Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
SlapActionJackson
Jul 27, 2006

And out of curiosity, what would you do with a 7950X3D?

Adbot
ADBOT LOVES YOU

SlapActionJackson
Jul 27, 2006

Linux drivers needed to read the thermal sensors in RPL are new, so very well might not be working right, especially depending on the kernel version you're on.

SlapActionJackson
Jul 27, 2006

You aren't going to get the memory controller up to 8gig

SlapActionJackson
Jul 27, 2006

Dr. Video Games 0031 posted:

I guess this is more about how shoddy AMD's support is for four-dimm configurations in general.

There are some pretty severe technical limitations that come into play to support 2 dpc, and the performance gap between 1 and 2 is only going to grow. The whole industry would be better served by acknowledging that modern ddr has line-of-sight to signaling rates > pcie g3 and 1 dpc is the right physical topology to support that.

SlapActionJackson
Jul 27, 2006

hobbesmaster posted:

Isn’t it the same as it’s always been and the real issue is ranks per channel, not DIMMs per channel per se?

2r2dpc has always been slow.

Multirank dimms pose some of the same challenges, but at a smaller scale. It's easier to make the interface work for one 2r dimm than two 1r dimms.
It's a good compromise for enabling high memory density on desktops until the ecosystem is ready to move to rdimm.

SlapActionJackson
Jul 27, 2006

Twerk from Home posted:

Every desktop CPU is just a server or laptop part that couldn't cut it. Desktop has been bottom feeding off the waste forever.

This isn't remotely true these days.

SlapActionJackson
Jul 27, 2006

Yudo posted:

Leaky chips that can clock = consumer desktop. Scalability and reusablity is the raison d'être of AMD's designs, with good reason. Or, at least, that is what I thought.

The basic idea is right, there's design and physical re-use up and down the stack. But a host of technical and logistical issues make it more complicated than "this CCD can't cut it as a server or laptop, stick it in a desktop". At the 10,000 foot level there are tradeoffs that can be made for more or fewer desktops vs fewer or more servers, but desktop can't subsist on server's rejects.

SlapActionJackson
Jul 27, 2006

BlankSystemDaemon posted:


Now if only AMD can invent an equivalent to GF-Ni, so we can get Reed-Solomon error correction offloaded.



Where would you use this?

SlapActionJackson
Jul 27, 2006

You ain't getting to DRAM in 19ns.

SlapActionJackson
Jul 27, 2006

They ain't building any AM4 anymore.

Adbot
ADBOT LOVES YOU

SlapActionJackson
Jul 27, 2006

Secure boot absolutely requires you to lock to something as a root of trust. The platform vendor is one possible sane option. Like all security measures, it involves usability tradeoffs.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply