speeds. Whatever you're using for an SSD RAID array will probably get in the way of using gen4 speeds, since you likely won't be able to get gen4 speeds over any cables or risers without redriver chips.

Re: AMD Ryzen 3000 announced

#139

Post by arianvanp »

ianai wrote: post #25
sadris wrote: post #20

Earlier quoted context omitted.

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

Read that often here tonight. Is there a killer app for single core performance other than UI/UX?

Most Flight simulators or any physics heavy game gains a lot from single-threaded performance

Re: AMD Ryzen 3000 announced

#140

Post by skrebbel »

tmd83 wrote: post #105

I'm really looking forward to this. But one issue with current AMD cpu is that you have to buy a desktop GPU even when you are not doing any kind of gaming and such. I know intel iGPU haven't been terribly good but for work they are good enough and is one less part and cheaper to boot. For same performance a Ryzen 3rd gen + gpu might still be cheaper but the price advantage gets reduced. I haven't really seen that me…

When I put together my computer a year or two ago, I just added some cheap $40 radeon board. It really didn't impact the bottom line much.
Post reply on HN