Intel a100 cpu: Stealey cpu | Processor | Nanotechnology Products

Stealey cpu | Processor | Nanotechnology Products

ADVANCED SEARCH

  • Home

  • Products

  • Stealey cpu

Manufacturer

Country

Document

datasheet 0.96 MB
Registration Date 6 Nov 2015
Share

Electronics
Processor
Product Number : A110

Processor

Applications

Computers

Computers

Properties

90 nm process technology
Physical Address Extensions 32-bit

Manufacturer’s Description

General information
Type CPU / Microprocessor
Market segment Mobile
Family Intel A100 Series
Model number  ?  A110
CPU part number
  • UM80536UC800512 is an OEM/tray microprocessor
Frequency  ?  800 MHz
Bus speed  ?  400 MHz
Clock multiplier  ?  8
Package 663-ball micro-FCBGA
0. 55″ x 0.75″ (14 mm x 19 mm)
Socket BGA663
Introduction date Apr 18, 2007
S-spec numbers
  ES/QS processors Production processors
Part number QSEJ SLAWA
UM80536UC800512 + +
Architecture / Microarchitecture
Microarchitecture Mobile
Processor core  ?  Stealey
Core stepping  ?  C0 (QSEJ)
CPUID 6D8 (QSEJ)
Manufacturing process 0.09 micron
Data width 32 bit
The number of cores 1
The number of threads 1
Floating Point Unit Integrated
Level 1 cache size  ?  32 KB instruction cache
32 KB write-back data cache
Level 2 cache size  ?  512 KB
Multiprocessing Uniprocessor
Features
  • MMX instructions
  • SSE / Streaming SIMD Extensions
  • SSE2 / Streaming SIMD Extensions 2
  • NX / XD / Execute disable bit  ? 
Low power features
  • AutoHALT mode  ? 
  • Stop Grant mode  ? 
  • Sleep mode  ? 
  • Deep Sleep mode  ? 
  • Deeper Sleep mode  ? 
  • Dynamic FSB power down
  • Enhanced SpeedStep technology  ? 
Integrated peripherals / components
Integrated graphics None
Electrical / Thermal parameters
V core  ?  0. 796V — 0.94V
Minimum/Maximum operating temperature  ?  0°C — 100°C
Minimum/Maximum power dissipation  ?  2.73 Watt (Deeper Sleep mode) / 7.35 Watt
Thermal Design Power  ?  3 Wat

Nvidia ditches Intel, cozies up to AMD with its new DGX A100

no, it won’t play crysis. stop asking —

Nvidia’s first Ampere hardware is headed for the data center, not the game room.

Jim Salter

  • This fully assembled DGX A100 node looks curiously lonely, all by itself without racks and wires and blinkenlights.

    Nvidia

Yesterday at Nvidia’s GPU Technology Conference, everybody found out what CEO Jensen Huang was cooking—an Ampere-powered successor to the Volta-powered DGX-2 deep learning system.

On Wednesday, we described mysterious hardware in Huang’s kitchen as likely «packing a few Xeon CPUs» in addition to the new successor to the Tesla v100 GPU. Egg’s on our face for that one—the new system packs a pair of AMD Epyc 7742 64-core, 128-thread CPUs, along with 1TiB of RAM, a pair of 1.9TiB NVMe SSDs in RAID1 for a boot drive, and up to four 3.8TiB PCIe4.0 NVMe drives in RAID 0 as secondary storage.

Goodbye Intel, hello AMD

  • It seems entirely likely that Nvidia didn’t want to monetarily support Intel’s plans to muscle in on its own profitable deep-learning turf.

    Intel Corporation

  • The Intel DG1 at least exists—but its leaked benchmarks don’t seem impressive, hinting at performance similar to years-old Nvidia GPUs.

    Intel Corporation

  • |

    Share

      Via, formerly one of the largest processor manufacturers, has developed a professional 7nm video accelerator for HPC, machine learning and other applications. It is a direct competitor to AMD MI100 and Nvidia A100, and in some calculations Via’s creation noticeably outperforms them.

