Ugh CES again. I always hate that time of year mainly because of the number of low-value news posts about each and every little tiny thing that some company announces there since on one wants to summarize (which is fair - artificial inflation of page views is a thing people have to do to keep their jobs).
Can anyone with the knowledge of details behing this tech explain here why the heck while the main source of heat is the controller they still keep using 12-14nm and not say 5-7nm.
Also is this supposedly to be the news if one can read one year ago: "China’s Domestic NVMe SSD Controller Manufacturer To Launch 14.5 GB/s PCIe Gen 5.0 Solution In 2023..." or another Chinese firm: "Yingren Technology's YR S900 PCIe 5.0 SSD controller has commenced mass production..."
I think there's more to an SSD controller than (temporarily) maxing out the PCIe bandwidth. That's probably more obvious in some of the more detailed articles about these controllers
Probably because E26 came out in January 2023, and they only started showing up around June 2023. And if you recall, 2020~2023 had a global chip shortage / high demand for the newer nodes.
Their E31T, announced in May 2023, uses 7nm. I would speculate they do have an E26 using 7nm in the works.
Scaling down buses is not the same as scaling down logic. Obviously SSD controllers have plenty of logic, but it's fair for a controller to not be brought forward to a newer node if only a little bit of the die benefits from the new node, and the yields are bad relative to a more mature node.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
8 Comments
Back to Article
PeachNCream - Friday, December 8, 2023 - link
Ugh CES again. I always hate that time of year mainly because of the number of low-value news posts about each and every little tiny thing that some company announces there since on one wants to summarize (which is fair - artificial inflation of page views is a thing people have to do to keep their jobs).ballsystemlord - Friday, December 8, 2023 - link
I only wish that they'd wait for full specs instead of these teaser articles which are rarely followed up on.SanX - Friday, December 8, 2023 - link
Can anyone with the knowledge of details behing this tech explain here why the heck while the main source of heat is the controller they still keep using 12-14nm and not say 5-7nm.SanX - Friday, December 8, 2023 - link
Also is this supposedly to be the news if one can read one year ago: "China’s Domestic NVMe SSD Controller Manufacturer To Launch 14.5 GB/s PCIe Gen 5.0 Solution In 2023..."or another Chinese firm:
"Yingren Technology's YR S900 PCIe 5.0 SSD controller has commenced mass production..."
nandnandnand - Friday, December 8, 2023 - link
I think there's more to an SSD controller than (temporarily) maxing out the PCIe bandwidth. That's probably more obvious in some of the more detailed articles about these controllersmeacupla - Friday, December 8, 2023 - link
Probably because E26 came out in January 2023, and they only started showing up around June 2023.And if you recall, 2020~2023 had a global chip shortage / high demand for the newer nodes.
Their E31T, announced in May 2023, uses 7nm. I would speculate they do have an E26 using 7nm in the works.
shabby - Monday, December 11, 2023 - link
More expensive process = less profit = less ivory back scratchers.lmcd - Monday, December 11, 2023 - link
Scaling down buses is not the same as scaling down logic. Obviously SSD controllers have plenty of logic, but it's fair for a controller to not be brought forward to a newer node if only a little bit of the die benefits from the new node, and the yields are bad relative to a more mature node.