Feeds

back to article Fujitsu Integrated Systems stack up servers, storage, and switches

Japanese IT giant Fujitsu will roll out a bounty of new – and, of course, cloudy – hardware and software offerings next week at its annual customer and partner confab in Tokyo. Fujitsu was the first and only Microsoft Azure partner that got a public cloud puffed up based on the Windows stack and its own iron that is compatible …

COMMENTS

This topic is closed for new posts.
Bronze badge

where do they get these numbers

Up to 300 VMs in a rack ?? 300 doing what? If I look at my own VM infrastructure and that of my previous companies for example we could probably get upwards of 1000-2000 VMs in a rack depending on server hardware config. The constraint is because of memory, not CPU and for the most part not I/O (relatively speaking). Like most physical servers, the virtual servers spend *most* of their lives idle from a CPU and I/O perspective. Though applications are always running and consuming memory so memory requirements are high relative to CPU and I/O.

Currently in the process of doubling the memory in our servers to 384GB (dual socket Opteron 6100), as the CPUs run at under 15% on average with peaks not being much higher. This workload isn't specific to my present position - I look back over the past decade at all the companies I have worked at and things are similar with very few exceptions(those few exceptions would have more than enough aggregate CPU capacity to burst in utilization without requiring a special configuration).

Are these 300 VMs running VMark or something is that where they get the data from? If not then how do these folks (and others too, not picking on Fujitsu specifically I see similar claims from other vendors constantly especially with these integrated solutions).

Just curious.

0
0

Re: where do they get these numbers

They are probably thinking about the higher utilisation you (should get) with public clouds.

Lets assume that instead of you buying a 'rack' of servers you rent each *VM* for $0.30 per hour that it is active. You are now likely to design your systems to have fewer VMs active at any one time, but when they are active use them more.

0
0

This post has been deleted by its author

Ok, here some facts derived from spec.org

Highest benchmark results for single system images is available with

the SPARC64 X 3GHz Architecture

which shows the best scalability, currently available up to 1024 cores in one system.

Top Performer in Spec2006 Benchmarks with 1024 cores

Benchmark "Hardware Vendor " System # Cores Result

CINT2006rate Fujitsu Fujitsu SPARC M10-4S SPARC64 X 3GHz 1024 23800,00

CFP2006rate Fujitsu Fujitsu SPARC M10-4S SPARC64 X 3GHz 1024 19700,00

For around 16 cores the Intel E5-2690 2.90 GHz still delivers best results (not counting the Cryo system, which was running under special cooling conditions at extreme clock rates).

Top 5 CINT2006 16 core

Benchmark "Hardware Vendor " System # Cores Result

CINT2006 Cryo Performance Computing Ltd Cryo Quad EUP-RO 4 61,70

CINT2006 Fujitsu PRIMERGY BX924 S3, Intel Xeon E5-2690, 2.90 GHz 16 60,80

CINT2006 Fujitsu PRIMERGY RX200 S7, Intel Xeon E5-2690, 2.90 GHz 16 60,70

CINT2006 Fujitsu PRIMERGY RX350 S7, Intel Xeon E5-2690, 2.90 GHz 16 60,70

CINT2006 Fujitsu PRIMERGY TX300 S7, Intel Xeon E5-2690, 2.90 GHz 16 60,70

Top 5 CFP2006 16 core

Benchmark "Hardware Vendor " System # Cores Result

CFP2006 Huawei Huawei CH121 V1 (Intel Xeon E5-2690) 16 97,30

CFP2006 Huawei Huawei RH2288 V2 (Intel Xeon E5-2690) 16 97,30

CFP2006 Oracle Sun Fire X4170 M3 (Intel Xeon E5-2690 2.9GHz) 16 96,80

CFP2006 Oracle Sun Server X3-2 (Intel Xeon E5-2690 2.9GHz) 16 96,80

CFP2006 Oracle Sun Blade X3-2B (Intel Xeon E5-2690 2.9GHz) 16 96,00

For around 128 cores the IBM Power Architecture still delivers best results.

Top 5 CINT2006rate 128 core (showing best results per system if more than one compiler and OS)

Benchmark "Hardware Vendor " System # Cores Result

CINT2006rate IBM Corporation IBM Power 795 (4.25 GHz) 128 6150,00

CINT2006rate IBM Corporation IBM Power 780 (3.7 GHz, RHEL) 128 6130,00

CINT2006rate Oracle Corporation SPARC T5-8 3.6GHz 128 3750,00

CINT2006rate Bull SAS bullx S6030 (Intel Xeon X7560 2.3GHz) 128 3050,00

CINT2006rate Fujitsu SPARC Enterprise M9000 SPARC64 VII+ 3.0GHz 128 1720,00

Top 5 CFP2006rate 128 core (showing best results per system if more than one compiler and OS)

Benchmark "Hardware Vendor " System # Cores Result

CFP2006rate IBM Corporation IBM Power 795 (4.25 GHz, 128 core) 128 5870,00

CFP2006rate IBM Corporation IBM Power 780 (3.7 GHz, 128 core, RHEL) 128 4180,00

CFP2006rate Oracle Corporation SPARC T5-8 3.6GHz 128 3020,00

CFP2006rate Bull SAS bullx S6030 (Intel Xeon X7560) 128 2050,00

CFP2006rate SGI Altix 4700 Bandwidth System (Itanium 9150M 1.66GHz/24M) 128 1950,00

For around 256 cores the IBM Power Architecture still delivers best results.

Top 5 CINT2006rate 256 core (showing best results per system if more than one compiler and OS)

Benchmark "Hardware Vendor " System # Cores Result

CINT2006rate IBM Corporation IBM Power 795 (4.0 GHz, RedHat) 256 11300,00

CINT2006rate Fujitsu Fujitsu SPARC M10-4S SPARC64 X 3GHz 256 6350,00

CINT2006rate SGI Altix 4700 Density System (Itanium 9150M 1.66GHz/24M) 256 3350,00

CINT2006rate Fujitsu SPARC Enterprise M9000 SPARC64 VII+ 3.0GHz 256 3150,00

CINT2006rate SGI Altix 4700 Bandwidth System (Itanium 2 9040 1.6GHz/18M) 256 2970,00

Top 5 CFP2006rate 256 core (showing best results per system if more than one compiler and OS)

Benchmark "Hardware Vendor " System # Cores Result

CFP2006rate IBM Corporation IBM Power 795 (4.0 GHz, 256 core) 256 10500,00

CFP2006rate Fujitsu Fujitsu SPARC M10-4S SPARC64 X 3GHz 256 5860,00

CFP2006rate SGI Altix 4700 Bandwidth System (Itanium 2 9040 1.6GHz/18M) 256 3510,00

CFP2006rate Fujitsu SPARC Enterprise M9000 SPARC64 VII+ 3.0GHz 256 2550,00

CFP2006rate Fujitsu SPARC Enterprise M9000 SPARC64 VII 2.5GHz 256 2100,00

0
0
This topic is closed for new posts.