      Via Super Video Card

      Via, the main competitor of Intel and AMD in the processor market in the distant past, is preparing to take away a significant part of the video card market from Nvidia and AMD. She has her own general-purpose graphics chip (General-purpose computing on graphics processing units, GPGPU), capable of overtaking the top-end developments of these companies in terms of performance. Via does not compete with Intel in this segment, since Intel, although it returned to the video card market many years later, still cannot offer anything that attracts attention.

      The new Via chip was codenamed Big Island, and it is possible that in the future it will become part of the accelerator series of the same name. It meets modern requirements for the production of GPUs, as it is produced on a 7-nanometer processor (this is done by the Taiwanese TSMC, which also produces AMD and Nvidia chips). The processor will be released under the brand Zhaoxin, owned by the joint venture of the same name, founded in 2013 by Via and the authorities of Shanghai (China).

      Having missed the processor market, Via wants to capture the video accelerator segment

      Big Island was developed by Via in a joint venture with the Chinese authorities from 2018 to 2020. The chip uses 2.5D CoWoS packaging and contains at least 24 billion transistors. Just like similar developments from AMD and Nvidia, Big Island received 32 GB of high-speed HBM2 memory (1.2 TB per second bandwidth) and became compatible with the PCIe 4.0 standard.

      Performance comparison

      According to TechSpot, the Big Island chip is not intended for use in conventional computers and is not suitable for video games and other entertainment. Instead, it is focused on use in machine learning, high performance computing (HPC), medical research, and security systems.

      Via is sure that its video chip is better than Nvidia and AMD

      analogs

      This means that you should not compare it with Radeon and GeForce, graphics cards for laptops and desktop PCs, but with specialized solutions from Nvidia and AMD for HPC and other listed industries. Nvidia has the A100 accelerator for this, AMD has the Instinct MI100.

      The performance of the Chinese GPGPU declared by the developer seems impressive. For example, when calculating FP16 (at half precision), the chip can deliver 147 teraflops, which is significantly better than the Nvidia F100 with its 78 teraflops. To AMD MI100 with 184.6 teraflops, the Chinese development is a little short.

      Nvidia A100

      In terms of FP32 computing capabilities, the Big Island should deliver up to 37 teraflops, which is higher than the A100 and MI100, which offer 19.5 and 23.1 teraflops, respectively.

      No deadlines

      Developers reported that Big Island is not only faster than competitors in some calculations, but also consumes less power (300 W) in comparison. Also, their statement refers to a lower price compared to their background.

      The creators of the chip did not fail to compare it with the products of Chinese companies. According to their calculations, the potential for commercial use of Big Island is one or two years ahead of the capabilities of «domestic counterparts. » What kind of developments are in question, they did not specify.

      Via clearly has big plans to distribute their new creation

      Zhaoxin has not announced the timing of the start of mass production of the chip. The exact prices of the final products also remain unknown.

      Leave to return

      In the recent past, Via, founded in 1987, was one of the three major players in the processor and chipset market along with Intel and AMD. It seriously competed with them and even released its own chipsets for Intel and AMD solutions, which caused dissatisfaction with these companies.

      Is there life after VMware?

      import independence

      The peak of popularity of Via products in the computer segment fell on the 90s of the 20th century and at the beginning of the 21st century, and it really posed such a serious threat to its competitors that Intel went to the trick and decided to fight it not by releasing more interesting products, but by litigation lawsuits. For example, in September 2001, it sued Via for patent infringement in an effort to force Via’s new Pentium 4 chipset out of the market.

      Via’s reaction was swift — the company filed a countersuit, claiming that Intel’s 845 chipset for the P4 processor violated its intellectual property rights.

      As of April 2021, Via still exists. She is developing processors as part of a joint venture with the Shanghai authorities Zhaoxin. For example, in April 2020, as part of this company, Via announced the KaiXian KX-U6780A desktop processor, comparable in performance to AMD Ryzen 3 3000 and 7th generation Intel Core i5. The outdated process technology (16 nanometers) and the lack of a number of modern technologies prevent it from competing with more modern models, but this is only one of Via’s first attempts over the past few years to create processors specifically for the consumer segment.

      Via’s developments attract the attention of major manufacturers of computer equipment.