Intel Optane's persistent memory starts at 850 USD for 128 GB



[ad_1]

Looking to the future: The Intel Optane persistent memory is now available for pre-order at $ 842 for a 128GB key and $ 2,461 for a 256GB key. This may seem like a lot, and the price is comparable to that of standard DDR4 keys. 16 GB or 32 GB, but in the world of servers, it represents one-tenth of the normal price.

Intel's Optimum Memory (OPM) is Intel's low-latency high-speed storage, placed in a form factor of memory. The presentation is simple: in case of failure of a critical system, the data stored in memory is not lost; and for memory-hungry applications, it offers surprisingly high capacity at low prices. The tradeoff is higher latency and lower raw speed compared to traditional volatile memory.

Thanks to two US online retailers who have pre-ordered the memory, we now know what the prices are. Although it is possible to buy conventional 128GB and 256GB USB sticks, they reach extreme levels with respectively $ 4,326 and $ 24,257. OPM costs 80 to 90% cheaper and adds 512 GB to the solution, although we do not know yet the price of this module.

128 GB Price / GB 256 GB Price / GB
OPM at CompSource $ 893 $ 6.98 $ 2,461 $ 9.61
OPM at ShopBLT $ 842 $ 6.58 $ 2,668 $ 10.42
Standard single stick $ 4,326 $ 33.80 $ 24,257 $ 94.75
Standard multi-stick $ 699 $ 5.46 $ 1,816 $ 7.09
Standard M.2 $ 51.99 $ 0.41 $ 87.99 $ 0.34

As cool as OPM is, it's not at all meant to reach consumers at this point. Even if a professional needs 256 GB of RAM, an 8 x 32 GB DDR4 kit costs less at $ 1,816. It's much faster too. OPM occupies its niche in a fantastic way, but it will take a long time before it can expand to other markets.

Servers with Intel OPM technology should arrive in June, but we do not know when individual keys will be available, and retailers do not offer release dates. Performance is also an important issue, not in terms of numbers and specifications, but how different workloads will be affected by lower speeds and increased capacity over traditional memory.

[ad_2]

Source link