The 990 EVO Plus and 9100 PRO, two new consumer SSDs that will presumably leverage either the PCIe Gen 4 or Gen 5 interfaces, is currently in development at Samsung, according to new trademark filings that have surfaced from the Korea Intellectual Property Rights Information Service, a group that covers domestic and foreign industrial property information, including patents, utility models, trademarks, and more. Samsung launched the 990 EVO in January, while the 990 PRO, an SSD that comes in a heatsink option and boasts what the company says is best-in-class read speeds, launched in September last year.
Links to the trademarks:
From a report:
- “Samsung might now plan on closing the gap between the 990 EVO and the 990 PRO by releasing a 990 EVO Plus model.”
- “Samsung also trademarked the 9100 PRO name at KIPRIS [but it ]doesn’t follow Samsung’s usual naming schemes for storage solutions, be it SSDs, USB sticks, or memory cards.”
- “Unlike the 990 EVO Plus, the 9100 PRO application also includes designated goods such as semiconductor devices, semiconductors, semiconductor memories, and semiconductor memory units.”
- “…one possible explanation as to why it exists might be that the 990 EVO Plus is an SSD and the 9100 PRO is an SSD component, like a memory chip.”
The trademarks:
Some promos for Samsung’s newer SSDs:
Samsung on the SSD 990 PRO:
Offering blazing-fast speeds and ultimate power efficiency, the 990 PRO series is optimized for massive data volumes, such as 3D/4K graphics work, data analytics and high-quality games, making it the ideal SSD for today’s PCs, laptops, game consoles and computing systems. With improved total bytes written (TBW) ratings of up to 2,400TB, the 990 PRO series ensures increased SSD reliability and longevity, ideal for those with highly demanding workloads and large storage capacity needs.
Join the discussion in The FPS Review Forums...
Discussion (3 replies)
Join Discussion →
Nice... maybe I'll get one 3 months after everyone else gets them to patch the problem that will be there at release. ;)
I hope they start focusing on raising the 4k random read speeds, as this ever increasing max bandwidth with little impact on actual system performance is getting tiring.
Give me something that performs more and more like an Intel Optane AIC and I will be a lot happier.
"Zarathustra, post: 86552, member: 203" wrote:I hope they start focusing on rising the 4k random read speeds, as this ever increasing max bandwidth with little impact on actual system performance is getting tiring.
Give me something that performs more and more like an Intel Optane AIC and I will be a lot happier.
What you want is effectively an HBA controller that you can slot in a 32 gig cache memory module into and a pair of PCIE Gen whatever NVME drives. Run them as a raid 1 on a good controller in a PCIE X 8 slot and you'll have your speed.
Speaking from a server specific standpoint a good raid controller in front of even a set of 8 NVME drives running at pcie 4 will get you some nice random read numbers. Only catch is to truly test it your test size will need to exceed the cache value. :) Otherwise the results are not realistic and will pretty much be the same regardless of disk type.
Recently went through some rounds of testing to find out some good results for SQL because Other than BLOB data our dataset LOVES 4k random reads.
Discussion (3 replies)
Join Discussion →Nice... maybe I'll get one 3 months after everyone else gets them to patch the problem that will be there at release. ;)
I hope they start focusing on raising the 4k random read speeds, as this ever increasing max bandwidth with little impact on actual system performance is getting tiring.
Give me something that performs more and more like an Intel Optane AIC and I will be a lot happier.
What you want is effectively an HBA controller that you can slot in a 32 gig cache memory module into and a pair of PCIE Gen whatever NVME drives. Run them as a raid 1 on a good controller in a PCIE X 8 slot and you'll have your speed.
Speaking from a server specific standpoint a good raid controller in front of even a set of 8 NVME drives running at pcie 4 will get you some nice random read numbers. Only catch is to truly test it your test size will need to exceed the cache value. :) Otherwise the results are not realistic and will pretty much be the same regardless of disk type.
Recently went through some rounds of testing to find out some good results for SQL because Other than BLOB data our dataset LOVES 4k random reads.