skip to main |
skip to sidebar
8:19:00 AM
valgeo
If Microsoft CEO Steve Ballmer is to be forever followed around by his
'developers, developers, developers' chant, now feels like the right
time for Amazon to be known for a call of 'iteration, iteration,
iteration'.
On Thursday the company cut prices in its two main datacentre regions
and released a new range of powerful 'Standard' rentable computers,
according to a post on its blog. These
two announcements follow Amazon's development rhythm over the past year
or so, which has seen around 10 significant announcements for the
service per month.
The two Second Generation M3 instances have more memory and computing
power than M1 rentable servers, but are restricted to using Amazon's
Elastic Block Store (EBS) technology by default. EBS has come in for a
good deal of criticism, with a former administrator of major social site
Reddit describing it as being "a barrel of laughs in terms of performance and reliability."
"With up to 50 percent higher absolute CPU performance, these
instances are optimised for applications such as media encoding, batch
processing, caching, and web serving," the company wrote in its blog
post.
The M3 Extra Large Instance has 15GB of memory and 13 EC2 Compute
Units (ECU) spread across four virtual cores, while the M3 Double Extra
Large Instance has 30GB of memory and 26 ECU across eight virtual cores.
An ECU is roughly equivalent to the CPU capacity of a one to 1.2Ghz
2007 AMD Opteron or Intel Xeon processor, according to Amazon.
In addition, Amazon has cut the price of rentable compute instances
by 18 percent in its main US East (Northern Virginia) datacentre hub,
along with its US West (Oregon) hub. As the Virginia hub is Amazon's
largest, it is also the one that has experienced the greatest amount of failures.
0 σχόλια:
Post a Comment